ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 365 - (show annotations)
Fri Jul 11 17:06:55 2008 UTC (13 years, 2 months ago) by ph10
File size: 159129 byte(s)
Further fixes for bumpalong processing in UTF-8 mode.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.0 02 Jul-08
5 ---------------------
7 1. Replaced UCP searching code with optimized version as implemented for Ad
8 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
9 stage table and inline lookup instead of a function, giving speed ups of 2
10 to 5 times on some simple patterns that I tested. Permission was given to
11 distribute the MultiStage2.py script that generates the tables (it's not in
12 the tarball, but is in the Subversion repository).
14 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
15 scripts.
17 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
18 a group with a zero qualifier. The result of the study could be incorrect,
19 or the function might crash, depending on the pattern.
21 4. Caseless matching was not working for non-ASCII characters in back
22 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
23 It now works when Unicode Property Support is available.
25 5. In pcretest, an escape such as \x{de} in the data was always generating
26 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
27 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
28 truncation.
30 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
32 7. Added two (int) casts to pcregrep when printing the difference of two
33 pointers, in case they are 64-bit values.
35 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
36 test 2 if it fails.
38 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
39 and a #define of that name to empty if it is not externally set. This is to
40 allow users of MSVC to set it if necessary.
42 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
43 the convenience functions in the pcre_get.c source file.
45 11. An option change at the start of a pattern that had top-level alternatives
46 could cause overwriting and/or a crash. This command provoked a crash in
47 some environments:
49 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
51 This potential security problem was recorded as CVE-2008-2371.
53 12. For a pattern where the match had to start at the beginning or immediately
54 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
55 pcre_dfa_exec() could read past the end of the passed subject if there was
56 no match. To help with detecting such bugs (e.g. with valgrind), I modified
57 pcretest so that it places the subject at the end of its malloc-ed buffer.
59 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
60 exec() might read past the end of the data buffer in UTF-8 mode.
62 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
63 the data contained the byte 0x85 as part of a UTF-8 character within its
64 first line. This applied both to normal and DFA matching.
67 Version 7.7 07-May-08
68 ---------------------
70 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
71 a string to a long long, pretend we don't even have a long long." This is
72 done by checking for the strtoq, strtoll, and _strtoi64 functions.
74 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
75 pre-7.6 versions, which defined a global no_arg variable instead of putting
76 it in the RE class. (See also #8 below.)
78 3. Remove a line of dead code, identified by coverity and reported by Nuno
79 Lopes.
81 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
83 (1) The include/exclude patterns were being applied to the whole pathnames
84 of files, instead of just to the final components.
86 (2) If there was more than one level of directory, the subdirectories were
87 skipped unless they satisfied the include/exclude conditions. This is
88 inconsistent with GNU grep (and could even be seen as contrary to the
89 pcregrep specification - which I improved to make it absolutely clear).
90 The action now is always to scan all levels of directory, and just
91 apply the include/exclude patterns to regular files.
93 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
94 --exclude_dir in the tests to avoid scanning .svn directories.
96 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
97 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
98 doesn't support NULs in patterns.
100 7. Added some missing "const"s to declarations of static tables in
101 pcre_compile.c and pcre_dfa_exec.c.
103 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
104 caused by fix #2 above. (Subsequently also a second patch to fix the
105 first patch. And a third patch - this was a messy problem.)
107 9. Applied Craig's patch to remove the use of push_back().
109 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
110 matching function regexec().
112 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
113 which, however, unlike Perl's \g{...}, are subroutine calls, not back
114 references. PCRE supports relative numbers with this syntax (I don't think
115 Oniguruma does).
117 12. Previously, a group with a zero repeat such as (...){0} was completely
118 omitted from the compiled regex. However, this means that if the group
119 was called as a subroutine from elsewhere in the pattern, things went wrong
120 (an internal error was given). Such groups are now left in the compiled
121 pattern, with a new opcode that causes them to be skipped at execution
122 time.
124 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
125 to the way PCRE behaves:
127 (a) A lone ] character is dis-allowed (Perl treats it as data).
129 (b) A back reference to an unmatched subpattern matches an empty string
130 (Perl fails the current match path).
132 (c) A data ] in a character class must be notated as \] because if the
133 first data character in a class is ], it defines an empty class. (In
134 Perl it is not possible to have an empty class.) The empty class []
135 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
136 The negative empty class [^] matches any one character, independently
137 of the DOTALL setting.
139 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
140 non-existent subpattern following a character class starting with ']' and
141 containing () gave an internal compiling error instead of "reference to
142 non-existent subpattern". Fortunately, when the pattern did exist, the
143 compiled code was correct. (When scanning forwards to check for the
144 existencd of the subpattern, it was treating the data ']' as terminating
145 the class, so got the count wrong. When actually compiling, the reference
146 was subsequently set up correctly.)
148 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
149 it was being rejected as not supported by pcre_dfa_exec(), even though
150 other assertions are supported. I have made pcre_dfa_exec() support
151 (*FAIL).
153 16. The implementation of 13c above involved the invention of a new opcode,
154 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
155 cannot be changed at match time, I realized I could make a small
156 improvement to matching performance by compiling OP_ALLANY instead of
157 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
158 on the OP_ANY path.
160 17. Compiling pcretest on Windows with readline support failed without the
161 following two fixes: (1) Make the unistd.h include conditional on
162 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
164 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
165 ncurses library to be included for pcretest when ReadLine support is
166 requested, but also to allow for it to be overridden. This patch came from
167 Daniel Bergström.
169 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
170 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
171 any errors with the current Unicode tables. Thanks to Peter Kankowski for
172 spotting this.
175 Version 7.6 28-Jan-08
176 ---------------------
178 1. A character class containing a very large number of characters with
179 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
180 overflow.
182 2. Patch to cut out the "long long" test in pcrecpp_unittest when
183 HAVE_LONG_LONG is not defined.
185 3. Applied Christian Ehrlicher's patch to update the CMake build files to
186 bring them up to date and include new features. This patch includes:
188 - Fixed PH's badly added libz and libbz2 support.
189 - Fixed a problem with static linking.
190 - Added pcredemo. [But later removed - see 7 below.]
191 - Fixed dftables problem and added an option.
192 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
194 - Added readline support for pcretest.
195 - Added an listing of the option settings after cmake has run.
197 4. A user submitted a patch to Makefile that makes it easy to create
198 "pcre.dll" under mingw when using Configure/Make. I added stuff to
199 Makefile.am that cause it to include this special target, without
200 affecting anything else. Note that the same mingw target plus all
201 the other distribution libraries and programs are now supported
202 when configuring with CMake (see 6 below) instead of with
203 Configure/Make.
205 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
206 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
207 exported in the Windows port". It has not yet been confirmed that the patch
208 solves the problem, but it does no harm.
210 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
211 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
212 with CMake, and also correct the comment about stack recursion.
214 7. Remove the automatic building of pcredemo from the ./configure system and
215 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
216 of a program that users should build themselves after PCRE is installed, so
217 building it automatically is not really right. What is more, it gave
218 trouble in some build environments.
220 8. Further tidies to CMakeLists.txt from Sheri and Christian.
223 Version 7.5 10-Jan-08
224 ---------------------
226 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
227 values in parens when parsing an RE using the C++ wrapper."
229 2. Negative specials like \S did not work in character classes in UTF-8 mode.
230 Characters greater than 255 were excluded from the class instead of being
231 included.
233 3. The same bug as (2) above applied to negated POSIX classes such as
234 [:^space:].
236 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
237 defined or documented. It seems to have been a typo for PCRE_STATIC, so
238 I have changed it.
240 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
241 first named subpattern) and a construct such as (?&a) would reference the
242 first named subpattern whose name started with "a" (in other words, the
243 length check was missing). Both these problems are fixed. "Subpattern name
244 expected" is now given for (?&) (a zero-length name), and this patch also
245 makes it give the same error for \k'' (previously it complained that that
246 was a reference to a non-existent subpattern).
248 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
249 this is right because (?- can be followed by option settings as well as by
250 digits. I have, however, made the messages clearer.
252 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
253 than the number used in the conditional) now cause a compile-time error.
254 This is actually not compatible with Perl, which accepts such patterns, but
255 treats the conditional as always being FALSE (as PCRE used to), but it
256 seems to me that giving a diagnostic is better.
258 8. Change "alphameric" to the more common word "alphanumeric" in comments
259 and messages.
261 9. Fix two occurrences of "backslash" in comments that should have been
262 "backspace".
264 10. Remove two redundant lines of code that can never be obeyed (their function
265 was moved elsewhere).
267 11. The program that makes PCRE's Unicode character property table had a bug
268 which caused it to generate incorrect table entries for sequences of
269 characters that have the same character type, but are in different scripts.
270 It amalgamated them into a single range, with the script of the first of
271 them. In other words, some characters were in the wrong script. There were
272 thirteen such cases, affecting characters in the following ranges:
274 U+002b0 - U+002c1
275 U+0060c - U+0060d
276 U+0061e - U+00612
277 U+0064b - U+0065e
278 U+0074d - U+0076d
279 U+01800 - U+01805
280 U+01d00 - U+01d77
281 U+01d9b - U+01dbf
282 U+0200b - U+0200f
283 U+030fc - U+030fe
284 U+03260 - U+0327f
285 U+0fb46 - U+0fbb1
286 U+10450 - U+1049d
288 12. The -o option (show only the matching part of a line) for pcregrep was not
289 compatible with GNU grep in that, if there was more than one match in a
290 line, it showed only the first of them. It now behaves in the same way as
291 GNU grep.
293 13. If the -o and -v options were combined for pcregrep, it printed a blank
294 line for every non-matching line. GNU grep prints nothing, and pcregrep now
295 does the same. The return code can be used to tell if there were any
296 non-matching lines.
298 14. Added --file-offsets and --line-offsets to pcregrep.
300 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
301 infinitely looping recursion. The bug was that positive lookaheads were not
302 being skipped when checking for a possible empty match (negative lookaheads
303 and both kinds of lookbehind were skipped).
305 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
306 inclusion of <windows.h> to before rather than after the definition of
307 INVALID_FILE_ATTRIBUTES (patch from David Byron).
309 17. Specifying a possessive quantifier with a specific limit for a Unicode
310 character property caused pcre_compile() to compile bad code, which led at
311 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
312 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
313 caused the error; without that there was no problem.
315 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
317 19. Added --enable-pcretest-libreadline.
319 20. In pcrecpp.cc, the variable 'count' was incremented twice in
320 RE::GlobalReplace(). As a result, the number of replacements returned was
321 double what it should be. I removed one of the increments, but Craig sent a
322 later patch that removed the other one (the right fix) and added unit tests
323 that check the return values (which was not done before).
325 21. Several CMake things:
327 (1) Arranged that, when cmake is used on Unix, the libraries end up with
328 the names libpcre and libpcreposix, not just pcre and pcreposix.
330 (2) The above change means that pcretest and pcregrep are now correctly
331 linked with the newly-built libraries, not previously installed ones.
335 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
336 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
337 UTF-8 newline character). The key issue is that the pattern starts .*;
338 this means that the match must be either at the beginning, or after a
339 newline. The bug was in the code for advancing after a failed match and
340 checking that the new position followed a newline. It was not taking
341 account of UTF-8 characters correctly.
343 23. PCRE was behaving differently from Perl in the way it recognized POSIX
344 character classes. PCRE was not treating the sequence [:...:] as a
345 character class unless the ... were all letters. Perl, however, seems to
346 allow any characters between [: and :], though of course it rejects as
347 unknown any "names" that contain non-letters, because all the known class
348 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
349 for example, whereas PCRE did not - it did not recognize a POSIX character
350 class. This seemed a bit dangerous, so the code has been changed to be
351 closer to Perl. The behaviour is not identical to Perl, because PCRE will
352 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
353 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
354 Perl does, and where it didn't before.
356 24. Rewrite so as to remove the single use of %n from pcregrep because in some
357 Windows environments %n is disabled by default.
360 Version 7.4 21-Sep-07
361 ---------------------
363 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
364 means that a class such as [\s] counted as "explicit reference to CR or
365 LF". That isn't really right - the whole point of the change was to try to
366 help when there was an actual mention of one of the two characters. So now
367 the change happens only if \r or \n (or a literal CR or LF) character is
368 encountered.
370 2. The 32-bit options word was also used for 6 internal flags, but the numbers
371 of both had grown to the point where there were only 3 bits left.
372 Fortunately, there was spare space in the data structure, and so I have
373 moved the internal flags into a new 16-bit field to free up more option
374 bits.
376 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
377 but did not set the internal JCHANGED flag - either of these is enough to
378 control the way the "get" function works - but the PCRE_INFO_JCHANGED
379 facility is supposed to tell if (?J) was ever used, so now (?J) at the
380 start sets both bits.
382 4. Added options (at build time, compile time, exec time) to change \R from
383 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
385 5. doc/pcresyntax.html was missing from the distribution.
387 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
388 compatibility, even though it is no longer used.
390 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
391 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
392 windows.h file is present (where different names are used). [This was
393 reversed later after testing - see 16 below.]
395 8. Changed all #include <config.h> to #include "config.h". There were also
396 some further <pcre.h> cases that I changed to "pcre.h".
398 9. When pcregrep was used with the --colour option, it missed the line ending
399 sequence off the lines that it output.
401 10. It was pointed out to me that arrays of string pointers cause lots of
402 relocations when a shared library is dynamically loaded. A technique of
403 using a single long string with a table of offsets can drastically reduce
404 these. I have refactored PCRE in four places to do this. The result is
405 dramatic:
407 Originally: 290
408 After changing UCP table: 187
409 After changing error message table: 43
410 After changing table of "verbs" 36
411 After changing table of Posix names 22
413 Thanks to the folks working on Gregex for glib for this insight.
415 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
416 unicode-properties was also set.
418 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
420 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
421 checked only for CRLF.
423 14. Added casts to pcretest.c to avoid compiler warnings.
425 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
427 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
428 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
429 entirely. This removes changes made in 7 above.
431 17. The CMake files have been updated, and there is now more information about
432 building with CMake in the NON-UNIX-USE document.
435 Version 7.3 28-Aug-07
436 ---------------------
438 1. In the rejigging of the build system that eventually resulted in 7.1, the
439 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
440 brackets there is not right, since it causes compilers to look for an
441 installed pcre.h, not the version that is in the source that is being
442 compiled (which of course may be different). I have changed it back to:
444 #include "pcre.h"
446 I have a vague recollection that the change was concerned with compiling in
447 different directories, but in the new build system, that is taken care of
448 by the VPATH setting the Makefile.
450 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
451 when the subject happened to end in the byte 0x85 (e.g. if the last
452 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
453 characters but of course it shouldn't be taken as a newline when it is part
454 of another character. The bug was that, for an unlimited repeat of . in
455 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
456 characters when looking for a newline.
458 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
460 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
461 in debug output.
463 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
464 long printing in the pcrecpp unittest when running under MinGW.
466 6. ESC_K was left out of the EBCDIC table.
468 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
469 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
470 limit also applies to "virtual nesting" when a pattern is recursive, and in
471 this case 1000 isn't so big. I have been able to remove this limit at the
472 expense of backing off one optimization in certain circumstances. Normally,
473 when pcre_exec() would call its internal match() function recursively and
474 immediately return the result unconditionally, it uses a "tail recursion"
475 feature to save stack. However, when a subpattern that can match an empty
476 string has an unlimited repetition quantifier, it no longer makes this
477 optimization. That gives it a stack frame in which to save the data for
478 checking that an empty string has been matched. Previously this was taken
479 from the 1000-entry workspace that had been reserved. So now there is no
480 explicit limit, but more stack is used.
482 8. Applied Daniel's patches to solve problems with the import/export magic
483 syntax that is required for Windows, and which was going wrong for the
484 pcreposix and pcrecpp parts of the library. These were overlooked when this
485 problem was solved for the main library.
487 9. There were some crude static tests to avoid integer overflow when computing
488 the size of patterns that contain repeated groups with explicit upper
489 limits. As the maximum quantifier is 65535, the maximum group length was
490 set at 30,000 so that the product of these two numbers did not overflow a
491 32-bit integer. However, it turns out that people want to use groups that
492 are longer than 30,000 bytes (though not repeat them that many times).
493 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
494 made it possible to implement the integer overflow checks in a much more
495 dynamic way, which I have now done. The artificial limitation on group
496 length has been removed - we now have only the limit on the total length of
497 the compiled pattern, which depends on the LINK_SIZE setting.
499 10. Fixed a bug in the documentation for get/copy named substring when
500 duplicate names are permitted. If none of the named substrings are set, the
501 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
502 empty string.
504 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
505 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
506 because the ] is interpreted as the first data character and the
507 terminating ] is not found. PCRE has been made compatible with Perl in this
508 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
509 cause memory overwriting.
511 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
512 string has been matched (to stop an infinite loop). It was not recognizing
513 a conditional subpattern that could match an empty string if that
514 subpattern was within another subpattern. For example, it looped when
515 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
516 condition was not nested. This bug has been fixed.
518 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
519 past the start of the subject in the presence of bytes with the top bit
520 set, for example "\x8aBCD".
522 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
523 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
525 14. Optimized (?!) to (*FAIL).
527 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
528 This restricts code points to be within the range 0 to 0x10FFFF, excluding
529 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
530 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
531 does: it's just the validity check that is more restrictive.
533 16. Inserted checks for integer overflows during escape sequence (backslash)
534 processing, and also fixed erroneous offset values for syntax errors during
535 backslash processing.
537 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
538 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
540 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
541 caused an overrun.
543 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
544 something other than just ASCII characters) inside a group that had an
545 unlimited repeat caused a loop at compile time (while checking to see
546 whether the group could match an empty string).
548 20. Debugging a pattern containing \p or \P could cause a crash. For example,
549 [\P{Any}] did so. (Error in the code for printing property names.)
551 21. An orphan \E inside a character class could cause a crash.
553 22. A repeated capturing bracket such as (A)? could cause a wild memory
554 reference during compilation.
556 23. There are several functions in pcre_compile() that scan along a compiled
557 expression for various reasons (e.g. to see if it's fixed length for look
558 behind). There were bugs in these functions when a repeated \p or \P was
559 present in the pattern. These operators have additional parameters compared
560 with \d, etc, and these were not being taken into account when moving along
561 the compiled data. Specifically:
563 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
564 length.
566 (b) An item such as \pL+ within a repeated group could cause crashes or
567 loops.
569 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
570 "reference to non-existent subpattern" error.
572 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
574 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
575 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
577 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
579 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
580 character were causing crashes (broken optimization).
582 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
583 \p or \P) caused a compile-time loop.
585 28. More problems have arisen in unanchored patterns when CRLF is a valid line
586 break. For example, the unstudied pattern [\r\n]A does not match the string
587 "\r\nA" because change 7.0/46 below moves the current point on by two
588 characters after failing to match at the start. However, the pattern \nA
589 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
590 the same is true. There doesn't seem any very clean way out of this, but
591 what I have chosen to do makes the common cases work: PCRE now takes note
592 of whether there can be an explicit match for \r or \n anywhere in the
593 pattern, and if so, 7.0/46 no longer applies. As part of this change,
594 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
595 pattern has explicit CR or LF references.
597 29. Added (*CR) etc for changing newline setting at start of pattern.
600 Version 7.2 19-Jun-07
601 ---------------------
603 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
604 which is apparently normally available under Windows.
606 2. Re-jig the pcregrep tests with different newline settings in an attempt
607 to make them independent of the local environment's newline setting.
609 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
611 4. Some of the "internals" tests were previously cut out when the link size
612 was not 2, because the output contained actual offsets. The recent new
613 "Z" feature of pcretest means that these can be cut out, making the tests
614 usable with all link sizes.
616 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
617 stack recursion. This gives a massive performance boost under BSD, but just
618 a small improvement under Linux. However, it saves one field in the frame
619 in all cases.
621 6. Added more features from the forthcoming Perl 5.10:
623 (a) (?-n) (where n is a string of digits) is a relative subroutine or
624 recursion call. It refers to the nth most recently opened parentheses.
626 (b) (?+n) is also a relative subroutine call; it refers to the nth next
627 to be opened parentheses.
629 (c) Conditions that refer to capturing parentheses can be specified
630 relatively, for example, (?(-2)... or (?(+3)...
632 (d) \K resets the start of the current match so that everything before
633 is not part of it.
635 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
637 (f) \g{name} is another synonym - part of Perl 5.10's unification of
638 reference syntax.
640 (g) (?| introduces a group in which the numbering of parentheses in each
641 alternative starts with the same number.
643 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
645 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
648 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
649 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
650 for detecting groups that can match an empty string.
652 9. A pattern with a very large number of alternatives (more than several
653 hundred) was running out of internal workspace during the pre-compile
654 phase, where pcre_compile() figures out how much memory will be needed. A
655 bit of new cunning has reduced the workspace needed for groups with
656 alternatives. The 1000-alternative test pattern now uses 12 bytes of
657 workspace instead of running out of the 4096 that are available.
659 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
661 11. Applied patch from Google to remove an optimization that didn't quite work.
662 The report of the bug said:
664 pcrecpp::RE("a*").FullMatch("aaa") matches, while
665 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
666 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
668 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
669 it matched the wrong number of bytes.
672 Version 7.1 24-Apr-07
673 ---------------------
675 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
676 that is more "standard", making use of automake and other Autotools. There
677 is some re-arrangement of the files and adjustment of comments consequent
678 on this.
680 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
681 for recursive directory scanning broke on some systems because the files
682 are not scanned in any specific order and on different systems the order
683 was different. A call to "sort" has been inserted into RunGrepTest for the
684 approprate test as a short-term fix. In the longer term there may be an
685 alternative.
687 3. I had an email from Eric Raymond about problems translating some of PCRE's
688 man pages to HTML (despite the fact that I distribute HTML pages, some
689 people do their own conversions for various reasons). The problems
690 concerned the use of low-level troff macros .br and .in. I have therefore
691 removed all such uses from the man pages (some were redundant, some could
692 be replaced by .nf/.fi pairs). The 132html script that I use to generate
693 HTML has been updated to handle .nf/.fi and to complain if it encounters
694 .br or .in.
696 4. Updated comments in configure.ac that get placed in config.h.in and also
697 arranged for config.h to be included in the distribution, with the name
698 config.h.generic, for the benefit of those who have to compile without
699 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
701 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
702 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
703 makevp.bat for the current PCRE, using the additional files
704 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
706 6. A Windows user reported a minor discrepancy with test 2, which turned out
707 to be caused by a trailing space on an input line that had got lost in his
708 copy. The trailing space was an accident, so I've just removed it.
710 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
711 that is needed.
713 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
714 as "const" (a) because they are and (b) because it helps the PHP
715 maintainers who have recently made a script to detect big data structures
716 in the php code that should be moved to the .rodata section. I remembered
717 to update Builducptable as well, so it won't revert if ucptable.h is ever
718 re-created.
720 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
721 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
722 order to be able to cut out the UTF-8 tables in the latter when UTF-8
723 support is not required. This saves 1.5-2K of code, which is important in
724 some applications.
726 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
727 so as not to refer to the tables, even though these functions will never be
728 called when UTF-8 support is disabled. Otherwise there are problems with a
729 shared library.
731 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
733 (a) It was defining its arguments as char * instead of void *.
735 (b) It was assuming that all moves were upwards in memory; this was true
736 a long time ago when I wrote it, but is no longer the case.
738 The emulated memove() is provided for those environments that have neither
739 memmove() nor bcopy(). I didn't think anyone used it these days, but that
740 is clearly not the case, as these two bugs were recently reported.
742 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
743 and Detrail to create the HTML documentation, the .txt form of the man
744 pages, and it removes trailing spaces from listed files. It also creates
745 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
746 case, it wraps all the #defines with #ifndefs. This script should be run
747 before "make dist".
749 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
750 with Unicode property support.
752 (a) For a maximizing quantifier, if the two different cases of the
753 character were of different lengths in their UTF-8 codings (there are
754 some cases like this - I found 11), and the matching function had to
755 back up over a mixture of the two cases, it incorrectly assumed they
756 were both the same length.
758 (b) When PCRE was configured to use the heap rather than the stack for
759 recursion during matching, it was not correctly preserving the data for
760 the other case of a UTF-8 character when checking ahead for a match
761 while processing a minimizing repeat. If the check also involved
762 matching a wide character, but failed, corruption could cause an
763 erroneous result when trying to check for a repeat of the original
764 character.
766 13. Some tidying changes to the testing mechanism:
768 (a) The RunTest script now detects the internal link size and whether there
769 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
770 values substituted by "configure". (The RunGrepTest script already did
771 this for UTF-8.) The configure.ac script no longer substitutes the
772 relevant variables.
774 (b) The debugging options /B and /D in pcretest show the compiled bytecode
775 with length and offset values. This means that the output is different
776 for different internal link sizes. Test 2 is skipped for link sizes
777 other than 2 because of this, bypassing the problem. Unfortunately,
778 there was also a test in test 3 (the locale tests) that used /B and
779 failed for link sizes other than 2. Rather than cut the whole test out,
780 I have added a new /Z option to pcretest that replaces the length and
781 offset values with spaces. This is now used to make test 3 independent
782 of link size. (Test 2 will be tidied up later.)
784 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
785 segmentation fault instead of returning the appropriate error message.
787 15. In multiline mode when the newline sequence was set to "any", the pattern
788 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
789 This doesn't seem right; it now treats the CRLF combination as the line
790 ending, and so does not match in that case. It's only a pattern such as ^$
791 that would hit this one: something like ^ABC$ would have failed after \r
792 and then tried again after \r\n.
794 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
795 in an attempt to make files that differ only in their line terminators
796 compare equal. This works on Linux.
798 17. Under certain error circumstances pcregrep might try to free random memory
799 as it exited. This is now fixed, thanks to valgrind.
801 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
802 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
803 was because its rules for how to advance for /g after matching an empty
804 string at the end of a line did not allow for this case. They now check for
805 it specially.
807 20. pcretest is supposed to handle patterns and data of any length, by
808 extending its buffers when necessary. It was getting this wrong when the
809 buffer for a data line had to be extended.
811 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
812 CRLF as a newline sequence.
814 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
815 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
816 I have nevertheless tidied it up.
818 23. Added some casts to kill warnings from HP-UX ia64 compiler.
820 24. Added a man page for pcre-config.
823 Version 7.0 19-Dec-06
824 ---------------------
826 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
827 moving to gcc 4.1.1.
829 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
830 sys/time.h, which is documented as needed for this function. It doesn't
831 seem to matter on Linux, but it showed up on some releases of OS X.
833 3. It seems that there are systems where bytes whose values are greater than
834 127 match isprint() in the "C" locale. The "C" locale should be the
835 default when a C program starts up. In most systems, only ASCII printing
836 characters match isprint(). This difference caused the output from pcretest
837 to vary, making some of the tests fail. I have changed pcretest so that:
839 (a) When it is outputting text in the compiled version of a pattern, bytes
840 other than 32-126 are always shown as hex escapes.
842 (b) When it is outputting text that is a matched part of a subject string,
843 it does the same, unless a different locale has been set for the match
844 (using the /L modifier). In this case, it uses isprint() to decide.
846 4. Fixed a major bug that caused incorrect computation of the amount of memory
847 required for a compiled pattern when options that changed within the
848 pattern affected the logic of the preliminary scan that determines the
849 length. The relevant options are -x, and -i in UTF-8 mode. The result was
850 that the computed length was too small. The symptoms of this bug were
851 either the PCRE error "internal error: code overflow" from pcre_compile(),
852 or a glibc crash with a message such as "pcretest: free(): invalid next
853 size (fast)". Examples of patterns that provoked this bug (shown in
854 pcretest format) are:
856 /(?-x: )/x
857 /(?x)(?-x: \s*#\s*)/
858 /((?i)[\x{c0}])/8
859 /(?i:[\x{c0}])/8
861 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
862 is now done differently.
864 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
865 wrapper classes; (b) implement a new function in the C++ scanner that is
866 more efficient than the old way of doing things because it avoids levels of
867 recursion in the regex matching; (c) add a paragraph to the documentation
868 for the FullMatch() function.
870 6. The escape sequence \n was being treated as whatever was defined as
871 "newline". Not only was this contrary to the documentation, which states
872 that \n is character 10 (hex 0A), but it also went horribly wrong when
873 "newline" was defined as CRLF. This has been fixed.
875 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
876 was being set to -1 for the "end of line" case (supposedly a value that no
877 character can have). Though this value is never used (the check for end of
878 line is "zero bytes in current character"), it caused compiler complaints.
879 I've changed it to 0xffffffff.
881 8. In pcre_version.c, the version string was being built by a sequence of
882 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
883 string (as it is for production releases) called a macro with an empty
884 argument. The C standard says the result of this is undefined. The gcc
885 compiler treats it as an empty string (which was what was wanted) but it is
886 reported that Visual C gives an error. The source has been hacked around to
887 avoid this problem.
889 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
890 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
891 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
892 of them did).
894 10. Originally, pcretest opened its input and output without "b"; then I was
895 told that "b" was needed in some environments, so it was added for release
896 5.0 to both the input and output. (It makes no difference on Unix-like
897 systems.) Later I was told that it is wrong for the input on Windows. I've
898 now abstracted the modes into two macros, to make it easier to fiddle with
899 them, and removed "b" from the input mode under Windows.
901 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
903 12. Added -help and --help to pcretest as an official way of being reminded
904 of the options.
906 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
907 and pcrecpp.cc because they annoy compilers at high warning levels.
909 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
911 15. Fixed an occurrence of == in configure.ac that should have been = (shell
912 scripts are not C programs :-) and which was not noticed because it works
913 on Linux.
915 16. pcretest is supposed to handle any length of pattern and data line (as one
916 line or as a continued sequence of lines) by extending its input buffer if
917 necessary. This feature was broken for very long pattern lines, leading to
918 a string of junk being passed to pcre_compile() if the pattern was longer
919 than about 50K.
921 17. I have done a major re-factoring of the way pcre_compile() computes the
922 amount of memory needed for a compiled pattern. Previously, there was code
923 that made a preliminary scan of the pattern in order to do this. That was
924 OK when PCRE was new, but as the facilities have expanded, it has become
925 harder and harder to keep it in step with the real compile phase, and there
926 have been a number of bugs (see for example, 4 above). I have now found a
927 cunning way of running the real compile function in a "fake" mode that
928 enables it to compute how much memory it would need, while actually only
929 ever using a few hundred bytes of working memory and without too many
930 tests of the mode. This should make future maintenance and development
931 easier. A side effect of this work is that the limit of 200 on the nesting
932 depth of parentheses has been removed (though this was never a serious
933 limitation, I suspect). However, there is a downside: pcre_compile() now
934 runs more slowly than before (30% or more, depending on the pattern). I
935 hope this isn't a big issue. There is no effect on runtime performance.
937 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
938 newline (only possible for the last line of a file) and it was a
939 pattern that set a locale (followed by /Lsomething), pcretest crashed.
941 19. Added additional timing features to pcretest. (1) The -tm option now times
942 matching only, not compiling. (2) Both -t and -tm can be followed, as a
943 separate command line item, by a number that specifies the number of
944 repeats to use when timing. The default is 50000; this gives better
945 precision, but takes uncomfortably long for very large patterns.
947 20. Extended pcre_study() to be more clever in cases where a branch of a
948 subpattern has no definite first character. For example, (a*|b*)[cd] would
949 previously give no result from pcre_study(). Now it recognizes that the
950 first character must be a, b, c, or d.
952 21. There was an incorrect error "recursive call could loop indefinitely" if
953 a subpattern (or the entire pattern) that was being tested for matching an
954 empty string contained only one non-empty item after a nested subpattern.
955 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
956 incorrectly, because the \d was being skipped in the check.
958 22. The pcretest program now has a new pattern option /B and a command line
959 option -b, which is equivalent to adding /B to every pattern. This causes
960 it to show the compiled bytecode, without the additional information that
961 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
962 is the same as /B/I).
964 23. A new optimization is now able automatically to treat some sequences such
965 as a*b as a*+b. More specifically, if something simple (such as a character
966 or a simple class like \d) has an unlimited quantifier, and is followed by
967 something that cannot possibly match the quantified thing, the quantifier
968 is automatically "possessified".
970 24. A recursive reference to a subpattern whose number was greater than 39
971 went wrong under certain circumstances in UTF-8 mode. This bug could also
972 have affected the operation of pcre_study().
974 25. Realized that a little bit of performance could be had by replacing
975 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
977 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
979 27. Possessive quantifiers such as a++ were previously implemented by turning
980 them into atomic groups such as ($>a+). Now they have their own opcodes,
981 which improves performance. This includes the automatically created ones
982 from 23 above.
984 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
985 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
986 the first matched character to be a colon. This applied both to named and
987 numbered groups.
989 29. The ucpinternal.h header file was missing its idempotency #ifdef.
991 30. I was sent a "project" file called libpcre.a.dev which I understand makes
992 building PCRE on Windows easier, so I have included it in the distribution.
994 31. There is now a check in pcretest against a ridiculously large number being
995 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
996 loop, the loop is abandoned.
998 32. Forward references to subpatterns in conditions such as (?(2)...) where
999 subpattern 2 is defined later cause pcre_compile() to search forwards in
1000 the pattern for the relevant set of parentheses. This search went wrong
1001 when there were unescaped parentheses in a character class, parentheses
1002 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1004 33. "Subroutine" calls and backreferences were previously restricted to
1005 referencing subpatterns earlier in the regex. This restriction has now
1006 been removed.
1008 34. Added a number of extra features that are going to be in Perl 5.10. On the
1009 whole, these are just syntactic alternatives for features that PCRE had
1010 previously implemented using the Python syntax or my own invention. The
1011 other formats are all retained for compatibility.
1013 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1014 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1015 also .NET compatible.
1017 (b) A recursion or subroutine call to a named group can now be defined as
1018 (?&name) as well as (?P>name).
1020 (c) A backreference to a named group can now be defined as \k<name> or
1021 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1022 5.10, are also .NET compatible.
1024 (d) A conditional reference to a named group can now use the syntax
1025 (?(<name>) or (?('name') as well as (?(name).
1027 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1028 groups (named and numbered) that are never evaluated inline, but can be
1029 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1030 is always false. There may be only one alternative in such a group.
1032 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1033 as the simple (?(R). The condition is true only if the most recent
1034 recursion is that of the given number or name. It does not search out
1035 through the entire recursion stack.
1037 (g) The escape \gN or \g{N} has been added, where N is a positive or
1038 negative number, specifying an absolute or relative reference.
1040 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1041 some "unreachable code" warnings.
1043 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1044 things, this adds five new scripts.
1046 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1047 There were also incompatibilities regarding the handling of \Q..\E inside
1048 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1049 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1051 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1052 matches an empty string, and forcibly breaks the loop. There were bugs in
1053 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1054 against aaaa the result was just "a" rather than "aaaa", for example. Two
1055 separate and independent bugs (that affected different cases) have been
1056 fixed.
1058 39. Refactored the code to abolish the use of different opcodes for small
1059 capturing bracket numbers. This is a tidy that I avoided doing when I
1060 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1061 The new approach is not only tidier, it makes it possible to reduce the
1062 memory needed to fix the previous bug (38).
1064 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1065 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1066 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1067 mode.
1069 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1070 report.
1072 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1073 copy construction and assignment in the C++ wrapper.
1075 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1076 couple of bugs that could have given wrong results in the "--newline=crlf"
1077 case.
1079 44. Added a number of casts and did some reorganization of signed/unsigned int
1080 variables following suggestions from Dair Grant. Also renamed the variable
1081 "this" as "item" because it is a C++ keyword.
1083 45. Arranged for dftables to add
1085 #include "pcre_internal.h"
1087 to pcre_chartables.c because without it, gcc 4.x may remove the array
1088 definition from the final binary if PCRE is built into a static library and
1089 dead code stripping is activated.
1091 46. For an unanchored pattern, if a match attempt fails at the start of a
1092 newline sequence, and the newline setting is CRLF or ANY, and the next two
1093 characters are CRLF, advance by two characters instead of one.
1096 Version 6.7 04-Jul-06
1097 ---------------------
1099 1. In order to handle tests when input lines are enormously long, pcretest has
1100 been re-factored so that it automatically extends its buffers when
1101 necessary. The code is crude, but this _is_ just a test program. The
1102 default size has been increased from 32K to 50K.
1104 2. The code in pcre_study() was using the value of the re argument before
1105 testing it for NULL. (Of course, in any sensible call of the function, it
1106 won't be NULL.)
1108 3. The memmove() emulation function in pcre_internal.h, which is used on
1109 systems that lack both memmove() and bcopy() - that is, hardly ever -
1110 was missing a "static" storage class specifier.
1112 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1113 containing an extended class (one that cannot be represented by a bitmap
1114 because it contains high-valued characters or Unicode property items, e.g.
1115 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1116 pattern, but PCRE should not loop if you do not (it no longer does).
1117 [Detail: two cases were found: (a) a repeated subpattern containing an
1118 extended class; (b) a recursive reference to a subpattern that followed a
1119 previous extended class. It wasn't skipping over the extended class
1120 correctly when UTF-8 mode was not set.]
1122 5. A negated single-character class was not being recognized as fixed-length
1123 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1124 compile error "lookbehind assertion is not fixed length".
1126 6. The RunPerlTest auxiliary script was showing an unexpected difference
1127 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1128 write a Perl script that can interpret lines of an input file either as
1129 byte characters or as UTF-8, which is what "perltest" was being required to
1130 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1131 can't do is switch easily at run time between having the "use utf8;" pragma
1132 or not. In the end, I fudged it by using the RunPerlTest script to insert
1133 "use utf8;" explicitly for the UTF-8 tests.
1135 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1136 the end of the subject string, contrary to the documentation and to what
1137 Perl does. This was true of both matching functions. Now it matches only at
1138 the start of the subject and immediately after *internal* newlines.
1140 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1141 a pointer to an int instead of a pointer to an unsigned long int. This
1142 caused problems on 64-bit systems.
1144 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1145 instance of the 'standard' template library not being so standard".
1147 10. There was no check on the number of named subpatterns nor the maximum
1148 length of a subpattern name. The product of these values is used to compute
1149 the size of the memory block for a compiled pattern. By supplying a very
1150 long subpattern name and a large number of named subpatterns, the size
1151 computation could be caused to overflow. This is now prevented by limiting
1152 the length of names to 32 characters, and the number of named subpatterns
1153 to 10,000.
1155 11. Subpatterns that are repeated with specific counts have to be replicated in
1156 the compiled pattern. The size of memory for this was computed from the
1157 length of the subpattern and the repeat count. The latter is limited to
1158 65535, but there was no limit on the former, meaning that integer overflow
1159 could in principle occur. The compiled length of a repeated subpattern is
1160 now limited to 30,000 bytes in order to prevent this.
1162 12. Added the optional facility to have named substrings with the same name.
1164 13. Added the ability to use a named substring as a condition, using the
1165 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1166 are numbers (not recommended). Forward references are permitted.
1168 14. Added forward references in named backreferences (if you see what I mean).
1170 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1171 pattern could run off the end of the subject. For example, the pattern
1172 "(?s)(.{1,5})"8 did this with the subject "ab".
1174 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1175 PCRE_CASELESS was set when matching characters that were quantified with ?
1176 or *.
1178 17. A character class other than a single negated character that had a minimum
1179 but no maximum quantifier - for example [ab]{6,} - was not handled
1180 correctly by pce_dfa_exec(). It would match only one character.
1182 18. A valid (though odd) pattern that looked like a POSIX character
1183 class but used an invalid character after [ (for example [[,abc,]]) caused
1184 pcre_compile() to give the error "Failed: internal error: code overflow" or
1185 in some cases to crash with a glibc free() error. This could even happen if
1186 the pattern terminated after [[ but there just happened to be a sequence of
1187 letters, a binary zero, and a closing ] in the memory that followed.
1189 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1190 over the years. Originally (before any Unicode support), just the bottom 8
1191 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1192 output from "man perlunicode" includes this:
1194 The regular expression compiler produces polymorphic opcodes. That
1195 is, the pattern adapts to the data and automatically switches to
1196 the Unicode character scheme when presented with Unicode data--or
1197 instead uses a traditional byte scheme when presented with byte
1198 data.
1200 Sadly, a wide octal escape does not cause a switch, and in a string with
1201 no other multibyte characters, these octal escapes are treated as before.
1202 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1203 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1204 Unicode string.
1206 I have not perpetrated such confusion in PCRE. Up till now, it took just
1207 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1208 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1209 translate to the appropriate multibyte character.
1211 29. Applied some refactoring to reduce the number of warnings from Microsoft
1212 and Borland compilers. This has included removing the fudge introduced
1213 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1214 a warning about an unused variable.
1216 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1217 characters since release 4.0, because Perl (from release 5.004) does not.
1218 [Or at least, is documented not to: some releases seem to be in conflict
1219 with the documentation.] However, when a pattern was studied with
1220 pcre_study() and all its branches started with \s, PCRE still included VT
1221 as a possible starting character. Of course, this did no harm; it just
1222 caused an unnecessary match attempt.
1224 22. Removed a now-redundant internal flag bit that recorded the fact that case
1225 dependency changed within the pattern. This was once needed for "required
1226 byte" processing, but is no longer used. This recovers a now-scarce options
1227 bit. Also moved the least significant internal flag bit to the most-
1228 significant bit of the word, which was not previously used (hangover from
1229 the days when it was an int rather than a uint) to free up another bit for
1230 the future.
1232 23. Added support for CRLF line endings as well as CR and LF. As well as the
1233 default being selectable at build time, it can now be changed at runtime
1234 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1235 specify that it is scanning data with non-default line endings.
1237 24. Changed the definition of CXXLINK to make it agree with the definition of
1238 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1240 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1241 recursions. This makes a big different to stack usage for some patterns.
1243 26. If a subpattern containing a named recursion or subroutine reference such
1244 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1245 the space required for the compiled pattern went wrong and gave too small a
1246 value. Depending on the environment, this could lead to "Failed: internal
1247 error: code overflow at offset 49" or "glibc detected double free or
1248 corruption" errors.
1250 27. Applied patches from Google (a) to support the new newline modes and (b) to
1251 advance over multibyte UTF-8 characters in GlobalReplace.
1253 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1254 difference for some implementation of PCRE in some Windows version.
1256 29. Added some extra testing facilities to pcretest:
1258 \q<number> in a data line sets the "match limit" value
1259 \Q<number> in a data line sets the "match recursion limt" value
1260 -S <number> sets the stack size, where <number> is in megabytes
1262 The -S option isn't available for Windows.
1265 Version 6.6 06-Feb-06
1266 ---------------------
1268 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1269 in pcreposix.h. I have copied the definition from pcre.h.
1271 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1272 because pcre.h is no longer a built file.
1274 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1275 not normally included in the compiled code.
1278 Version 6.5 01-Feb-06
1279 ---------------------
1281 1. When using the partial match feature with pcre_dfa_exec(), it was not
1282 anchoring the second and subsequent partial matches at the new starting
1283 point. This could lead to incorrect results. For example, with the pattern
1284 /1234/, partially matching against "123" and then "a4" gave a match.
1286 2. Changes to pcregrep:
1288 (a) All non-match returns from pcre_exec() were being treated as failures
1289 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1290 error message is output. Some extra information is given for the
1292 probably the only errors that are likely to be caused by users (by
1293 specifying a regex that has nested indefinite repeats, for instance).
1294 If there are more than 20 of these errors, pcregrep is abandoned.
1296 (b) A binary zero was treated as data while matching, but terminated the
1297 output line if it was written out. This has been fixed: binary zeroes
1298 are now no different to any other data bytes.
1300 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1301 used to set a locale for matching. The --locale=xxxx long option has
1302 been added (no short equivalent) to specify a locale explicitly on the
1303 pcregrep command, overriding the environment variables.
1305 (d) When -B was used with -n, some line numbers in the output were one less
1306 than they should have been.
1308 (e) Added the -o (--only-matching) option.
1310 (f) If -A or -C was used with -c (count only), some lines of context were
1311 accidentally printed for the final match.
1313 (g) Added the -H (--with-filename) option.
1315 (h) The combination of options -rh failed to suppress file names for files
1316 that were found from directory arguments.
1318 (i) Added the -D (--devices) and -d (--directories) options.
1320 (j) Added the -F (--fixed-strings) option.
1322 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1324 (l) Added the --colo(u)r option.
1326 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1327 is not present by default.
1329 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1330 items such as (?R) or (?1), when the recursion could match a number of
1331 alternatives. If it matched one of the alternatives, but subsequently,
1332 outside the recursion, there was a failure, the code tried to back up into
1333 the recursion. However, because of the way PCRE is implemented, this is not
1334 possible, and the result was an incorrect result from the match.
1336 In order to prevent this happening, the specification of recursion has
1337 been changed so that all such subpatterns are automatically treated as
1338 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1340 4. I had overlooked the fact that, in some locales, there are characters for
1341 which isalpha() is true but neither isupper() nor islower() are true. In
1342 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1343 and ordfeminine) are like this. This affected the treatment of \w and \W
1344 when they appeared in character classes, but not when they appeared outside
1345 a character class. The bit map for "word" characters is now created
1346 separately from the results of isalnum() instead of just taking it from the
1347 upper, lower, and digit maps. (Plus the underscore character, of course.)
1349 5. The above bug also affected the handling of POSIX character classes such as
1350 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1351 permanent tables. Instead, the bit maps for such a class were previously
1352 created as the appropriate unions of the upper, lower, and digit bitmaps.
1353 Now they are created by subtraction from the [[:word:]] class, which has
1354 its own bitmap.
1356 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1357 It is created by subtracting the vertical space characters (\x09, \x0a,
1358 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1359 subtraction was done in the overall bitmap for a character class, meaning
1360 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1361 be recognized. This bug has been fixed.
1363 7. Patches from the folks at Google:
1365 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1366 real life, but is still worth protecting against".
1368 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1369 regular expressions".
1371 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1372 have it.
1374 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1375 "configure" and the latter not, in order to fix a problem somebody had
1376 with compiling the Arg class on HP-UX.
1378 (e) Improve the error-handling of the C++ wrapper a little bit.
1380 (f) New tests for checking recursion limiting.
1382 8. The pcre_memmove() function, which is used only if the environment does not
1383 have a standard memmove() function (and is therefore rarely compiled),
1384 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1385 returning a result (though PCRE never actually uses the result).
1387 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1388 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1389 returned instead of calling malloc() with an overflowing number that would
1390 most likely cause subsequent chaos.
1392 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1394 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1395 with this option is matched, the nmatch and pmatch options of regexec() are
1396 ignored.
1398 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1399 provided in case anyone wants to the the POSIX interface with UTF-8
1400 strings.
1402 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1403 C++ linking (needed for some HP-UX environments).
1405 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1406 (unused parameter) and in the pcre_printint() function (omitted "default"
1407 switch label when the default is to do nothing).
1409 15. Added some code to make it possible, when PCRE is compiled as a C++
1410 library, to replace subject pointers for pcre_exec() with a smart pointer
1411 class, thus making it possible to process discontinuous strings.
1413 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1414 much the same function. They were added by different people who were trying
1415 to make PCRE easy to compile on non-Unix systems. It has been suggested
1416 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1417 for compiling on Windows systems. I have therefore replaced it with
1418 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1419 defaults to "extern" for C or "extern C" for C++, which works fine on
1420 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1421 SCOPE with something explicit in config.h. In addition:
1423 (a) pcreposix.h still had just "extern" instead of either of these macros;
1424 I have replaced it with PCRE_DATA_SCOPE.
1426 (b) Functions such as _pcre_xclass(), which are internal to the library,
1427 but external in the C sense, all had PCRE_EXPORT in their definitions.
1428 This is apparently wrong for the Windows case, so I have removed it.
1429 (It makes no difference on Unix-like systems.)
1431 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1432 of recursive calls to match(). This is different to MATCH_LIMIT because
1433 that limits the total number of calls to match(), not all of which increase
1434 the depth of recursion. Limiting the recursion depth limits the amount of
1435 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1436 when PCRE is compiled, and changed at run time. A patch from Google adds
1437 this functionality to the C++ interface.
1439 18. Changes to the handling of Unicode character properties:
1441 (a) Updated the table to Unicode 4.1.0.
1443 (b) Recognize characters that are not in the table as "Cn" (undefined).
1445 (c) I revised the way the table is implemented to a much improved format
1446 which includes recognition of ranges. It now supports the ranges that
1447 are defined in UnicodeData.txt, and it also amalgamates other
1448 characters into ranges. This has reduced the number of entries in the
1449 table from around 16,000 to around 3,000, thus reducing its size
1450 considerably. I realized I did not need to use a tree structure after
1451 all - a binary chop search is just as efficient. Having reduced the
1452 number of entries, I extended their size from 6 bytes to 8 bytes to
1453 allow for more data.
1455 (d) Added support for Unicode script names via properties such as \p{Han}.
1457 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1458 matching that character.
1460 20. When matching a repeated Unicode property with a minimum greater than zero,
1461 (for example \pL{2,}), PCRE could look past the end of the subject if it
1462 reached it while seeking the minimum number of characters. This could
1463 happen only if some of the characters were more than one byte long, because
1464 there is a check for at least the minimum number of bytes.
1466 21. Refactored the implementation of \p and \P so as to be more general, to
1467 allow for more different types of property in future. This has changed the
1468 compiled form incompatibly. Anybody with saved compiled patterns that use
1469 \p or \P will have to recompile them.
1471 22. Added "Any" and "L&" to the supported property types.
1473 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1474 but give a compile time error if the value is greater than 0xff.
1476 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1477 accidentally not being installed or uninstalled.
1479 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1480 made were to insert the current release number. This seemed silly, because
1481 it made things harder for people building PCRE on systems that don't run
1482 "configure". I have turned pcre.h into a distributed file, no longer built
1483 by "configure", with the version identification directly included. There is
1484 no longer a pcre.h.in file.
1486 However, this change necessitated a change to the pcre-config script as
1487 well. It is built from pcre-config.in, and one of the substitutions was the
1488 release number. I have updated configure.ac so that ./configure now finds
1489 the release number by grepping pcre.h.
1491 26. Added the ability to run the tests under valgrind.
1494 Version 6.4 05-Sep-05
1495 ---------------------
1497 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1498 "--" to be printed when multiple files were scanned, even when none of the
1499 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1500 consider it to be a bug, and have restored the previous behaviour.
1502 2. A couple of code tidies to get rid of compiler warnings.
1504 3. The pcretest program used to cheat by referring to symbols in the library
1505 whose names begin with _pcre_. These are internal symbols that are not
1506 really supposed to be visible externally, and in some environments it is
1507 possible to suppress them. The cheating is now confined to including
1508 certain files from the library's source, which is a bit cleaner.
1510 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1511 file's purpose clearer.
1513 5. Reorganized pcre_ucp_findchar().
1516 Version 6.3 15-Aug-05
1517 ---------------------
1519 1. The file libpcre.pc.in did not have general read permission in the tarball.
1521 2. There were some problems when building without C++ support:
1523 (a) If C++ support was not built, "make install" and "make test" still
1524 tried to test it.
1526 (b) There were problems when the value of CXX was explicitly set. Some
1527 changes have been made to try to fix these, and ...
1529 (c) --disable-cpp can now be used to explicitly disable C++ support.
1531 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1532 backslash in a target when C++ was disabled. This confuses some
1533 versions of "make", apparently. Using an intermediate variable solves
1534 this. (Same for CPP_LOBJ.)
1536 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1537 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1538 necessary on certain architectures.
1540 4. Added a setting of -export-symbols-regex to the link command to remove
1541 those symbols that are exported in the C sense, but actually are local
1542 within the library, and not documented. Their names all begin with
1543 "_pcre_". This is not a perfect job, because (a) we have to except some
1544 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1545 available (and never for static libraries). I have made a note to try to
1546 find a way round (a) in the future.
1549 Version 6.2 01-Aug-05
1550 ---------------------
1552 1. There was no test for integer overflow of quantifier values. A construction
1553 such as {1111111111111111} would give undefined results. What is worse, if
1554 a minimum quantifier for a parenthesized subpattern overflowed and became
1555 negative, the calculation of the memory size went wrong. This could have
1556 led to memory overwriting.
1558 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1560 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1561 operating environments where this matters.
1563 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1564 PCRE options from within the C++ wrapper.
1566 5. Named capturing subpatterns were not being correctly counted when a pattern
1567 was compiled. This caused two problems: (a) If there were more than 100
1568 such subpatterns, the calculation of the memory needed for the whole
1569 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1570 back references of the form \12, where the number was greater than 9, were
1571 not recognized as back references, even though there were sufficient
1572 previous subpatterns.
1574 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1575 versions of gcc, e.g. 2.95.4.
1578 Version 6.1 21-Jun-05
1579 ---------------------
1581 1. There was one reference to the variable "posix" in pcretest.c that was not
1582 surrounded by "#if !defined NOPOSIX".
1584 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1585 the cross-check on the old pcre_info() function, for the benefit of the
1586 cut-down version of PCRE that is currently imported into Exim.
1588 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1589 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1590 patterns that start (?i) but I don't think that matters. The use of (?i) is
1591 just an example; this all applies to the other options as well.
1593 4. Since libtool seems to echo the compile commands it is issuing, the output
1594 from "make" can be reduced a bit by putting "@" in front of each libtool
1595 compile command.
1597 5. Patch from the folks at Google for configure.in to be a bit more thorough
1598 in checking for a suitable C++ installation before trying to compile the
1599 C++ stuff. This should fix a reported problem when a compiler was present,
1600 but no suitable headers.
1602 6. The man pages all had just "PCRE" as their title. I have changed them to
1603 be the relevant file name. I have also arranged that these names are
1604 retained in the file doc/pcre.txt, which is a concatenation in text format
1605 of all the man pages except the little individual ones for each function.
1607 7. The NON-UNIX-USE file had not been updated for the different set of source
1608 files that come with release 6. I also added a few comments about the C++
1609 wrapper.
1612 Version 6.0 07-Jun-05
1613 ---------------------
1615 1. Some minor internal re-organization to help with my DFA experiments.
1617 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1618 didn't matter for the library itself when fully configured, but did matter
1619 when compiling without UCP support, or within Exim, where the ucp files are
1620 not imported.
1622 3. Refactoring of the library code to split up the various functions into
1623 different source modules. The addition of the new DFA matching code (see
1624 below) to a single monolithic source would have made it really too
1625 unwieldy, quite apart from causing all the code to be include in a
1626 statically linked application, when only some functions are used. This is
1627 relevant even without the DFA addition now that patterns can be compiled in
1628 one application and matched in another.
1630 The downside of splitting up is that there have to be some external
1631 functions and data tables that are used internally in different modules of
1632 the library but which are not part of the API. These have all had their
1633 names changed to start with "_pcre_" so that they are unlikely to clash
1634 with other external names.
1636 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1637 a different (DFA) algorithm. Although it is slower than the original
1638 function, it does have some advantages for certain types of matching
1639 problem.
1641 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1642 including restarting after a partial match.
1644 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1645 defined when compiling for Windows was sent to me. I have put it into the
1646 code, though I have no means of testing or verifying it.
1648 7. Added the pcre_refcount() auxiliary function.
1650 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1651 match before or at the first newline in the subject string. In pcretest,
1652 the /f option on a pattern can be used to set this.
1654 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1655 would behave wrongly. This has been present in PCRE since release 4.0.
1657 10. A number of changes to the pcregrep command:
1659 (a) Refactored how -x works; insert ^(...)$ instead of setting
1660 PCRE_ANCHORED and checking the length, in preparation for adding
1661 something similar for -w.
1663 (b) Added the -w (match as a word) option.
1665 (c) Refactored the way lines are read and buffered so as to have more
1666 than one at a time available.
1668 (d) Implemented a pcregrep test script.
1670 (e) Added the -M (multiline match) option. This allows patterns to match
1671 over several lines of the subject. The buffering ensures that at least
1672 8K, or the rest of the document (whichever is the shorter) is available
1673 for matching (and similarly the previous 8K for lookbehind assertions).
1675 (f) Changed the --help output so that it now says
1677 -w, --word-regex(p)
1679 instead of two lines, one with "regex" and the other with "regexp"
1680 because that confused at least one person since the short forms are the
1681 same. (This required a bit of code, as the output is generated
1682 automatically from a table. It wasn't just a text change.)
1684 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1685 option but starts with a hyphen. Could be a pattern or a path name
1686 starting with a hyphen, for instance.
1688 (h) "-" can be given as a file name to represent stdin.
1690 (i) When file names are being printed, "(standard input)" is used for
1691 the standard input, for compatibility with GNU grep. Previously
1692 "<stdin>" was used.
1694 (j) The option --label=xxx can be used to supply a name to be used for
1695 stdin when file names are being printed. There is no short form.
1697 (k) Re-factored the options decoding logic because we are going to add
1698 two more options that take data. Such options can now be given in four
1699 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1701 (l) Added the -A, -B, and -C options for requesting that lines of context
1702 around matches be printed.
1704 (m) Added the -L option to print the names of files that do not contain
1705 any matching lines, that is, the complement of -l.
1707 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1708 continue to scan other files.
1710 (o) The -s option was incorrectly implemented. For compatibility with other
1711 greps, it now suppresses the error message for a non-existent or non-
1712 accessible file (but not the return code). There is a new option called
1713 -q that suppresses the output of matching lines, which was what -s was
1714 previously doing.
1716 (p) Added --include and --exclude options to specify files for inclusion
1717 and exclusion when recursing.
1719 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1720 Hopefully, it now does.
1722 12. Missing cast in pcre_study().
1724 13. Added an "uninstall" target to the makefile.
1726 14. Replaced "extern" in the function prototypes in Makefile.in with
1727 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1728 world, but is set differently for Windows.
1730 15. Added a second compiling function called pcre_compile2(). The only
1731 difference is that it has an extra argument, which is a pointer to an
1732 integer error code. When there is a compile-time failure, this is set
1733 non-zero, in addition to the error test pointer being set to point to an
1734 error message. The new argument may be NULL if no error number is required
1735 (but then you may as well call pcre_compile(), which is now just a
1736 wrapper). This facility is provided because some applications need a
1737 numeric error indication, but it has also enabled me to tidy up the way
1738 compile-time errors are handled in the POSIX wrapper.
1740 16. Added VPATH=.libs to the makefile; this should help when building with one
1741 prefix path and installing with another. (Or so I'm told by someone who
1742 knows more about this stuff than I do.)
1744 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1745 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1746 match everything, including newlines. This is not POSIX-compatible, but
1747 somebody wanted the feature. From pcretest it can be activated by using
1748 both the P and the s flags.
1750 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1752 19. libpcre.pc was being incorrectly installed as executable.
1754 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1755 it now also looks for '\r' so that it will work unmodified on Windows.
1757 21. Added Google's contributed C++ wrapper to the distribution.
1759 22. Added some untidy missing memory free() calls in pcretest, to keep
1760 Electric Fence happy when testing.
1764 Version 5.0 13-Sep-04
1765 ---------------------
1767 1. Internal change: literal characters are no longer packed up into items
1768 containing multiple characters in a single byte-string. Each character
1769 is now matched using a separate opcode. However, there may be more than one
1770 byte in the character in UTF-8 mode.
1772 2. The pcre_callout_block structure has two new fields: pattern_position and
1773 next_item_length. These contain the offset in the pattern to the next match
1774 item, and its length, respectively.
1776 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1777 insertion of callouts before each pattern item. Added the /C option to
1778 pcretest to make use of this.
1780 4. On the advice of a Windows user, the lines
1782 #if defined(_WIN32) || defined(WIN32)
1783 _setmode( _fileno( stdout ), 0x8000 );
1784 #endif /* defined(_WIN32) || defined(WIN32) */
1786 have been added to the source of pcretest. This apparently does useful
1787 magic in relation to line terminators.
1789 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1790 for the benefit of those environments where the "b" makes a difference.
1792 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1793 to know about it. I have put a hack into configure.in that adds in code
1794 to set GCC=yes if CC=icc. This seems to end up at a point in the
1795 generated configure script that is early enough to affect the setting of
1796 compiler options, which is what is needed, but I have no means of testing
1797 whether it really works. (The user who reported this had patched the
1798 generated configure script, which of course I cannot do.)
1800 LATER: After change 22 below (new libtool files), the configure script
1801 seems to know about icc (and also ecc). Therefore, I have commented out
1802 this hack in configure.in.
1804 7. Added support for pkg-config (2 patches were sent in).
1806 8. Negated POSIX character classes that used a combination of internal tables
1807 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1808 [[:^ascii]]. Typically, they would match almost any characters. The other
1809 POSIX classes were not broken in this way.
1811 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1812 to find the match, as PCRE was deluded into thinking that the match had to
1813 start at the start point or following a newline. The same bug applied to
1814 patterns with negative forward assertions or any backward assertions
1815 preceding ".*" at the start, unless the pattern required a fixed first
1816 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1818 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1819 starting at the last subject character, bytes beyond the end of the subject
1820 string were read.
1822 11. Renamed the variable "class" as "classbits" to make life easier for C++
1823 users. (Previously there was a macro definition, but it apparently wasn't
1824 enough.)
1826 12. Added the new field "tables" to the extra data so that tables can be passed
1827 in at exec time, or the internal tables can be re-selected. This allows
1828 a compiled regex to be saved and re-used at a later time by a different
1829 program that might have everything at different addresses.
1831 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1832 -R library as well as a -L library.
1834 14. The debugging options of pcretest (-d on the command line or D on a
1835 pattern) showed incorrect output for anything following an extended class
1836 that contained multibyte characters and which was followed by a quantifier.
1838 15. Added optional support for general category Unicode character properties
1839 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1840 support. It adds about 90K to the size of the library. The meanings of the
1841 inbuilt class escapes such as \d and \s have NOT been changed.
1843 16. Updated pcredemo.c to include calls to free() to release the memory for the
1844 compiled pattern.
1846 17. The generated file chartables.c was being created in the source directory
1847 instead of in the building directory. This caused the build to fail if the
1848 source directory was different from the building directory, and was
1849 read-only.
1851 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1852 file. No doubt somebody will tell me if they don't make sense... Also added
1853 Dan Mooney's comments about building on OpenVMS.
1855 19. Added support for partial matching via the PCRE_PARTIAL option for
1856 pcre_exec() and the \P data escape in pcretest.
1858 20. Extended pcretest with 3 new pattern features:
1860 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1861 write the compiled pattern to the file whose name is "rest-of-line".
1862 This is a straight binary dump of the data, with the saved pointer to
1863 the character tables forced to be NULL. The study data, if any, is
1864 written too. After writing, pcretest reads a new pattern.
1866 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1867 compiled pattern from the given file. There must not be any
1868 occurrences of "<" in the file name (pretty unlikely); if there are,
1869 pcretest will instead treat the initial "<" as a pattern delimiter.
1870 After reading in the pattern, pcretest goes on to read data lines as
1871 usual.
1873 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1874 and 16-bit fields in a compiled pattern, to simulate a pattern that
1875 was compiled on a host of opposite endianness.
1877 21. The pcre-exec() function can now cope with patterns that were compiled on
1878 hosts of opposite endianness, with this restriction:
1880 As for any compiled expression that is saved and used later, the tables
1881 pointer field cannot be preserved; the extra_data field in the arguments
1882 to pcre_exec() should be used to pass in a tables address if a value
1883 other than the default internal tables were used at compile time.
1885 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1886 now diagnosed as an error. Previously, most of the time, a negative number
1887 would have been treated as zero, but if in addition "ovector" was passed as
1888 NULL, a crash could occur.
1890 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1891 new versions from the libtool 1.5 distribution (the last one is a copy of
1892 a file called libtool.m4). This seems to have fixed the need to patch
1893 "configure" to support Darwin 1.3 (which I used to do). However, I still
1894 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1895 workstation).
1897 24. Changed the PCRE licence to be the more standard "BSD" licence.
1900 Version 4.5 01-Dec-03
1901 ---------------------
1903 1. There has been some re-arrangement of the code for the match() function so
1904 that it can be compiled in a version that does not call itself recursively.
1905 Instead, it keeps those local variables that need separate instances for
1906 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1907 needs to "recurse". Keeping track of where control must go is done by means
1908 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1909 hide most of the details from the main code, and operates only if
1910 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1911 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1912 operating.
1914 To make it easier for callers to provide specially tailored get/free
1915 functions for this usage, two new functions, pcre_stack_malloc, and
1916 pcre_stack_free, are used. They are always called in strict stacking order,
1917 and the size of block requested is always the same.
1919 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1920 PCRE has been compiled to use the stack or the heap for recursion. The
1921 -C option of pcretest uses this to show which version is compiled.
1923 A new data escape \S, is added to pcretest; it causes the amounts of store
1924 obtained and freed by both kinds of malloc/free at match time to be added
1925 to the output.
1927 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1928 what's available on my current Linux desktop machine.
1930 3. When matching a UTF-8 string, the test for a valid string at the start has
1931 been extended. If start_offset is not zero, PCRE now checks that it points
1932 to a byte that is the start of a UTF-8 character. If not, it returns
1933 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1934 this is necessary because there may be backward assertions in the pattern.
1935 When matching the same subject several times, it may save resources to use
1936 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1938 4. The code for checking the validity of UTF-8 strings has been tightened so
1939 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1940 containing "overlong sequences".
1942 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1943 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1944 should have been "&", but it just so happened that all the cases this let
1945 through by mistake were picked up later in the function.
1947 6. I had used a variable called "isblank" - this is a C99 function, causing
1948 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1950 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1951 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1952 so the version is shown; (c) stop "make test" echoing "./RunTest".
1954 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1956 9. The prototype for memmove() for systems that don't have it was using
1957 size_t, but the inclusion of the header that defines size_t was later. I've
1958 moved the #includes for the C headers earlier to avoid this.
1960 10. Added some adjustments to the code to make it easier to compiler on certain
1961 special systems:
1963 (a) Some "const" qualifiers were missing.
1964 (b) Added the macro EXPORT before all exported functions; by default this
1965 is defined to be empty.
1966 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1967 that it reads its output file name as an argument instead of writing
1968 to the standard output and assuming this can be redirected.
1970 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1971 class containing characters with values greater than 255, PCRE compilation
1972 went into a loop.
1974 12. A recursive reference to a subpattern that was within another subpattern
1975 that had a minimum quantifier of zero caused PCRE to crash. For example,
1976 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1977 recursion. If the recursively-called subpattern itself had a zero repeat,
1978 that was OK.
1980 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1981 buffer into which it was copied (for escape processing) was still set at
1982 1024, so long lines caused crashes.
1984 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1985 "internal error: code overflow...". This applied to any character class
1986 that was followed by a possessive quantifier.
1988 15. Modified the Makefile to add libpcre.la as a prerequisite for
1989 libpcreposix.la because I was told this is needed for a parallel build to
1990 work.
1992 16. If a pattern that contained .* following optional items at the start was
1993 studied, the wrong optimizing data was generated, leading to matching
1994 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1995 matching string must start with a or b or c. The correct conclusion for
1996 this pattern is that a match can start with any character.
1999 Version 4.4 13-Aug-03
2000 ---------------------
2002 1. In UTF-8 mode, a character class containing characters with values between
2003 127 and 255 was not handled correctly if the compiled pattern was studied.
2004 In fixing this, I have also improved the studying algorithm for such
2005 classes (slightly).
2007 2. Three internal functions had redundant arguments passed to them. Removal
2008 might give a very teeny performance improvement.
2010 3. Documentation bug: the value of the capture_top field in a callout is *one
2011 more than* the number of the hightest numbered captured substring.
2013 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2014 in incorrectly linking with a previously installed version. They now link
2015 explicitly with libpcre.la.
2017 5. configure.in no longer needs to recognize Cygwin specially.
2019 6. A problem in pcre.in for Windows platforms is fixed.
2021 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2022 pcretest, it used to include the size of the study block as part of its
2023 output. Unfortunately, the structure contains a field that has a different
2024 size on different hardware architectures. This meant that the tests that
2025 showed this size failed. As the block is currently always of a fixed size,
2026 this information isn't actually particularly useful in pcretest output, so
2027 I have just removed it.
2029 8. Three pre-processor statements accidentally did not start in column 1.
2030 Sadly, there are *still* compilers around that complain, even though
2031 standard C has not required this for well over a decade. Sigh.
2033 9. In pcretest, the code for checking callouts passed small integers in the
2034 callout_data field, which is a void * field. However, some picky compilers
2035 complained about the casts involved for this on 64-bit systems. Now
2036 pcretest passes the address of the small integer instead, which should get
2037 rid of the warnings.
2039 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2040 both compile and run time, and gives an error if an invalid UTF-8 sequence
2041 is found. There is a option for disabling this check in cases where the
2042 string is known to be correct and/or the maximum performance is wanted.
2044 11. In response to a bug report, I changed one line in Makefile.in from
2046 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2047 to
2048 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2050 to look similar to other lines, but I have no way of telling whether this
2051 is the right thing to do, as I do not use Windows. No doubt I'll get told
2052 if it's wrong...
2055 Version 4.3 21-May-03
2056 ---------------------
2058 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2059 Makefile.
2061 2. Some refactoring to improve the quality of the code:
2063 (i) The utf8_table... variables are now declared "const".
2065 (ii) The code for \cx, which used the "case flipping" table to upper case
2066 lower case letters, now just substracts 32. This is ASCII-specific,
2067 but the whole concept of \cx is ASCII-specific, so it seems
2068 reasonable.
2070 (iii) PCRE was using its character types table to recognize decimal and
2071 hexadecimal digits in the pattern. This is silly, because it handles
2072 only 0-9, a-f, and A-F, but the character types table is locale-
2073 specific, which means strange things might happen. A private
2074 table is now used for this - though it costs 256 bytes, a table is
2075 much faster than multiple explicit tests. Of course, the standard
2076 character types table is still used for matching digits in subject
2077 strings against \d.
2079 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2080 ending in _t are). So I've renamed it as ESC_tee.
2082 3. The first argument for regexec() in the POSIX wrapper should have been
2083 defined as "const".
2085 4. Changed pcretest to use malloc() for its buffers so that they can be
2086 Electric Fenced for debugging.
2088 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2089 to read one or more bytes before the start of the subject string. Often this
2090 had no effect on PCRE's behaviour, but in some circumstances it could
2091 provoke a segmentation fault.
2093 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2094 to try to read one or more bytes before the start of the subject string.
2096 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2097 UTF-8 support could misbehave in various ways if the subject string
2098 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2099 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2100 back over UTF-8 characters.)
2103 Version 4.2 14-Apr-03
2104 ---------------------
2106 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2108 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2109 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2110 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2111 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2112 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2114 Note: automatic setting of the BUILD variables is not yet working
2115 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2116 compile-time but not at link-time
2117 [LINK]: use for linking executables only
2118 make different versions for Windows and non-Windows
2119 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2120 libraries
2121 [LINK_FOR_BUILD]: new variable
2122 [OBJEXT]: use throughout
2123 [EXEEXT]: use throughout
2124 <winshared>: new target
2125 <wininstall>: new target
2126 <dftables.o>: use native compiler
2127 <dftables>: use native linker
2128 <install>: handle Windows platform correctly
2129 <clean>: ditto
2130 <check>: ditto
2131 copy DLL to top builddir before testing
2133 As part of these changes, -no-undefined was removed again. This was reported
2134 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2135 in any case.
2137 3. Some tidies to get rid of compiler warnings:
2139 . In the match_data structure, match_limit was an unsigned long int, whereas
2140 match_call_count was an int. I've made them both unsigned long ints.
2142 . In pcretest the fact that a const uschar * doesn't automatically cast to
2143 a void * provoked a warning.
2145 . Turning on some more compiler warnings threw up some "shadow" variables
2146 and a few more missing casts.
2148 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2149 option, a class that contained a single character with a value between 128
2150 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2152 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2153 option, a class that contained several characters, but with at least one
2154 whose value was between 128 and 255 caused PCRE to crash.
2157 Version 4.1 12-Mar-03
2158 ---------------------
2160 1. Compiling with gcc -pedantic found a couple of places where casts were
2161 needed, and a string in dftables.c that was longer than standard compilers are
2162 required to support.
2164 2. Compiling with Sun's compiler found a few more places where the code could
2165 be tidied up in order to avoid warnings.
2167 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2168 first of these names is deprecated in the latest Autoconf in favour of the name
2169 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2170 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2171 analogy I have changed it to CFLAGS_FOR_BUILD.
2173 4. Added -no-undefined to the linking command in the Makefile, because this is
2174 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2175 linking step for the pcreposix library.
2177 5. PCRE was failing to diagnose the case of two named groups with the same
2178 name.
2180 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2181 literal character that is needed in the subject for a match, and scans along to
2182 ensure that it is present before embarking on the full matching process. This
2183 saves time in cases of nested unlimited repeats that are never going to match.
2184 Problem: the scan can take a lot of time if the subject is very long (e.g.
2185 megabytes), thus penalizing straightforward matches. It is now done only if the
2186 amount of subject to be scanned is less than 1000 bytes.
2188 7. A lesser problem with the same optimization is that it was recording the
2189 first character of an anchored pattern as "needed", thus provoking a search
2190 right along the subject, even when the first match of the pattern was going to
2191 fail. The "needed" character is now not set for anchored patterns, unless it
2192 follows something in the pattern that is of non-fixed length. Thus, it still
2193 fulfils its original purpose of finding quick non-matches in cases of nested
2194 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2197 Version 4.0 17-Feb-03
2198 ---------------------
2200 1. If a comment in an extended regex that started immediately after a meta-item
2201 extended to the end of string, PCRE compiled incorrect data. This could lead to
2202 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2204 2. Moved to autoconf 2.53 and libtool 1.4.2.
2206 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2207 the special perltest8 script is no longer needed - all the tests can be run
2208 from a single perltest script.
2210 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2211 by \s. It has now been removed in PCRE. This means it isn't recognized as
2212 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2213 class [:space:] *does* include VT, thereby creating a mess.
2215 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2216 space and tab.
2218 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2219 its new features into the main test script, reducing the number of scripts.
2221 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2222 were backward compatible, and made the (?i) apply to the whole pattern, as if
2223 /i were given. Now it behaves more logically, and applies the option setting
2224 only to what follows. PCRE has been changed to follow suit. However, if it
2225 finds options settings right at the start of the pattern, it extracts them into
2226 the global options, as before. Thus, they show up in the info data.
2228 8. Added support for the \Q...\E escape sequence. Characters in between are
2229 treated as literals. This is slightly different from Perl in that $ and @ are
2230 also handled as literals inside the quotes. In Perl, they will cause variable
2231 interpolation. Note the following examples:
2233 Pattern PCRE matches Perl matches
2235 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2236 \Qabc\$xyz\E abc\$xyz abc\$xyz
2237 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2239 For compatibility with Perl, \Q...\E sequences are recognized inside character
2240 classes as well as outside them.
2242 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2243 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2244 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2245 signed/unsigned warnings.
2247 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2248 option for pcretest, so I've replaced it by a simple function that does just
2249 that job.
2251 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2252 "pcregrep -".
2254 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2255 Java package. This provides some syntactic sugar for simple cases of what my
2256 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2257 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2258 item, you can use this simplified notation. Note that only makes sense with
2259 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2260 greediness, whatever the setting of the PCRE_UNGREEDY option.
2262 13. A change of greediness default within a pattern was not taking effect at
2263 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2264 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2265 was abstracted outside.
2267 14. PCRE now supports the \G assertion. It is true when the current matching
2268 position is at the start point of the match. This differs from \A when the
2269 starting offset is non-zero. Used with the /g option of pcretest (or similar
2270 code), it works in the same way as it does for Perl's /g option. If all
2271 alternatives of a regex begin with \G, the expression is anchored to the start
2272 match position, and the "anchored" flag is set in the compiled expression.
2274 15. Some bugs concerning the handling of certain option changes within patterns
2275 have been fixed. These applied to options other than (?ims). For example,
2276 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2277 been the other way round. Some of this was related to change 7 above.
2279 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2280 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2281 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2282 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2284 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2285 mode. Unlike ".", it always matches newline, whatever the setting of
2286 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2287 assertions. Perl allows it, but it doesn't (in general) work because it can't
2288 calculate the length of the lookbehind. At least, that's the case for Perl
2289 5.8.0 - I've been told they are going to document that it doesn't work in
2290 future.
2292 18. Added an error diagnosis for escapes that PCRE does not support: these are
2293 \L, \l, \N, \P, \p, \U, \u, and \X.
2295 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2296 reading past the end of the pattern in cases such as /[abcd/.
2298 20. PCRE was getting more memory than necessary for patterns with classes that
2299 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2301 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2302 compiling PCRE for use with Virtual Pascal.
2304 22. Small fix to the Makefile to make it work properly if the build is done
2305 outside the source tree.
2307 23. Added a new extension: a condition to go with recursion. If a conditional
2308 subpattern starts with (?(R) the "true" branch is used if recursion has
2309 happened, whereas the "false" branch is used only at the top level.
2311 24. When there was a very long string of literal characters (over 255 bytes
2312 without UTF support, over 250 bytes with UTF support), the computation of how
2313 much memory was required could be incorrect, leading to segfaults or other
2314 strange effects.
2316 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2317 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2318 there was a subsequent back reference to those brackets. This meant that, for
2319 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2320 possible to check for precisely this case. All we can do is abandon the
2321 optimization if .* occurs inside capturing brackets when there are any back
2322 references whatsoever. (See below for a better fix that came later.)
2324 26. The handling of the optimization for finding the first character of a
2325 non-anchored pattern, and for finding a character that is required later in the
2326 match were failing in some cases. This didn't break the matching; it just
2327 failed to optimize when it could. The way this is done has been re-implemented.
2329 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2331 28. Added a new feature that provides some of the functionality that Perl
2332 provides with (?{...}). The facility is termed a "callout". The way it is done
2333 in PCRE is for the caller to provide an optional function, by setting
2334 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2335 global variable. By default it is unset, which disables all calling out. To get
2336 the function called, the regex must include (?C) at appropriate points. This
2337 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2338 This provides a means of identifying different callout points. When PCRE
2339 reaches such a point in the regex, if pcre_callout has been set, the external
2340 function is called. It is provided with data in a structure called
2341 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2342 matching continues; if it returns a non-zero value, the match at the current
2343 point fails. However, backtracking will occur if possible. [This was changed
2344 later and other features added - see item 49 below.]
2346 29. pcretest is upgraded to test the callout functionality. It provides a
2347 callout function that displays information. By default, it shows the start of
2348 the match and the current position in the text. There are some new data escapes
2349 to vary what happens:
2351 \C+ in addition, show current contents of captured substrings
2352 \C- do not supply a callout function
2353 \C!n return 1 when callout number n is reached
2354 \C!n!m return 1 when callout number n is reached for the mth time
2356 30. If pcregrep was called with the -l option and just a single file name, it
2357 output "<stdin>" if a match was found, instead of the file name.
2359 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2360 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2361 pcre_exec(). This saves a malloc/free per call. The default value of
2362 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2363 when configuring.
2365 32. The default maximum size of a compiled pattern is 64K. There have been a
2366 few cases of people hitting this limit. The code now uses macros to handle the
2367 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2368 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2369 configuring. Tests 2 and 5 work only with 2-byte links because they output
2370 debugging information about compiled patterns.
2372 33. Internal code re-arrangements:
2374 (a) Moved the debugging function for printing out a compiled regex into
2375 its own source file (printint.c) and used #include to pull it into
2376 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2377 separate copies.
2379 (b) Defined the list of op-code names for debugging as a macro in
2380 internal.h so that it is next to the definition of the opcodes.
2382 (c) Defined a table of op-code lengths for simpler skipping along compiled
2383 code. This is again a macro in internal.h so that it is next to the
2384 definition of the opcodes.
2386 34. Added support for recursive calls to individual subpatterns, along the
2387 lines of Robin Houston's patch (but implemented somewhat differently).
2389 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2390 allow it to read and process whole directories in Win32. This code was
2391 contributed by Lionel Fourquaux; it has not been tested by me.
2393 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2394 used to name a group. Names consist of alphanumerics and underscores, and must
2395 be unique. Back references use the syntax (?P=name) and recursive calls use
2396 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2397 numbers. The function pcre_fullinfo() can be used after compilation to extract
2398 a name/number map. There are three relevant calls:
2400 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2401 PCRE_INFO_NAMECOUNT yields the number of entries
2402 PCRE_INFO_NAMETABLE yields a pointer to the map.
2404 The map is a vector of fixed-size entries. The size of each entry depends on
2405 the length of the longest name used. The first two bytes of each entry are the
2406 group number, most significant byte first. There follows the corresponding
2407 name, zero terminated. The names are in alphabetical order.
2409 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2410 case instead of 255. Making it the same both with and without UTF-8 support
2411 means that the same test output works with both.
2413 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2414 calling malloc() with a zero argument.
2416 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2417 optimization. I've improved things by keeping a bitmap of backreferences with
2418 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2419 fact referenced, the optimization can be applied. It is unlikely that a
2420 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2421 the match to follow \n) will appear inside brackets with a number greater than
2422 31, but if it does, any back reference > 31 suppresses the optimization.
2424 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2425 of disabling numbered capturing parentheses. Any opening parenthesis that is
2426 not followed by ? behaves as if it were followed by ?: but named parentheses
2427 can still be used for capturing (and they will acquire numbers in the usual
2428 way).
2430 41. Redesigned the return codes from the match() function into yes/no/error so
2431 that errors can be passed back from deep inside the nested calls. A malloc
2432 failure while inside a recursive subpattern call now causes the
2433 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2435 42. It is now possible to set a limit on the number of times the match()
2436 function is called in a call to pcre_exec(). This facility makes it possible to
2437 limit the amount of recursion and backtracking, though not in a directly
2438 obvious way, because the match() function is used in a number of different
2439 circumstances. The count starts from zero for each position in the subject
2440 string (for non-anchored patterns). The default limit is, for compatibility, a
2441 large number, namely 10 000 000. You can change this in two ways:
2443 (a) When configuring PCRE before making, you can use --with-match-limit=n
2444 to set a default value for the compiled library.
2446 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2447 a different value is set. See 45 below.
2449 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2451 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2452 of things that can be changed at compile time. The first argument specifies
2453 what is wanted and the second points to where the information is to be placed.
2454 The current list of available information is:
2458 The output is an integer that is set to one if UTF-8 support is available;
2459 otherwise it is set to zero.
2463 The output is an integer that it set to the value of the code that is used for
2464 newline. It is either LF (10) or CR (13).
2468 The output is an integer that contains the number of bytes used for internal
2469 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2473 The output is an integer that contains the threshold above which the POSIX
2474 interface uses malloc() for output vectors. See item 31 above.
2478 The output is an unsigned integer that contains the default limit of the number
2479 of match() calls in a pcre_exec() execution. See 42 above.
2481 44. pcretest has been upgraded by the addition of the -C option. This causes it
2482 to extract all the available output from the new pcre_config() function, and to
2483 output it. The program then exits immediately.
2485 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2486 order to support additional features. One way would have been to define
2487 pcre_exec2() (for example) with extra arguments, but this would not have been
2488 extensible, and would also have required all calls to the original function to
2489 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2490 is used for passing in "extra" data from pcre_study().
2492 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2493 contains the following fields:
2495 flags a bitmap indicating which of the following fields are set
2496 study_data opaque data from pcre_study()
2497 match_limit a way of specifying a limit on match() calls for a specific
2498 call to pcre_exec()
2499 callout_data data for callouts (see 49 below)
2501 The flag bits are also defined in pcre.h, and are
2507 The pcre_study() function now returns one of these new pcre_extra blocks, with
2508 the actual study data pointed to by the study_data field, and the
2509 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2510 before. That is, this change is entirely upwards-compatible and requires no
2511 change to existing code.
2513 If you want to pass in additional data to pcre_exec(), you can either place it
2514 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2515 block.
2517 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2518 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2519 times with different match limits, until it finds the minimum value needed for
2520 pcre_exec() to complete. The value is then output. This can be instructive; for
2521 most simple matches the number is quite small, but for pathological cases it
2522 gets very large very quickly.
2524 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2525 returns the size of the data block pointed to by the study_data field in a
2526 pcre_extra block, that is, the value that was passed as the argument to
2527 pcre_malloc() when PCRE was getting memory in which to place the information
2528 created by pcre_study(). The fourth argument should point to a size_t variable.
2529 pcretest has been extended so that this information is shown after a successful
2530 pcre_study() call when information about the compiled regex is being displayed.
2532 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2533 because what follows is always an absolute path. (Later: it turns out that this
2534 is more than cosmetic for MinGW, because it doesn't like empty path
2535 components.)
2537 49. Some changes have been made to the callout feature (see 28 above):
2539 (i) A callout function now has three choices for what it returns:
2541 0 => success, carry on matching
2542 > 0 => failure at this point, but backtrack if possible
2543 < 0 => serious error, return this value from pcre_exec()
2545 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2546 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2547 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2548 use by callout functions. It will never be used by PCRE itself.
2550 (ii) The pcre_extra structure (see 45 above) has a void * field called
2551 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2552 pcre_callout_block structure has a field of the same name. The contents of
2553 the field passed in the pcre_extra structure are passed to the callout
2554 function in the corresponding field in the callout block. This makes it
2555 easier to use the same callout-containing regex from multiple threads. For
2556 testing, the pcretest program has a new data escape
2558 \C*n pass the number n (may be negative) as callout_data
2560 If the callout function in pcretest receives a non-zero value as
2561 callout_data, it returns that value.
2563 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2564 there were some redundant $(CFLAGS) in commands that are now specified as
2565 $(LINK), which already includes $(CFLAGS).
2567 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2568 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2569 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2570 one-byte characters throughout. Note that case-insensitive matching applies
2571 only to characters whose values are less than 256. PCRE doesn't support the
2572 notion of cases for higher-valued characters.
2574 (i) A character class whose characters are all within 0-255 is handled as
2575 a bit map, and the map is inverted for negative classes. Previously, a
2576 character > 255 always failed to match such a class; however it should
2577 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2579 (ii) A negated character class with a single character < 255 is coded as
2580 "not this character" (OP_NOT). This wasn't working properly when the test
2581 character was multibyte, either singly or repeated.
2583 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2584 mode, for example: \x{100}{2,3}.
2586 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2587 singly or repeated) now correctly test multibyte characters. However,
2588 PCRE doesn't recognize any characters with values greater than 255 as
2589 digits, spaces, or word characters. Such characters always match \D, \S,
2590 and \W, and never match \d, \s, or \w.
2592 (v) Classes may now contain characters and character ranges with values
2593 greater than 255. For example: [ab\x{100}-\x{400}].
2595 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2596 PCRE in UTF-8 mode.
2598 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2599 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2600 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2601 value.)
2603 53. The single man page has become too large. I have therefore split it up into
2604 a number of separate man pages. These also give rise to individual HTML pages;
2605 these are now put in a separate directory, and there is an index.html page that
2606 lists them all. Some hyperlinking between the pages has been installed.
2608 54. Added convenience functions for handling named capturing parentheses.
2610 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2611 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2612 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2613 are faulted.
2615 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2616 calling configure. These values are used when compiling the dftables.c program
2617 which is run to generate the source of the default character tables. They
2618 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2619 you will need to set these values.
2621 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2624 Version 3.9 02-Jan-02
2625 ---------------------
2627 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2629 2. If --disable-static was given, the building process failed when trying to
2630 build pcretest and pcregrep. (For some reason it was using libtool to compile
2631 them, which is not right, as they aren't part of the library.)
2634 Version 3.8 18-Dec-01
2635 ---------------------
2637 1. The experimental UTF-8 code was completely screwed up. It was packing the
2638 bytes in the wrong order. How dumb can you get?
2641 Version 3.7 29-Oct-01
2642 ---------------------
2644 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2645 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2646 this didn't happen under Solaris 8, where I normally test things.
2648 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2649 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2650 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2651 this fix an hour or so after the initial 3.7 release.)
2654 Version 3.6 23-Oct-01
2655 ---------------------
2657 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2658 offsets passed as NULL with zero offset count.
2660 2. The config.guess and config.sub files had not been updated when I moved to
2661 the latest autoconf.
2664 Version 3.5 15-Aug-01
2665 ---------------------
2667 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2668 had been forgotten.
2670 2. By using declared but undefined structures, we can avoid using "void"
2671 definitions in pcre.h while keeping the internal definitions of the structures
2672 private.
2674 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2675 user point of view, this means that both static and shared libraries are built
2676 by default, but this can be individually controlled. More of the work of
2677 handling this static/shared cases is now inside libtool instead of PCRE's make
2678 file.
2680 4. The pcretest utility is now installed along with pcregrep because it is
2681 useful for users (to test regexs) and by doing this, it automatically gets
2682 relinked by libtool. The documentation has been turned into a man page, so
2683 there are now .1, .txt, and .html versions in /doc.
2685 5. Upgrades to pcregrep:
2686 (i) Added long-form option names like gnu grep.
2687 (ii) Added --help to list all options with an explanatory phrase.
2688 (iii) Added -r, --recursive to recurse into sub-directories.
2689 (iv) Added -f, --file to read patterns from a file.
2691 6. pcre_exec() was referring to its "code" argument before testing that
2692 argument for NULL (and giving an error if it was NULL).
2694 7. Upgraded Makefile.in to allow for compiling in a different directory from
2695 the source directory.
2697 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2698 options bits, the pointer it was passed was to an int instead of to an unsigned
2699 long int. This mattered only on 64-bit systems.
2701 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2702 generated) instead of pcre.in, which it its source. Also made the same change
2703 in several of the .c files.
2705 10. A new release of gcc defines printf() as a macro, which broke pcretest
2706 because it had an ifdef in the middle of a string argument for printf(). Fixed
2707 by using separate calls to printf().
2709 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2710 script, to force use of CR or LF instead of \n in the source. On non-Unix
2711 systems, the value can be set in config.h.
2713 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2714 absolute limit. Changed the text of the error message to make this clear, and
2715 likewise updated the man page.
2717 13. The limit of 99 on the number of capturing subpatterns has been removed.
2718 The new limit is 65535, which I hope will not be a "real" limit.
2721 Version 3.4 22-Aug-00
2722 ---------------------
2724 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2726 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2729 Version 3.3 01-Aug-00
2730 ---------------------
2732 1. If an octal character was given, but the value was greater than \377, it
2733 was not getting masked to the least significant bits, as documented. This could
2734 lead to crashes in some systems.
2736 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2737 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2739 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2740 These just pass their arguments on to (pcre_free)(), but they are provided
2741 because some uses of PCRE bind it to non-C systems that can call its functions,
2742 but cannot call free() or pcre_free() directly.
2744 4. Add "make test" as a synonym for "make check". Corrected some comments in
2745 the Makefile.
2747 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2748 Makefile.
2750 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2751 command called pgrep for grepping around the active processes.
2753 7. Added the beginnings of support for UTF-8 character strings.
2755 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2756 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2757 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2758 out for the ar command.)
2761 Version 3.2 12-May-00
2762 ---------------------
2764 This is purely a bug fixing release.
2766 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2767 of ZA. This was just one example of several cases that could provoke this bug,
2768 which was introduced by change 9 of version 2.00. The code for breaking
2769 infinite loops after an iteration that matches an empty string was't working
2770 correctly.
2772 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2773 when matched against abbab (for example). After matching an empty string, it
2774 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2775 caused it to match further down the string than it should.
2777 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2778 was there because it doesn't seem to be needed, and it causes trouble on some
2779 systems, as it is not a Standard C header. It has been removed.
2781 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2782 were reported on the Macintosh. The changes were from
2784 while ((c = *(++ptr)) != 0 && c != '\n');
2785 to
2786 while ((c = *(++ptr)) != 0 && c != '\n') ;
2788 Totally extraordinary, but if that's what it takes...
2790 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2791 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2792 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2793 assumes the way PCRE uses memmove() (always moving upwards).
2795 6. PCRE is being used in one environment where strchr() is not available. There
2796 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2797 faster code anyway.
2800 Version 3.1 09-Feb-00
2801 ---------------------
2803 The only change in this release is the fixing of some bugs in Makefile.in for
2804 the "install" target:
2806 (1) It was failing to install pcreposix.h.
2808 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2811 Version 3.0 01-Feb-00
2812 ---------------------
2814 1. Add support for the /+ modifier to perltest (to output $` like it does in
2815 pcretest).
2817 2. Add support for the /g modifier to perltest.
2819 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2820 matches null strings.
2822 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2823 pattern. Perl treats empty patterns specially - it reuses the most recent
2824 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2825 effect.
2827 5. The POSIX interface was broken in that it was just handing over the POSIX
2828 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2829 required a bigger vector, with some working space on the end. This means that
2830 the POSIX wrapper now has to get and free some memory, and copy the results.
2832 6. Added some simple autoconf support, placing the test data and the
2833 documentation in separate directories, re-organizing some of the
2834 information files, and making it build pcre-config (a GNU standard). Also added
2835 libtool support for building PCRE as a shared library, which is now the
2836 default.
2838 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2839 09 are not valid octal constants. Single digits will be used for minor values
2840 less than 10.
2842 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2843 existing programs that set these in the POSIX interface can use PCRE without
2844 modification.
2846 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2847 return all that pcre_info() returns, plus additional data. The pcre_info()
2848 function is retained for compatibility, but is considered to be obsolete.
2850 10. Added experimental recursion feature (?R) to handle one common case that
2851 Perl 5.6 will be able to do with (?p{...}).
2853 11. Added support for POSIX character classes like [:alpha:], which Perl is
2854 adopting.
2857 Version 2.08 31-Aug-99
2858 ----------------------
2860 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2861 trying to match at the startoffset position, but instead was moving forward to
2862 the next newline as if a previous match had failed.
2864 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2865 and could get into a loop if a null string was matched other than at the start
2866 of the subject.
2868 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2869 be distinguished at compile time, and for completeness also added PCRE_DATE.
2871 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2872 in GnuWin32 environments.
2875 Version 2.07 29-Jul-99
2876 ----------------------
2878 1. The documentation is now supplied in plain text form and HTML as well as in
2879 the form of man page sources.
2881 2. C++ compilers don't like assigning (void *) values to other pointer types.
2882 In particular this affects malloc(). Although there is no problem in Standard
2883 C, I've put in casts to keep C++ compilers happy.
2885 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2886 should be (const char *).
2888 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2889 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2890 However, I haven't made this a standard facility. The documentation doesn't
2891 mention it, and the Makefile doesn't support it.
2893 5. The Makefile now contains an "install" target, with editable destinations at
2894 the top of the file. The pcretest program is not installed.
2896 6. pgrep -V now gives the PCRE version number and date.
2898 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2899 causing the entire string to be ignored, instead of just the last character.
2901 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2902 non-matching string, it can take a very, very long time, even for strings of
2903 quite modest length, because of the nested recursion. PCRE now does better in
2904 some of these cases. It does this by remembering the last required literal
2905 character in the pattern, and pre-searching the subject to ensure it is present
2906 before running the real match. In other words, it applies a heuristic to detect
2907 some types of certain failure quickly, and in the above example, if presented
2908 with a string that has no trailing " it gives "no match" very quickly.
2910 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2911 other alternatives are tried instead.
2914 Version 2.06 09-Jun-99
2915 ----------------------
2917 1. Change pcretest's output for amount of store used to show just the code
2918 space, because the remainder (the data block) varies in size between 32-bit and
2919 64-bit systems.
2921 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2922 start matching at. This allows lookbehinds to work when searching for multiple
2923 occurrences in a string.
2925 3. Added additional options to pcretest for testing multiple occurrences:
2927 /+ outputs the rest of the string that follows a match
2928 /g loops for multiple occurrences, using the new startoffset argument
2929 /G loops for multiple occurrences by passing an incremented pointer
2931 4. PCRE wasn't doing the "first character" optimization for patterns starting
2932 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2933 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2934 the letter 'x'. On long subject strings, this gives a significant speed-up.
2937 Version 2.05 21-Apr-99
2938 ----------------------
2940 1. Changed the type of magic_number from int to long int so that it works
2941 properly on 16-bit systems.
2943 2. Fixed a bug which caused patterns starting with .* not to work correctly
2944 when the subject string contained newline characters. PCRE was assuming
2945 anchoring for such patterns in all cases, which is not correct because .* will
2946 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2947 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2948 must be retried after every newline in the subject.
2951 Version 2.04 18-Feb-99
2952 ----------------------
2954 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2955 computation of the store needed to hold the pattern was incorrect (too large).
2956 If such patterns were nested a few deep, this could multiply and become a real
2957 problem.
2959 2. Added /M option to pcretest to show the memory requirement of a specific
2960 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2962 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2963 compiled in such a way that the backtracking after subsequent failure was
2964 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2965 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2968 Version 2.03 02-Feb-99
2969 ----------------------
2971 1. Fixed typo and small mistake in man page.
2973 2. Added 4th condition (GPL supersedes if conflict) and created separate
2974 LICENCE file containing the conditions.
2976 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2977 Perl, that is the internal \ allows the delimiter to be included in the
2978 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2979 the final delimiter, add \ to the end of the pattern (to test the error).
2981 4. Added the convenience functions for extracting substrings after a successful
2982 match. Updated pcretest to make it able to test these functions.
2985 Version 2.02 14-Jan-99
2986 ----------------------
2988 1. Initialized the working variables associated with each extraction so that
2989 their saving and restoring doesn't refer to uninitialized store.
2991 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2992 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2993 fix the problem.
2995 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2996 calls, and wasn't printing the correct value for compiling calls. Increased the
2997 default value of LOOPREPEAT, and the number of significant figures in the
2998 times.
3000 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3002 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3003 a building problem on Windows NT with a FAT file system.
3006 Version 2.01 21-Oct-98
3007 ----------------------
3009 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3010 to character tables built by pcre_maketables() in the current locale. If NULL
3011 is passed, the default tables are used.
3014 Version 2.00 24-Sep-98
3015 ----------------------
3017 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3018 it any more.
3020 2. Allow quantification of (?>) groups, and make it work correctly.
3022 3. The first character computation wasn't working for (?>) groups.
3024 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3025 end of the subject) and add 5.005's \z, which really does match only at the
3026 very end of the subject.
3028 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3030 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3031 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3032 localized options. All options to pcre_study() were also removed.
3034 7. Add other new features from 5.005:
3036 $(?<= positive lookbehind
3037 $(?<! negative lookbehind
3038 (?imsx-imsx) added the unsetting capability
3039 such a setting is global if at outer level; local otherwise
3040 (?imsx-imsx:) non-capturing groups with option setting
3041 (?(cond)re|re) conditional pattern matching
3043 A backreference to itself in a repeated group matches the previous
3044 captured string.
3046 8. General tidying up of studying (both automatic and via "study")
3047 consequential on the addition of new assertions.
3049 9. As in 5.005, unlimited repeated groups that could match an empty substring
3050 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3051 runtime if any iteration does actually match an empty substring.
3053 10. Include the RunTest script in the distribution.
3055 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3056 discrepancies, some of which were old and were also with respect to 5.004. They
3057 have now been fixed.
3060 Version 1.09 28-Apr-98
3061 ----------------------
3063 1. A negated single character class followed by a quantifier with a minimum
3064 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3065 program crashes, or just wrong answers. This did not apply to negated classes
3066 containing more than one character, or to minima other than one.
3069 Version 1.08 27-Mar-98
3070 ----------------------
3072 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3074 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3075 latter must appear before anything that relies on it in the pattern.
3078 Version 1.07 16-Feb-98
3079 ----------------------
3081 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3082 repeat of a potentially empty string).
3085 Version 1.06 23-Jan-98
3086 ----------------------
3088 1. Added Markus Oberhumer's little patches for C++.
3090 2. Literal strings longer than 255 characters were broken.
3093 Version 1.05 23-Dec-97
3094 ----------------------
3096 1. Negated character classes containing more than one character were failing if
3097 PCRE_CASELESS was set at run time.
3100 Version 1.04 19-Dec-97
3101 ----------------------
3103 1. Corrected the man page, where some "const" qualifiers had been omitted.
3105 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3106 input syntax.
3108 3. Fixed memory leak which occurred when a regex with back references was
3109 matched with an offsets vector that wasn't big enough. The temporary memory
3110 that is used in this case wasn't being freed if the match failed.
3112 4. Tidied pcretest to ensure it frees memory that it gets.
3114 5. Temporary memory was being obtained in the case where the passed offsets
3115 vector was exactly big enough.
3117 6. Corrected definition of offsetof() from change 5 below.
3119 7. I had screwed up change 6 below and broken the rules for the use of
3120 setjmp(). Now fixed.
3123 Version 1.03 18-Dec-97
3124 ----------------------
3126 1. A erroneous regex with a missing opening parenthesis was correctly
3127 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3128 on some systems.
3130 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3131 it was reported that one broken compiler failed on the former because "code" is
3132 also an independent variable.
3134 3. The erroneous regex a[]b caused an array overrun reference.
3136 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3137 fail on data ending with that character. (It was going on too far, and checking
3138 the next character, typically a binary zero.) This was specific to the
3139 optimized code for single-character negative classes.
3141 5. Added a contributed patch from the TIN world which does the following:
3143 + Add an undef for memmove, in case the the system defines a macro for it.
3145 + Add a definition of offsetof(), in case there isn't one. (I don't know
3146 the reason behind this - offsetof() is part of the ANSI standard - but
3147 it does no harm).
3149 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3150 most of the places where whitespace preceded '#'. I have given up and
3151 allowed the remaining 2 cases to be at the margin.
3153 + Rename some variables in pcre to eliminate shadowing. This seems very
3154 pedantic, but does no harm, of course.
3156 6. Moved the call to setjmp() into its own function, to get rid of warnings
3157 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3159 7. Constructs such as \d{8,} were compiling into the equivalent of
3160 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3161 outcome, but in this particular case used more store than had been allocated,
3162 which caused the bug to be discovered because it threw up an internal error.
3164 8. The debugging code in both pcre and pcretest for outputting the compiled
3165 form of a regex was going wrong in the case of back references followed by
3166 curly-bracketed repeats.
3169 Version 1.02 12-Dec-97
3170 ----------------------
3172 1. Typos in pcre.3 and comments in the source fixed.
3174 2. Applied a contributed patch to get rid of places where it used to remove
3175 'const' from variables, and fixed some signed/unsigned and uninitialized
3176 variable warnings.
3178 3. Added the "runtest" target to Makefile.
3180 4. Set default compiler flag to -O2 rather than just -O.
3183 Version 1.01 19-Nov-97
3184 ----------------------
3186 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3187 like /([ab]*)*/, that is, for classes with more than one character in them.
3189 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3190 as /((?>a*))*/ (a PCRE_EXTRA facility).
3193 Version 1.00 18-Nov-97
3194 ----------------------
3196 1. Added compile-time macros to support systems such as SunOS4 which don't have
3197 memmove() or strerror() but have other things that can be used instead.
3199 2. Arranged that "make clean" removes the executables.
3202 Version 0.99 27-Oct-97
3203 ----------------------
3205 1. Fixed bug in code for optimizing classes with only one character. It was
3206 initializing a 32-byte map regardless, which could cause it to run off the end
3207 of the memory it had got.
3209 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3212 Version 0.98 22-Oct-97
3213 ----------------------
3215 1. Fixed bug in code for handling temporary memory usage when there are more
3216 back references than supplied space in the ovector. This could cause segfaults.
3219 Version 0.97 21-Oct-97
3220 ----------------------
3222 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3224 2. Optimized negated single characters not to use a bit map.
3226 3. Brought error texts together as macro definitions; clarified some of them;
3227 fixed one that was wrong - it said "range out of order" when it meant "invalid
3228 escape sequence".
3230 4. Changed some char * arguments to const char *.
3232 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3234 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3235 pcretest.
3238 Version 0.96 16-Oct-97
3239 ----------------------
3241 1. Added a simple "pgrep" utility to the distribution.
3243 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3244 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3245 where "ddd" means "one or more decimal digits".
3247 3. Fixed serious bug. If a pattern had a back reference, but the call to
3248 pcre_exec() didn't supply a large enough ovector to record the related
3249 identifying subpattern, the match always failed. PCRE now remembers the number
3250 of the largest back reference, and gets some temporary memory in which to save
3251 the offsets during matching if necessary, in order to ensure that
3252 backreferences always work.
3254 4. Increased the compatibility with Perl in a number of ways:
3256 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3257 to request this handling. The option can be set at compile or exec time.
3259 (b) $ matches before a terminating newline by default; an option
3260 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3261 mode). The option can be set at compile or exec time.
3263 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3264 the same as Perl's. If the decimal number it represents is less than 10
3265 or there aren't that many previous left capturing parentheses, an octal
3266 escape is read. Inside a character class, it's always an octal escape,
3267 even if it is a single digit.
3269 (d) An escaped but undefined alphabetic character is taken as a literal,
3270 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3271 escapes.
3273 (e) {0} is now permitted. (The previous item is removed from the compiled
3274 pattern).
3276 5. Changed all the names of code files so that the basic parts are no longer
3277 than 10 characters, and abolished the teeny "globals.c" file.
3279 6. Changed the handling of character classes; they are now done with a 32-byte
3280 bit map always.
3282 7. Added the -d and /D options to pcretest to make it possible to look at the
3283 internals of compilation without having to recompile pcre.
3286 Version 0.95 23-Sep-97
3287 ----------------------
3289 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3290 \x20 at the start of a run of normal characters. These were being treated as
3291 real characters, instead of the source characters being re-checked.
3294 Version 0.94 18-Sep-97
3295 ----------------------
3297 1. The functions are now thread-safe, with the caveat that the global variables
3298 containing pointers to malloc() and free() or alternative functions are the
3299 same for all threads.
3301 2. Get pcre_study() to generate a bitmap of initial characters for non-
3302 anchored patterns when this is possible, and use it if passed to pcre_exec().
3305 Version 0.93 15-Sep-97
3306 ----------------------
3308 1. /(b)|(:+)/ was computing an incorrect first character.
3310 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3311 but not actually doing anything yet.
3313 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3314 as Perl does (e.g. [-az] or [az-]).
3316 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3317 all possible positions.
3319 5. Split up into different modules to avoid including unneeded functions in a
3320 compiled binary. However, compile and exec are still in one module. The "study"
3321 function is split off.
3323 6. The character tables are now in a separate module whose source is generated
3324 by an auxiliary program - but can then be edited by hand if required. There are
3325 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3326 toupper() in the code.
3328 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3329 make them global. Abolish the function for setting them, as the caller can now
3330 set them directly.
3333 Version 0.92 11-Sep-97
3334 ----------------------
3336 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3337 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3339 2. Caseless matching was not working in character classes if the characters in
3340 the pattern were in upper case.
3342 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3344 4. Make PCRE_ANCHORED public and accept as a compile option.
3346 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3347 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3348 pass them.
3350 6. Give an error if bad option bits passed at compile or run time.
3352 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3353 pcretest to cause it to pass that flag.
3355 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3356 options, and the first character, if set.
3358 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3361 Version 0.91 10-Sep-97
3362 ----------------------
3364 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3365 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3367 2. PCRE was looping on encountering an indefinitely repeated back reference to
3368 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3369 Perl does - treats the match as successful.
3371 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5