ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1553 - (show annotations)
Tue Apr 28 11:36:24 2015 UTC (4 years, 9 months ago) by ph10
File size: 277235 byte(s)
Error occurred while calculating annotation data.
Final tidies for 8.37
1 ChangeLog for PCRE
2 ------------------
4 Version 8.37 28-April-2015
5 --------------------------
7 1. When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
8 for those parentheses to be closed with whatever has been captured so far.
9 However, it was failing to mark any other groups between the hightest
10 capture so far and the currrent group as "unset". Thus, the ovector for
11 those groups contained whatever was previously there. An example is the
12 pattern /(x)|((*ACCEPT))/ when matched against "abcd".
14 2. If an assertion condition was quantified with a minimum of zero (an odd
15 thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
17 3. If a pattern in pcretest input had the P (POSIX) modifier followed by an
18 unrecognized modifier, a crash could occur.
20 4. An attempt to do global matching in pcretest with a zero-length ovector
21 caused a crash.
23 5. Fixed a memory leak during matching that could occur for a subpattern
24 subroutine call (recursive or otherwise) if the number of captured groups
25 that had to be saved was greater than ten.
27 6. Catch a bad opcode during auto-possessification after compiling a bad UTF
28 string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
29 UTF with NO_UTF_CHECK is documented as having an undefined outcome.
31 7. A UTF pattern containing a "not" match of a non-ASCII character and a
32 subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
34 8. When a pattern is compiled, it remembers the highest back reference so that
35 when matching, if the ovector is too small, extra memory can be obtained to
36 use instead. A conditional subpattern whose condition is a check on a
37 capture having happened, such as, for example in the pattern
38 /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
39 setting the highest backreference number. This mattered only if pcre_exec()
40 was called with an ovector that was too small to hold the capture, and there
41 was no other kind of back reference (a situation which is probably quite
42 rare). The effect of the bug was that the condition was always treated as
43 FALSE when the capture could not be consulted, leading to a incorrect
44 behaviour by pcre_exec(). This bug has been fixed.
46 9. A reference to a duplicated named group (either a back reference or a test
47 for being set in a conditional) that occurred in a part of the pattern where
48 PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
49 to be incorrectly calculated, leading to overwriting.
51 10. A mutually recursive set of back references such as (\2)(\1) caused a
52 segfault at study time (while trying to find the minimum matching length).
53 The infinite loop is now broken (with the minimum length unset, that is,
54 zero).
56 11. If an assertion that was used as a condition was quantified with a minimum
57 of zero, matching went wrong. In particular, if the whole group had
58 unlimited repetition and could match an empty string, a segfault was
59 likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
60 assertions to be quantified, but not if they are being used as conditions,
61 so the above pattern is faulted by Perl. PCRE has now been changed so that
62 it also rejects such patterns.
64 12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
65 failed to allow the zero-repeat case if pcre2_exec() was called with an
66 ovector too small to capture the group.
68 13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
69 Red Hat Product Security:
71 (a) A crash if /K and /F were both set with the option to save the compiled
72 pattern.
74 (b) Another crash if the option to print captured substrings in a callout
75 was combined with setting a null ovector, for example \O\C+ as a subject
76 string.
78 14. A pattern such as "((?2){0,1999}())?", which has a group containing a
79 forward reference repeated a large (but limited) number of times within a
80 repeated outer group that has a zero minimum quantifier, caused incorrect
81 code to be compiled, leading to the error "internal error:
82 previously-checked referenced subpattern not found" when an incorrect
83 memory address was read. This bug was reported as "heap overflow",
84 discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
85 CVE-2015-2325.
87 23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
88 call within a group that also contained a recursive back reference caused
89 incorrect code to be compiled. This bug was reported as "heap overflow",
90 discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
91 number CVE-2015-2326.
93 24. Computing the size of the JIT read-only data in advance has been a source
94 of various issues, and new ones are still appear unfortunately. To fix
95 existing and future issues, size computation is eliminated from the code,
96 and replaced by on-demand memory allocation.
98 25. A pattern such as /(?i)[A-`]/, where characters in the other case are
99 adjacent to the end of the range, and the range contained characters with
100 more than one other case, caused incorrect behaviour when compiled in UTF
101 mode. In that example, the range a-j was left out of the class.
103 26. Fix JIT compilation of conditional blocks, which assertion
104 is converted to (*FAIL). E.g: /(?(?!))/.
106 27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
107 discovered by the LLVM fuzzer.
109 28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
110 when this assertion was used as a condition, for example (?(?!)a|b). In
111 pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
112 error about an unsupported item.
114 29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
115 possessification code could take exponential time to complete. A recursion
116 depth limit of 1000 has been imposed to limit the resources used by this
117 optimization.
119 30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
120 such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
121 because \S ensures they are all in the class. The code for doing this was
122 interacting badly with the code for computing the amount of space needed to
123 compile the pattern, leading to a buffer overflow. This bug was discovered
124 by the LLVM fuzzer.
126 31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
127 other kinds of group caused stack overflow at compile time. This bug was
128 discovered by the LLVM fuzzer.
130 32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
131 between a subroutine call and its quantifier was incorrectly compiled,
132 leading to buffer overflow or other errors. This bug was discovered by the
133 LLVM fuzzer.
135 33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
136 assertion after (?(. The code was failing to check the character after
137 (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
138 was discovered by the LLVM fuzzer.
140 34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
141 a fixed maximum following a group that contains a subroutine reference was
142 incorrectly compiled and could trigger buffer overflow. This bug was
143 discovered by the LLVM fuzzer.
145 35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
146 caused a stack overflow instead of the diagnosis of a non-fixed length
147 lookbehind assertion. This bug was discovered by the LLVM fuzzer.
149 36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
150 (e.g. /(?<=\Ka)/) could make pcregrep loop.
152 37. There was a similar problem to 36 in pcretest for global matches.
154 38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
155 and a subsequent item in the pattern caused a non-match, backtracking over
156 the repeated \X did not stop, but carried on past the start of the subject,
157 causing reference to random memory and/or a segfault. There were also some
158 other cases where backtracking after \C could crash. This set of bugs was
159 discovered by the LLVM fuzzer.
161 39. The function for finding the minimum length of a matching string could take
162 a very long time if mutual recursion was present many times in a pattern,
163 for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
164 method has been implemented. This infelicity was discovered by the LLVM
165 fuzzer.
167 40. Static linking against the PCRE library using the pkg-config module was
168 failing on missing pthread symbols.
171 Version 8.36 26-September-2014
172 ------------------------------
174 1. Got rid of some compiler warnings in the C++ modules that were shown up by
175 -Wmissing-field-initializers and -Wunused-parameter.
177 2. The tests for quantifiers being too big (greater than 65535) were being
178 applied after reading the number, and stupidly assuming that integer
179 overflow would give a negative number. The tests are now applied as the
180 numbers are read.
182 3. Tidy code in pcre_exec.c where two branches that used to be different are
183 now the same.
185 4. The JIT compiler did not generate match limit checks for certain
186 bracketed expressions with quantifiers. This may lead to exponential
187 backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
188 issue should be resolved now.
190 5. Fixed an issue, which occures when nested alternatives are optimized
191 with table jumps.
193 6. Inserted two casts and changed some ints to size_t in the light of some
194 reported 64-bit compiler warnings (Bugzilla 1477).
196 7. Fixed a bug concerned with zero-minimum possessive groups that could match
197 an empty string, which sometimes were behaving incorrectly in the
198 interpreter (though correctly in the JIT matcher). This pcretest input is
199 an example:
201 '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
204 the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
205 JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
206 for an empty string was breaking the inner loop and carrying on at a lower
207 level, when possessive repeated groups should always return to a higher
208 level as they have no backtrack points in them. The empty string test now
209 occurs at the outer level.
211 8. Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
212 ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
214 9. Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
215 doesn't).
217 10. Change 8.34/15 introduced a bug that caused the amount of memory needed
218 to hold a pattern to be incorrectly computed (too small) when there were
219 named back references to duplicated names. This could cause "internal
220 error: code overflow" or "double free or corruption" or other memory
221 handling errors.
223 11. When named subpatterns had the same prefixes, back references could be
224 confused. For example, in this pattern:
226 /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
228 the reference to 'Name' was incorrectly treated as a reference to a
229 duplicate name.
231 12. A pattern such as /^s?c/mi8 where the optional character has more than
232 one "other case" was incorrectly compiled such that it would only try to
233 match starting at "c".
235 13. When a pattern starting with \s was studied, VT was not included in the
236 list of possible starting characters; this should have been part of the
237 8.34/18 patch.
239 14. If a character class started [\Qx]... where x is any character, the class
240 was incorrectly terminated at the ].
242 15. If a pattern that started with a caseless match for a character with more
243 than one "other case" was studied, PCRE did not set up the starting code
244 unit bit map for the list of possible characters. Now it does. This is an
245 optimization improvement, not a bug fix.
247 16. The Unicode data tables have been updated to Unicode 7.0.0.
249 17. Fixed a number of memory leaks in pcregrep.
251 18. Avoid a compiler warning (from some compilers) for a function call with
252 a cast that removes "const" from an lvalue by using an intermediate
253 variable (to which the compiler does not object).
255 19. Incorrect code was compiled if a group that contained an internal recursive
256 back reference was optional (had quantifier with a minimum of zero). This
257 example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
258 caused segmentation faults because of stack overflows at compile time.
260 20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
261 group that is quantified with an indefinite repeat, caused a compile-time
262 loop which used up all the system stack and provoked a segmentation fault.
263 This was not the same bug as 19 above.
265 21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
266 Patch by Mike Frysinger.
269 Version 8.35 04-April-2014
270 --------------------------
272 1. A new flag is set, when property checks are present in an XCLASS.
273 When this flag is not set, PCRE can perform certain optimizations
274 such as studying these XCLASS-es.
276 2. The auto-possessification of character sets were improved: a normal
277 and an extended character set can be compared now. Furthermore
278 the JIT compiler optimizes more character set checks.
280 3. Got rid of some compiler warnings for potentially uninitialized variables
281 that show up only when compiled with -O2.
283 4. A pattern such as (?=ab\K) that uses \K in an assertion can set the start
284 of a match later then the end of the match. The pcretest program was not
285 handling the case sensibly - it was outputting from the start to the next
286 binary zero. It now reports this situation in a message, and outputs the
287 text from the end to the start.
289 5. Fast forward search is improved in JIT. Instead of the first three
290 characters, any three characters with fixed position can be searched.
291 Search order: first, last, middle.
293 6. Improve character range checks in JIT. Characters are read by an inprecise
294 function now, which returns with an unknown value if the character code is
295 above a certain threshold (e.g: 256). The only limitation is that the value
296 must be bigger than the threshold as well. This function is useful when
297 the characters above the threshold are handled in the same way.
299 7. The macros whose names start with RAWUCHAR are placeholders for a future
300 mode in which only the bottom 21 bits of 32-bit data items are used. To
301 make this more memorable for those maintaining the code, the names have
302 been changed to start with UCHAR21, and an extensive comment has been added
303 to their definition.
305 8. Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
306 to the export list in Makefile.am (they were accidentally omitted from the
307 8.34 tarball).
309 9. The informational output from pcretest used the phrase "starting byte set"
310 which is inappropriate for the 16-bit and 32-bit libraries. As the output
311 for "first char" and "need char" really means "non-UTF-char", I've changed
312 "byte" to "char", and slightly reworded the output. The documentation about
313 these values has also been (I hope) clarified.
315 10. Another JIT related optimization: use table jumps for selecting the correct
316 backtracking path, when more than four alternatives are present inside a
317 bracket.
319 11. Empty match is not possible, when the minimum length is greater than zero,
320 and there is no \K in the pattern. JIT should avoid empty match checks in
321 such cases.
323 12. In a caseless character class with UCP support, when a character with more
324 than one alternative case was not the first character of a range, not all
325 the alternative cases were added to the class. For example, s and \x{17f}
326 are both alternative cases for S: the class [RST] was handled correctly,
327 but [R-T] was not.
329 13. The configure.ac file always checked for pthread support when JIT was
330 enabled. This is not used in Windows, so I have put this test inside a
331 check for the presence of windows.h (which was already tested for).
333 14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
334 The algorithm provides a way to skip certain starting offsets, and usually
335 faster than linear prefix searches.
337 15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
338 as for 'fr_FR' and 'french'. For some reason, however, it then used the
339 Windows-specific input and output files, which have 'french' screwed in.
340 So this could never have worked. One of the problems with locales is that
341 they aren't always the same. I have now updated RunTest so that it checks
342 the output of the locale test (test 3) against three different output
343 files, and it allows the test to pass if any one of them matches. With luck
344 this should make the test pass on some versions of Solaris where it was
345 failing. Because of the uncertainty, the script did not used to stop if
346 test 3 failed; it now does. If further versions of a French locale ever
347 come to light, they can now easily be added.
349 16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
350 there was a message during ./configure, but it did not stop. This now
351 provokes an error. The invalid example in README has been corrected.
352 If a value less than the minimum is given, the minimum value has always
353 been used, but now a warning is given.
355 17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
356 was a bug in the test system, which is now fixed. Also, the list of various
357 configurations that are tested for each release did not have one with both
358 16/32 bits and --enable-bar-anycrlf. It now does.
360 18. pcretest was missing "-C bsr" for displaying the \R default setting.
362 19. Little endian PowerPC systems are supported now by the JIT compiler.
364 20. The fast forward newline mechanism could enter to an infinite loop on
365 certain invalid UTF-8 input. Although we don't support these cases
366 this issue can be fixed by a performance optimization.
368 21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
369 not take account if existing stack usage. There is now a new global
370 variable called pcre_stack_guard that can be set to point to an external
371 function to check stack availability. It is called at the start of
372 processing every parenthesized group.
374 22. A typo in the code meant that in ungreedy mode the max/min qualifier
375 behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
376 not match "ab".
378 23. When UTF was disabled, the JIT program reported some incorrect compile
379 errors. These messages are silenced now.
381 24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
382 compiler.
384 25. Change all the temporary files used in RunGrepTest to be different to those
385 used by RunTest so that the tests can be run simultaneously, for example by
386 "make -j check".
389 Version 8.34 15-December-2013
390 -----------------------------
392 1. Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
393 executable memory. Patch inspired by Carsten Klein.
395 2. ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
396 this macro is never tested and has no effect, because the work to support
397 coverage involves only compiling and linking options and special targets in
398 the Makefile. The comment in config.h implied that defining the macro would
399 enable coverage support, which is totally false. There was also support for
400 setting this macro in the CMake files (my fault, I just copied it from
401 configure). SUPPORT_GCOV has now been removed.
403 3. Make a small performance improvement in strlen16() and strlen32() in
404 pcretest.
406 4. Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
407 detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
408 these cases). There was also one in pcretest.c.
410 5. Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
412 6. In UTF mode, the code for checking whether a group could match an empty
413 string (which is used for indefinitely repeated groups to allow for
414 breaking an infinite loop) was broken when the group contained a repeated
415 negated single-character class with a character that occupied more than one
416 data item and had a minimum repetition of zero (for example, [^\x{100}]* in
417 UTF-8 mode). The effect was undefined: the group might or might not be
418 deemed as matching an empty string, or the program might have crashed.
420 7. The code for checking whether a group could match an empty string was not
421 recognizing that \h, \H, \v, \V, and \R must match a character.
423 8. Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
424 an empty string. If it can, pcretest shows this in its information output.
426 9. Fixed two related bugs that applied to Unicode extended grapheme clusters
427 that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
428 matched by pcre_exec() without using JIT:
430 (a) If the rest of the pattern did not match after a maximal run of
431 grapheme clusters, the code for backing up to try with fewer of them
432 did not always back up over a full grapheme when characters that do not
433 have the modifier quality were involved, e.g. Hangul syllables.
435 (b) If the match point in a subject started with modifier character, and
436 there was no match, the code could incorrectly back up beyond the match
437 point, and potentially beyond the first character in the subject,
438 leading to a segfault or an incorrect match result.
440 10. A conditional group with an assertion condition could lead to PCRE
441 recording an incorrect first data item for a match if no other first data
442 item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
443 first data item, and therefore matched "ca" after "c" instead of at the
444 start.
446 11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
447 bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
449 12. The source of pcregrep now includes z/OS-specific code so that it can be
450 compiled for z/OS as part of the special z/OS distribution.
452 13. Added the -T and -TM options to pcretest.
454 14. The code in pcre_compile.c for creating the table of named capturing groups
455 has been refactored. Instead of creating the table dynamically during the
456 actual compiling pass, the information is remembered during the pre-compile
457 pass (on the stack unless there are more than 20 named groups, in which
458 case malloc() is used) and the whole table is created before the actual
459 compile happens. This has simplified the code (it is now nearly 150 lines
460 shorter) and prepared the way for better handling of references to groups
461 with duplicate names.
463 15. A back reference to a named subpattern when there is more than one of the
464 same name now checks them in the order in which they appear in the pattern.
465 The first one that is set is used for the reference. Previously only the
466 first one was inspected. This change makes PCRE more compatible with Perl.
468 16. Unicode character properties were updated from Unicode 6.3.0.
470 17. The compile-time code for auto-possessification has been refactored, based
471 on a patch by Zoltan Herczeg. It now happens after instead of during
472 compilation. The code is cleaner, and more cases are handled. The option
473 PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
474 options in pcretest are provided to set it. It can also be set by
475 (*NO_AUTO_POSSESS) at the start of a pattern.
477 18. The character VT has been added to the default ("C" locale) set of
478 characters that match \s and are generally treated as white space,
479 following this same change in Perl 5.18. There is now no difference between
480 "Perl space" and "POSIX space". Whether VT is treated as white space in
481 other locales depends on the locale.
483 19. The code for checking named groups as conditions, either for being set or
484 for being recursed, has been refactored (this is related to 14 and 15
485 above). Processing unduplicated named groups should now be as fast at
486 numerical groups, and processing duplicated groups should be faster than
487 before.
489 20. Two patches to the CMake build system, by Alexander Barkov:
491 (1) Replace the "source" command by "." in CMakeLists.txt because
492 "source" is a bash-ism.
494 (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
495 without these the CMake build does not work on Solaris.
497 21. Perl has changed its handling of \8 and \9. If there is no previously
498 encountered capturing group of those numbers, they are treated as the
499 literal characters 8 and 9 instead of a binary zero followed by the
500 literals. PCRE now does the same.
502 22. Following Perl, added \o{} to specify codepoints in octal, making it
503 possible to specify values greater than 0777 and also making them
504 unambiguous.
506 23. Perl now gives an error for missing closing braces after \x{... instead of
507 treating the string as literal. PCRE now does the same.
509 24. RunTest used to grumble if an inappropriate test was selected explicitly,
510 but just skip it when running all tests. This make it awkward to run ranges
511 of tests when one of them was inappropriate. Now it just skips any
512 inappropriate tests, as it always did when running all tests.
514 25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
515 character types such as \d or \w, too many callouts were inserted, and the
516 data that they returned was rubbish.
518 26. In UCP mode, \s was not matching two of the characters that Perl matches,
519 namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
520 were matched by \h. The code has now been refactored so that the lists of
521 the horizontal and vertical whitespace characters used for \h and \v (which
522 are defined only in one place) are now also used for \s.
524 27. Add JIT support for the 64 bit TileGX architecture.
525 Patch by Jiong Wang (Tilera Corporation).
527 28. Possessive quantifiers for classes (both explicit and automatically
528 generated) now use special opcodes instead of wrapping in ONCE brackets.
530 29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
531 (because it's meaningless), this was not happening when PCRE_CASELESS was
532 set. Not wrong, but inefficient.
534 30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
535 properties for \w, \d, etc) is present in a test regex. Otherwise if the
536 test contains no characters greater than 255, Perl doesn't realise it
537 should be using Unicode semantics.
539 31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
540 [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
541 does in Unicode mode.
543 32. Added the "forbid" facility to pcretest so that putting tests into the
544 wrong test files can sometimes be quickly detected.
546 33. There is now a limit (default 250) on the depth of nesting of parentheses.
547 This limit is imposed to control the amount of system stack used at compile
548 time. It can be changed at build time by --with-parens-nest-limit=xxx or
549 the equivalent in CMake.
551 34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
552 errors. Perl warns for these when in warning mode, but PCRE has no facility
553 for giving warnings.
555 35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
556 However, this was not working for (?!) because it is optimized to (*FAIL),
557 for which PCRE does not allow quantifiers. The optimization is now disabled
558 when a quantifier follows (?!). I can't see any use for this, but it makes
559 things uniform.
561 36. Perl no longer allows group names to start with digits, so I have made this
562 change also in PCRE. It simplifies the code a bit.
564 37. In extended mode, Perl ignores spaces before a + that indicates a
565 possessive quantifier. PCRE allowed a space before the quantifier, but not
566 before the possessive +. It now does.
568 38. The use of \K (reset reported match start) within a repeated possessive
569 group such as (a\Kb)*+ was not working.
571 40. Document that the same character tables must be used at compile time and
572 run time, and that the facility to pass tables to pcre_exec() and
573 pcre_dfa_exec() is for use only with saved/restored patterns.
575 41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
576 features for Builds with MSVC:
578 1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
579 builds). The .rc files can be used to set FileDescription and many other
580 attributes.
582 2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
583 This allows higher-level build scripts which want .pdb files to avoid
584 hard-coding the exact files needed."
586 42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
587 mean "start of word" and "end of word", respectively, as a transition aid.
589 43. A minimizing repeat of a class containing codepoints greater than 255 in
590 non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
591 compiled to use the heap for recursion.
593 44. Got rid of some compiler warnings for unused variables when UTF but not UCP
594 is configured.
597 Version 8.33 28-May-2013
598 ------------------------
600 1. Added 'U' to some constants that are compared to unsigned integers, to
601 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
602 variables that are added to signed variables, to ensure the result is
603 signed and can be negated.
605 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
606 CMake config files.
608 3. Revise the creation of config.h.generic so that all boolean macros are
609 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
610 overriding via -D on the command line possible.
612 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
613 to unsigned int is reported to make a quite noticeable speed difference in
614 a specific Windows environment. Testing on Linux did also appear to show
615 some benefit (and it is clearly not harmful). Also fixed the definition of
616 Xop which should be unsigned.
618 5. Related to (4), changing the definition of the intermediate variable cc
619 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
620 improvements.
622 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
623 unnecessary spaces.
625 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
626 and later.
628 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
630 9. Optimizing fast_forward_start_bits in JIT.
632 10. Adding support for callouts in JIT, and fixing some issues revealed
633 during this work. Namely:
635 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
637 (b) Minimum length was not checked before the matching is started.
639 11. The value of capture_last that is passed to callouts was incorrect in some
640 cases when there was a capture on one path that was subsequently abandoned
641 after a backtrack. Also, the capture_last value is now reset after a
642 recursion, since all captures are also reset in this case.
644 12. The interpreter no longer returns the "too many substrings" error in the
645 case when an overflowing capture is in a branch that is subsequently
646 abandoned after a backtrack.
648 13. In the pathological case when an offset vector of size 2 is used, pcretest
649 now prints out the matched string after a yield of 0 or 1.
651 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
652 Only supported by the JIT compiler at the moment.
654 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
656 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
657 offset of the starting point of the matching process, provided the offsets
658 vector is large enough.
660 17. The \A escape now records a lookbehind value of 1, though its execution
661 does not actually inspect the previous character. This is to ensure that,
662 in partial multi-segment matching, at least one character from the old
663 segment is retained when a new segment is processed. Otherwise, if there
664 are no lookbehinds in the pattern, \A might match incorrectly at the start
665 of a new segment.
667 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
669 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
670 gives some modest performance improvement in 8-bit mode.
672 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
673 be expressed in certain programming languages using Universal Character
674 Names.
676 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
677 which points out that "non characters" are not "characters that may not
678 appear in Unicode strings" but rather "characters that are reserved for
679 internal use and have only local meaning".
681 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
682 there was a conditional group that depended on an assertion, if the
683 assertion was false, the callout that immediately followed the alternation
684 in the condition was skipped when pcre_exec() was used for matching.
686 23. Allow an explicit callout to be inserted before an assertion that is the
687 condition for a conditional group, for compatibility with automatic
688 callouts, which always insert a callout at this point.
690 24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
691 confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
693 25. (*PRUNE) is now supported by the JIT compiler.
695 26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
697 27. Fix the case where there are two or more SKIPs with arguments that may be
698 ignored.
700 28. (*SKIP) is now supported by the JIT compiler.
702 29. (*THEN) is now supported by the JIT compiler.
704 30. Update RunTest with additional test selector options.
706 31. The way PCRE handles backtracking verbs has been changed in two ways.
708 (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
709 SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
710 backtracking. In some cases this makes it more Perl-compatible, but Perl's
711 rather obscure rules do not always do the same thing.
713 (2) Previously, backtracking verbs were confined within assertions. This is
714 no longer the case for positive assertions, except for (*ACCEPT). Again,
715 this sometimes improves Perl compatibility, and sometimes does not.
717 32. A number of tests that were in test 2 because Perl did things differently
718 have been moved to test 1, because either Perl or PCRE has changed, and
719 these tests are now compatible.
721 32. Backtracking control verbs are now handled in the same way in JIT and
722 interpreter.
724 33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
725 contained a forward subroutine reference caused a compile error.
727 34. Auto-detect and optimize limited repetitions in JIT.
729 35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
730 blocking (*UTF) etc.
732 36. In the interpreter, maximizing pattern repetitions for characters and
733 character types now use tail recursion, which reduces stack usage.
735 37. The value of the max lookbehind was not correctly preserved if a compiled
736 and saved regex was reloaded on a host of different endianness.
738 38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
739 of the compiled pattern block, expand the flags field from 16 to 32 bits
740 because it was almost full.
742 39. Try madvise first before posix_madvise.
744 40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
745 with a pattern such as ^$. It has taken 4 years for anybody to notice! The
746 original change locked out all matches of empty strings. This has been
747 changed so that one match of an empty string per line is recognized.
748 Subsequent searches on the same line (for colouring or for --only-matching,
749 for example) do not recognize empty strings.
751 41. Applied a user patch to fix a number of spelling mistakes in comments.
753 42. Data lines longer than 65536 caused pcretest to crash.
755 43. Clarified the data type for length and startoffset arguments for pcre_exec
756 and pcre_dfa_exec in the function-specific man pages, where they were
757 explicitly stated to be in bytes, never having been updated. I also added
758 some clarification to the pcreapi man page.
760 44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
761 a segmentation fault.
764 Version 8.32 30-November-2012
765 -----------------------------
767 1. Improved JIT compiler optimizations for first character search and single
768 character iterators.
770 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
771 Patch by Daniel Richard G.
773 3. Single character iterator optimizations in the JIT compiler.
775 4. Improved JIT compiler optimizations for character ranges.
777 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
778 Reported by Giuseppe D'Angelo.
780 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
781 of a line, was being set incorrectly in cases where .* appeared inside
782 atomic brackets at the start of a pattern, or where there was a subsequent
783 *PRUNE or *SKIP.
785 7. Improved instruction cache flush for POWER/PowerPC.
786 Patch by Daniel Richard G.
788 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
789 grep:
791 (a) There is now no limit to the number of patterns to be matched.
793 (b) An error is given if a pattern is too long.
795 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
796 are now supported.
798 (d) --exclude-from and --include-from (multiple use) have been added.
800 (e) Exclusions and inclusions now apply to all files and directories, not
801 just to those obtained from scanning a directory recursively.
803 (f) Multiple uses of -f and --file-list are now supported.
805 (g) In a Windows environment, the default for -d has been changed from
806 "read" (the GNU grep default) to "skip", because otherwise the presence
807 of a directory in the file list provokes an error.
809 (h) The documentation has been revised and clarified in places.
811 9. Improve the matching speed of capturing brackets.
813 10. Changed the meaning of \X so that it now matches a Unicode extended
814 grapheme cluster.
816 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
817 out POSIX threads when JIT support is configured.
819 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
821 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
822 structure could go wrong in environments where size_t is not the same size
823 as int.
825 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
826 set.
828 15. The EBCDIC support had decayed; later updates to the code had included
829 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
830 general tidy up of EBCDIC-related issues, and the documentation was also
831 not quite right. There is now a test that can be run on ASCII systems to
832 check some of the EBCDIC-related things (but is it not a full test).
834 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
835 in a small tidy to the code.
837 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
839 18. If the --only-matching (-o) option in pcregrep is specified multiple
840 times, each one causes appropriate output. For example, -o1 -o2 outputs the
841 substrings matched by the 1st and 2nd capturing parentheses. A separating
842 string can be specified by --om-separator (default empty).
844 19. Improving the first n character searches.
846 20. Turn case lists for horizontal and vertical white space into macros so that
847 they are defined only once.
849 21. This set of changes together give more compatible Unicode case-folding
850 behaviour for characters that have more than one other case when UCP
851 support is available.
853 (a) The Unicode property table now has offsets into a new table of sets of
854 three or more characters that are case-equivalent. The MultiStage2.py
855 script that generates these tables (the pcre_ucd.c file) now scans
856 CaseFolding.txt instead of UnicodeData.txt for character case
857 information.
859 (b) The code for adding characters or ranges of characters to a character
860 class has been abstracted into a generalized function that also handles
861 case-independence. In UTF-mode with UCP support, this uses the new data
862 to handle characters with more than one other case.
864 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
865 whose other case is greater than 256 are now correctly matched
866 caselessly. Previously, the high codepoint matched the low one, but not
867 vice versa.
869 (d) The processing of \h, \H, \v, and \ in character classes now makes use
870 of the new class addition function, using character lists defined as
871 macros alongside the case definitions of 20 above.
873 (e) Caseless back references now work with characters that have more than
874 one other case.
876 (f) General caseless matching of characters with more than one other case
877 is supported.
879 22. Unicode character properties were updated from Unicode 6.2.0
881 23. Improved CMake support under Windows. Patch by Daniel Richard G.
883 24. Add support for 32-bit character strings, and UTF-32
885 25. Major JIT compiler update (code refactoring and bugfixing).
886 Experimental Sparc 32 support is added.
888 26. Applied a modified version of Daniel Richard G's patch to create
889 pcre.h.generic and config.h.generic by "make" instead of in the
890 PrepareRelease script.
892 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
893 pcre_compile.c when checking for a zero character.
895 28. Introducing a native interface for JIT. Through this interface, the compiled
896 machine code can be directly executed. The purpose of this interface is to
897 provide fast pattern matching, so several sanity checks are not performed.
898 However, feature tests are still performed. The new interface provides
899 1.4x speedup compared to the old one.
901 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
902 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
903 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
905 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
906 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
907 the "old" RFC 2279). Instead, it ended up passing a negative length to
908 pcre_exec().
910 31. Add support for GCC's visibility feature to hide internal functions.
912 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
913 "unknown -C option" after outputting 0 or 1.
915 33. There is now support for generating a code coverage report for the test
916 suite in environments where gcc is the compiler and lcov is installed. This
917 is mainly for the benefit of the developers.
919 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
920 unaddressable using valgrind annotations, allowing valgrind to detect
921 invalid memory accesses. This is mainly for the benefit of the developers.
923 25. (*UTF) can now be used to start a pattern in any of the three libraries.
925 26. Give configure error if --enable-cpp but no C++ compiler found.
928 Version 8.31 06-July-2012
929 -------------------------
931 1. Fixing a wrong JIT test case and some compiler warnings.
933 2. Removed a bashism from the RunTest script.
935 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
936 to unsigned type, result still unsigned" that was given by an MS compiler
937 on encountering the code "-sizeof(xxx)".
939 4. Partial matching support is added to the JIT compiler.
941 5. Fixed several bugs concerned with partial matching of items that consist
942 of more than one character:
944 (a) /^(..)\1/ did not partially match "aba" because checking references was
945 done on an "all or nothing" basis. This also applied to repeated
946 references.
948 (b) \R did not give a hard partial match if \r was found at the end of the
949 subject.
951 (c) \X did not give a hard partial match after matching one or more
952 characters at the end of the subject.
954 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
955 a partial match for the string "\r".
957 (e) When newline was set to CRLF, the metacharacter "." did not recognize
958 a partial match for a CR character at the end of the subject string.
960 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
961 running pcretest, the text "(JIT)" added to the output whenever JIT is
962 actually used to run the match.
964 7. Individual JIT compile options can be set in pcretest by following -s+[+]
965 or /S+[+] with a digit between 1 and 7.
967 8. OP_NOT now supports any UTF character not just single-byte ones.
969 9. (*MARK) control verb is now supported by the JIT compiler.
971 10. The command "./RunTest list" lists the available tests without actually
972 running any of them. (Because I keep forgetting what they all are.)
976 12. Applied a (slightly modified) user-supplied patch that improves performance
977 when the heap is used for recursion (compiled with --disable-stack-for-
978 recursion). Instead of malloc and free for each heap frame each time a
979 logical recursion happens, frames are retained on a chain and re-used where
980 possible. This sometimes gives as much as 30% improvement.
982 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
983 call.
985 14. As documented, (*COMMIT) is now confined to within a positive assertion.
987 15. It is now possible to link pcretest with libedit as an alternative to
988 libreadline.
990 16. (*COMMIT) control verb is now supported by the JIT compiler.
992 17. The Unicode data tables have been updated to Unicode 6.1.0.
994 18. Added --file-list option to pcregrep.
996 19. Added binary file support to pcregrep, including the -a, --binary-files,
997 -I, and --text options.
999 20. The madvise function is renamed for posix_madvise for QNX compatibility
1000 reasons. Fixed by Giuseppe D'Angelo.
1002 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1004 22. Changed the option for creating symbolic links for 16-bit man pages from
1005 -s to -sf so that re-installing does not cause issues.
1007 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1009 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1010 matches in certain environments (the workspace was not being correctly
1011 retained). Also added to pcre_dfa_exec() a simple plausibility check on
1012 some of the workspace data at the beginning of a restart.
1014 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1015 was not doing so when it should - probably a typo introduced by SVN 528
1016 (change 8.10/14).
1018 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1019 \w+ when the character tables indicated that \x{c4} was a word character.
1020 There were several related cases, all because the tests for doing a table
1021 lookup were testing for characters less than 127 instead of 255.
1023 27. If a pattern contains capturing parentheses that are not used in a match,
1024 their slots in the ovector are set to -1. For those that are higher than
1025 any matched groups, this happens at the end of processing. In the case when
1026 there were back references that the ovector was too small to contain
1027 (causing temporary malloc'd memory to be used during matching), and the
1028 highest capturing number was not used, memory off the end of the ovector
1029 was incorrectly being set to -1. (It was using the size of the temporary
1030 memory instead of the true size.)
1032 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1033 ovector size, it uses memory at the end of the block that it has got.
1035 29. Check for an overlong MARK name and give an error at compile time. The
1036 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1038 30. JIT compiler update.
1040 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1041 Rill for the patch.
1043 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1045 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1047 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1048 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1049 was enabled.
1051 35. Improve JIT code generation for greedy plus quantifier.
1053 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1054 1 to "aa" instead of to an empty string. The bug affected repeated groups
1055 that could potentially match an empty string.
1057 37. Optimizing single character iterators in JIT.
1059 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1060 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1061 codepoints that are too big for the mode are faulted, and in a UTF mode,
1062 disallowed codepoints are also faulted.
1064 39. If PCRE was compiled with UTF support, in three places in the DFA
1065 matcher there was code that should only have been obeyed in UTF mode, but
1066 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1067 processing when bytes with values greater than 127 were present. In 16-bit
1068 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1069 both cases the values are those that cannot be the first data item in a UTF
1070 character. The three items that might have provoked this were recursions,
1071 possessively repeated groups, and atomic groups.
1073 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1074 and pcregrep, because some OS require shared objects to be explicitly
1075 passed to ld, causing the link step to fail if they are not.
1077 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1078 mode, patterns that started with \h* or \R* might be incorrectly matched.
1081 Version 8.30 04-February-2012
1082 -----------------------------
1084 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
1085 name is defined in ctype.h.
1087 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
1088 only in quite long subpatterns.
1090 3. Removed the function pcre_info(), which has been obsolete and deprecated
1091 since it was replaced by pcre_fullinfo() in February 2000.
1093 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
1094 match a (*MARK), and the match failed at the start of the subject, a
1095 reference to memory before the start of the subject could occur. This bug
1096 was introduced by fix 17 of release 8.21.
1098 5. A reference to an unset group with zero minimum repetition was giving
1099 totally wrong answers (in non-JavaScript-compatibility mode). For example,
1100 /(another)?(\1?)test/ matched against "hello world test". This bug was
1101 introduced in release 8.13.
1103 6. Add support for 16-bit character strings (a large amount of work involving
1104 many changes and refactorings).
1106 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1107 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1108 from a file.
1110 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1111 rounding is not applied in this particular case).
1113 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1114 if they appear, or are escaped, in patterns.
1116 10. Get rid of a number of -Wunused-but-set-variable warnings.
1118 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1119 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1120 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1121 also returns the mark "x". This bug applied to capturing parentheses,
1122 non-capturing parentheses, and atomic parentheses. It also applied to some
1123 assertions.
1125 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1126 information out of configure.ac instead of relying on pcre.h.generic, which
1127 is not stored in the repository.
1129 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1130 -lreadline.
1132 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1134 15. Applied Graycode's patch to put the top-level frame on the stack rather
1135 than the heap when not using the stack for recursion. This gives a
1136 performance improvement in many cases when recursion is not deep.
1138 16. Experimental code added to "pcretest -C" to output the stack frame size.
1141 Version 8.21 12-Dec-2011
1142 ------------------------
1144 1. Updating the JIT compiler.
1146 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1147 are added as well.
1149 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1150 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1151 calling _pcre_jit_exec. Some extra comments are added.
1153 4. (*MARK) settings inside atomic groups that do not contain any capturing
1154 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1155 was introduced by change 18 for 8.20.
1157 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1158 ECMA-262 standard.
1160 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1161 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1162 This bug was probably introduced by change 9 of 8.13.
1164 7. While fixing 6 above, I noticed that a number of other items were being
1165 incorrectly rejected as "not fixed length". This arose partly because newer
1166 opcodes had not been added to the fixed-length checking code. I have (a)
1167 corrected the bug and added tests for these items, and (b) arranged for an
1168 error to occur if an unknown opcode is encountered while checking for fixed
1169 length instead of just assuming "not fixed length". The items that were
1170 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1171 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1172 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1174 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1175 being incorrectly compiled and would have given unpredicatble results.
1177 9. A possessively repeated subpattern with minimum repeat count greater than
1178 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1179 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1180 the first (A) could occur when it should not.
1182 10. Add a cast and remove a redundant test from the code.
1184 11. JIT should use pcre_malloc/pcre_free for allocation.
1186 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1187 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1188 is anything other than /usr, -L is still shown).
1190 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1192 14. Perl does not support \N without a following name in a [] class; PCRE now
1193 also gives an error.
1195 15. If a forward reference was repeated with an upper limit of around 2000,
1196 it caused the error "internal error: overran compiling workspace". The
1197 maximum number of forward references (including repeats) was limited by the
1198 internal workspace, and dependent on the LINK_SIZE. The code has been
1199 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1200 the default depends on LINK_SIZE. There is a new upper limit (for safety)
1201 of around 200,000 forward references. While doing this, I also speeded up
1202 the filling in of repeated forward references.
1204 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1205 incorrectly expecting the subject to contain another "a" after the start.
1207 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1208 in the match, the SKIP should be ignored. This was not happening; instead
1209 the SKIP was being treated as NOMATCH. For patterns such as
1210 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1211 tested.
1213 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1214 now much more compatible with Perl, in particular in cases where the result
1215 is a non-match for a non-anchored pattern. For example, if
1216 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1217 "m", where previously it did not return a name. A side effect of this
1218 change is that for partial matches, the last encountered mark name is
1219 returned, as for non matches. A number of tests that were previously not
1220 Perl-compatible have been moved into the Perl-compatible test files. The
1221 refactoring has had the pleasing side effect of removing one argument from
1222 the match() function, thus reducing its stack requirements.
1224 19. If the /S+ option was used in pcretest to study a pattern using JIT,
1225 subsequent uses of /S (without +) incorrectly behaved like /S+.
1227 21. Retrieve executable code size support for the JIT compiler and fixing
1228 some warnings.
1230 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1231 not work when the shorter character appeared right at the end of the
1232 subject string.
1234 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1235 systems.
1237 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1238 output it when the /M option is used in pcretest.
1240 25. The CheckMan script was not being included in the distribution. Also, added
1241 an explicit "perl" to run Perl scripts from the PrepareRelease script
1242 because this is reportedly needed in Windows.
1244 26. If study data was being save in a file and studying had not found a set of
1245 "starts with" bytes for the pattern, the data written to the file (though
1246 never used) was taken from uninitialized memory and so caused valgrind to
1247 complain.
1249 27. Updated RunTest.bat as provided by Sheri Pierce.
1251 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1253 29. Computation of memory usage for the table of capturing group names was
1254 giving an unnecessarily large value.
1257 Version 8.20 21-Oct-2011
1258 ------------------------
1260 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1261 a POSIX class. After further experiments with Perl, which convinced me that
1262 Perl has bugs and confusions, a closing square bracket is no longer allowed
1263 in a POSIX name. This bug also affected patterns with classes that started
1264 with full stops.
1266 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1267 captured substring, but while checking the failing first alternative,
1268 substring 1 is temporarily captured. If the output vector supplied to
1269 pcre_exec() was not big enough for this capture, the yield of the function
1270 was still zero ("insufficient space for captured substrings"). This cannot
1271 be totally fixed without adding another stack variable, which seems a lot
1272 of expense for a edge case. However, I have improved the situation in cases
1273 such as /(a)(b)x|abc/ matched against "abc", where the return code
1274 indicates that fewer than the maximum number of slots in the ovector have
1275 been set.
1277 3. Related to (2) above: when there are more back references in a pattern than
1278 slots in the output vector, pcre_exec() uses temporary memory during
1279 matching, and copies in the captures as far as possible afterwards. It was
1280 using the entire output vector, but this conflicts with the specification
1281 that only 2/3 is used for passing back captured substrings. Now it uses
1282 only the first 2/3, for compatibility. This is, of course, another edge
1283 case.
1285 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
1286 main code base, and can be used by building with --enable-jit. When this is
1287 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1288 runtime --no-jit option is given.
1290 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
1291 ovector, the return from the function was zero, implying that there were
1292 other matches that did not fit. The correct "exactly full" value is now
1293 returned.
1295 6. If a subpattern that was called recursively or as a subroutine contained
1296 (*PRUNE) or any other control that caused it to give a non-standard return,
1297 invalid errors such as "Error -26 (nested recursion at the same subject
1298 position)" or even infinite loops could occur.
1300 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1301 computing the minimum length on reaching *ACCEPT, and so ended up with the
1302 wrong value of 1 rather than 0. Further investigation indicates that
1303 computing a minimum subject length in the presence of *ACCEPT is difficult
1304 (think back references, subroutine calls), and so I have changed the code
1305 so that no minimum is registered for a pattern that contains *ACCEPT.
1307 8. If (*THEN) was present in the first (true) branch of a conditional group,
1308 it was not handled as intended. [But see 16 below.]
1310 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1311 Sheri Pierce.
1313 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1314 the first byte in a match must be "a".
1316 11. Change 17 for 8.13 increased the recursion depth for patterns like
1317 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1318 pattern contains any instances of (*THEN). If it does not, the old
1319 optimizations are restored. It would be nice to do this on a per-group
1320 basis, but at the moment that is not feasible.
1322 12. In some environments, the output of pcretest -C is CRLF terminated. This
1323 broke RunTest's code that checks for the link size. A single white space
1324 character after the value is now allowed for.
1326 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1327 For "fr", it uses the Windows-specific input and output files.
1329 14. If (*THEN) appeared in a group that was called recursively or as a
1330 subroutine, it did not work as intended. [But see next item.]
1332 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1333 pattern fragments (but not containing any | characters). If A and B are
1334 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1335 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1336 D. In other words, Perl considers parentheses that do not contain any |
1337 characters to be part of a surrounding alternative, whereas PCRE was
1338 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1339 differently. PCRE now behaves in the same way as Perl, except in the case
1340 of subroutine/recursion calls such as (?1) which have in any case always
1341 been different (but PCRE had them first :-).
1343 16. Related to 15 above: Perl does not treat the | in a conditional group as
1344 creating alternatives. Such a group is treated in the same way as an
1345 ordinary group without any | characters when processing (*THEN). PCRE has
1346 been changed to match Perl's behaviour.
1348 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1349 RunGrepTest script failed.
1351 18. Change 22 for version 13 caused atomic groups to use more stack. This is
1352 inevitable for groups that contain captures, but it can lead to a lot of
1353 stack use in large patterns. The old behaviour has been restored for atomic
1354 groups that do not contain any capturing parentheses.
1356 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1357 suppress the check for a minimum subject length at run time. (If it was
1358 given to pcre_exec() or pcre_dfa_exec() it did work.)
1360 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1361 fail to work when decoding hex characters in data strings in EBCDIC
1362 environments.
1364 21. It appears that in at least one Mac OS environment, the isxdigit() function
1365 is implemented as a macro that evaluates to its argument more than once,
1366 contravening the C 90 Standard (I haven't checked a later standard). There
1367 was an instance in pcretest which caused it to go wrong when processing
1368 \x{...} escapes in subject strings. The has been rewritten to avoid using
1369 things like p++ in the argument of isxdigit().
1372 Version 8.13 16-Aug-2011
1373 ------------------------
1375 1. The Unicode data tables have been updated to Unicode 6.0.0.
1377 2. Two minor typos in pcre_internal.h have been fixed.
1379 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1380 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1381 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1383 4. There were a number of related bugs in the code for matching backrefences
1384 caselessly in UTF-8 mode when codes for the characters concerned were
1385 different numbers of bytes. For example, U+023A and U+2C65 are an upper
1386 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1387 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1388 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1389 2-byte code at the end of the subject (it thought there wasn't enough data
1390 left).
1392 5. Comprehensive information about what went wrong is now returned by
1393 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1394 as the output vector has at least 2 elements. The offset of the start of
1395 the failing character and a reason code are placed in the vector.
1397 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
1398 now returned is for the first byte of the failing character, instead of the
1399 last byte inspected. This is an incompatible change, but I hope it is small
1400 enough not to be a problem. It makes the returned offset consistent with
1401 pcre_exec() and pcre_dfa_exec().
1403 7. pcretest now gives a text phrase as well as the error number when
1404 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1405 failure, the offset and reason code are output.
1407 8. When \R was used with a maximizing quantifier it failed to skip backwards
1408 over a \r\n pair if the subsequent match failed. Instead, it just skipped
1409 back over a single character (\n). This seems wrong (because it treated the
1410 two characters as a single entity when going forwards), conflicts with the
1411 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1412 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1413 has been changed.
1415 9. Some internal refactoring has changed the processing so that the handling
1416 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1417 time (the PCRE_DOTALL option was changed this way some time ago: version
1418 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1419 which was always a bit of a fudge. It also means that there is one less
1420 argument for the match() function, which reduces its stack requirements
1421 slightly. This change also fixes an incompatibility with Perl: the pattern
1422 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1424 10. More internal refactoring has drastically reduced the number of recursive
1425 calls to match() for possessively repeated groups such as (abc)++ when
1426 using pcre_exec().
1428 11. While implementing 10, a number of bugs in the handling of groups were
1429 discovered and fixed:
1431 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1432 (a|)*(?1) gave a compile-time internal error.
1433 ((a|)+)+ did not notice that the outer group could match an empty string.
1434 (^a|^)+ was not marked as anchored.
1435 (.*a|.*)+ was not marked as matching at start or after a newline.
1437 12. Yet more internal refactoring has removed another argument from the match()
1438 function. Special calls to this function are now indicated by setting a
1439 value in a variable in the "match data" data block.
1441 13. Be more explicit in pcre_study() instead of relying on "default" for
1442 opcodes that mean there is no starting character; this means that when new
1443 ones are added and accidentally left out of pcre_study(), testing should
1444 pick them up.
1446 14. The -s option of pcretest has been documented for ages as being an old
1447 synonym of -m (show memory usage). I have changed it to mean "force study
1448 for every regex", that is, assume /S for every regex. This is similar to -i
1449 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1450 using it. It makes it easier to run collections of tests with and without
1451 study enabled, and thereby test pcre_study() more easily. All the standard
1452 tests are now run with and without -s (but some patterns can be marked as
1453 "never study" - see 20 below).
1455 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1456 restoration of the capturing data to the outer values was not happening
1457 correctly.
1459 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1460 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1461 pattern had matched an empty string, and so incorrectly returned a no
1462 match.
1464 17. There was optimizing code for the last branch of non-capturing parentheses,
1465 and also for the obeyed branch of a conditional subexpression, which used
1466 tail recursion to cut down on stack usage. Unfortunately, now that there is
1467 the possibility of (*THEN) occurring in these branches, tail recursion is
1468 no longer possible because the return has to be checked for (*THEN). These
1469 two optimizations have therefore been removed. [But see 8.20/11 above.]
1471 18. If a pattern containing \R was studied, it was assumed that \R always
1472 matched two bytes, thus causing the minimum subject length to be
1473 incorrectly computed because \R can also match just one byte.
1475 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1476 was incorrectly computed.
1478 20. If /S is present twice on a test pattern in pcretest input, it now
1479 *disables* studying, thereby overriding the use of -s on the command line
1480 (see 14 above). This is necessary for one or two tests to keep the output
1481 identical in both cases.
1483 21. When (*ACCEPT) was used in an assertion that matched an empty string and
1484 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1486 22. When an atomic group that contained a capturing parenthesis was
1487 successfully matched, but the branch in which it appeared failed, the
1488 capturing was not being forgotten if a higher numbered group was later
1489 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1490 group 1 to "a", when in fact it should be unset. This applied to multi-
1491 branched capturing and non-capturing groups, repeated or not, and also to
1492 positive assertions (capturing in negative assertions does not happen
1493 in PCRE) and also to nested atomic groups.
1495 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1496 subject after a captured substring, to make it easier to tell which of a
1497 number of identical substrings has been captured.
1499 24. The way atomic groups are processed by pcre_exec() has been changed so that
1500 if they are repeated, backtracking one repetition now resets captured
1501 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1502 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1503 "aaa". Previously, it would have been "a". As part of this code
1504 refactoring, the way recursive calls are handled has also been changed.
1506 25. If an assertion condition captured any substrings, they were not passed
1507 back unless some other capturing happened later. For example, if
1508 (?(?=(a))a) was matched against "a", no capturing was returned.
1510 26. When studying a pattern that contained subroutine calls or assertions,
1511 the code for finding the minimum length of a possible match was handling
1512 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1513 group 1 called group 2 while simultaneously a separate group 2 called group
1514 1). A stack overflow occurred in this case. I have fixed this by limiting
1515 the recursion depth to 10.
1517 27. Updated RunTest.bat in the distribution to the version supplied by Tom
1518 Fortmann. This supports explicit test numbers on the command line, and has
1519 argument validation and error reporting.
1521 28. An instance of \X with an unlimited repeat could fail if at any point the
1522 first character it looked at was a mark character.
1524 29. Some minor code refactoring concerning Unicode properties and scripts
1525 should reduce the stack requirement of match() slightly.
1527 30. Added the '=' option to pcretest to check the setting of unused capturing
1528 slots at the end of the pattern, which are documented as being -1, but are
1529 not included in the return count.
1531 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1532 compiled something random. Now it gives a compile-time error (as does
1533 Perl).
1535 32. A *MARK encountered during the processing of a positive assertion is now
1536 recorded and passed back (compatible with Perl).
1538 33. If --only-matching or --colour was set on a pcregrep call whose pattern
1539 had alternative anchored branches, the search for a second match in a line
1540 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1541 matched the line "0102" twice. The same bug affected patterns that started
1542 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1543 twice.
1545 34. Previously, PCRE did not allow quantification of assertions. However, Perl
1546 does, and because of capturing effects, quantifying parenthesized
1547 assertions may at times be useful. Quantifiers are now allowed for
1548 parenthesized assertions.
1550 35. A minor code tidy in pcre_compile() when checking options for \R usage.
1552 36. \g was being checked for fancy things in a character class, when it should
1553 just be a literal "g".
1555 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1556 appearance of a nested POSIX class supersedes an apparent external class.
1557 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1558 unescaped square brackets may also appear as part of class names. For
1559 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1560 more like Perl. (But see 8.20/1 above.)
1562 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1563 was because it thought it was \N{name}, which is not supported).
1565 39. Add minix to OS list not supporting the -S option in pcretest.
1567 40. PCRE tries to detect cases of infinite recursion at compile time, but it
1568 cannot analyze patterns in sufficient detail to catch mutual recursions
1569 such as ((?1))((?2)). There is now a runtime test that gives an error if a
1570 subgroup is called recursively as a subpattern for a second time at the
1571 same position in the subject string. In previous releases this might have
1572 been caught by the recursion limit, or it might have run out of stack.
1574 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1575 happen only once. PCRE was, however incorrectly giving a compile time error
1576 "recursive call could loop indefinitely" because it cannot analyze the
1577 pattern in sufficient detail. The compile time test no longer happens when
1578 PCRE is compiling a conditional subpattern, but actual runaway loops are
1579 now caught at runtime (see 40 above).
1581 42. It seems that Perl allows any characters other than a closing parenthesis
1582 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1583 has been changed to be the same.
1585 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1586 as not to get warnings when autogen.sh is called. Also changed
1587 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1589 44. To help people who use pcregrep to scan files containing exceedingly long
1590 lines, the following changes have been made:
1592 (a) The default value of the buffer size parameter has been increased from
1593 8K to 20K. (The actual buffer used is three times this size.)
1595 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1596 PCRE is built.
1598 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1599 to be set at run time.
1601 (d) Numerical values in pcregrep options can be followed by K or M, for
1602 example --buffer-size=50K.
1604 (e) If a line being scanned overflows pcregrep's buffer, an error is now
1605 given and the return code is set to 2.
1607 45. Add a pointer to the latest mark to the callout data block.
1609 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1610 partial match of an empty string instead of no match. This was specific to
1611 the use of ".".
1613 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1614 complete match instead of a partial match. This bug was dependent on both
1615 the PCRE_UTF8 and PCRE_DOTALL options being set.
1617 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1618 starting byte set, because \b was not being ignored.
1621 Version 8.12 15-Jan-2011
1622 ------------------------
1624 1. Fixed some typos in the markup of the man pages, and wrote a script that
1625 checks for such things as part of the documentation building process.
1627 2. On a big-endian 64-bit system, pcregrep did not correctly process the
1628 --match-limit and --recursion-limit options (added for 8.11). In
1629 particular, this made one of the standard tests fail. (The integer value
1630 went into the wrong half of a long int.)
1632 3. If the --colour option was given to pcregrep with -v (invert match), it
1633 did strange things, either producing crazy output, or crashing. It should,
1634 of course, ignore a request for colour when reporting lines that do not
1635 match.
1637 4. Another pcregrep bug caused similar problems if --colour was specified with
1638 -M (multiline) and the pattern match finished with a line ending.
1640 5. In pcregrep, when a pattern that ended with a literal newline sequence was
1641 matched in multiline mode, the following line was shown as part of the
1642 match. This seems wrong, so I have changed it.
1644 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
1645 the check for further matches in the same line (so they could be coloured)
1646 to overrun the end of the current line. If another match was found, it was
1647 incorrectly shown (and then shown again when found in the next line).
1649 7. If pcregrep was compiled under Windows, there was a reference to the
1650 function pcregrep_exit() before it was defined. I am assuming this was
1651 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1652 reported by a user. I've moved the definition above the reference.
1655 Version 8.11 10-Dec-2010
1656 ------------------------
1658 1. (*THEN) was not working properly if there were untried alternatives prior
1659 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1660 backtracked to try for "b" instead of moving to the next alternative branch
1661 at the same level (in this case, to look for "c"). The Perl documentation
1662 is clear that when (*THEN) is backtracked onto, it goes to the "next
1663 alternative in the innermost enclosing group".
1665 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1666 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1667 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1668 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1669 (*THEN).
1671 3. If \s appeared in a character class, it removed the VT character from
1672 the class, even if it had been included by some previous item, for example
1673 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1674 of \s, but is part of the POSIX "space" class.)
1676 4. A partial match never returns an empty string (because you can always
1677 match an empty string at the end of the subject); however the checking for
1678 an empty string was starting at the "start of match" point. This has been
1679 changed to the "earliest inspected character" point, because the returned
1680 data for a partial match starts at this character. This means that, for
1681 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1682 (previously it gave "no match").
1684 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1685 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1686 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1687 has an implication that the given string is incomplete (because a partial
1688 match is preferred over a full match). For this reason, these items now
1689 give a partial match in this situation. [Aside: previously, the one case
1690 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1691 match rather than a full match, which was wrong by the old rules, but is
1692 now correct.]
1694 6. There was a bug in the handling of #-introduced comments, recognized when
1695 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1696 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1697 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1698 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1699 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1700 places in pcre_compile().
1702 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1703 comments when looking ahead for named forward references to subpatterns,
1704 the only newline sequence it recognized was NL. It now handles newlines
1705 according to the set newline convention.
1707 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1708 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1709 cater for a lack of strerror(). These oversights have been fixed.
1711 9. Added --match-limit and --recursion-limit to pcregrep.
1713 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1715 11. When the -o option was used, pcregrep was setting a return code of 1, even
1716 when matches were found, and --line-buffered was not being honoured.
1718 12. Added an optional parentheses number to the -o and --only-matching options
1719 of pcregrep.
1721 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1722 can match an empty string. The code to do it in pcretest and pcredemo
1723 needed fixing:
1725 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1726 only one byte after an empty string match just before CRLF (this case
1727 just got forgotten; "any" and "anycrlf" were OK).
1729 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1730 mode when an empty string match preceded an ASCII character followed by
1731 a non-ASCII character. (The code for advancing by one character rather
1732 than one byte was nonsense.)
1734 (c) The pcredemo.c sample program did not have any code at all to handle
1735 the cases when CRLF is a valid newline sequence.
1737 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1738 as a starting offset was within the subject string. There is now a new
1739 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1740 negative or greater than the length of the string. In order to test this,
1741 pcretest is extended to allow the setting of negative starting offsets.
1743 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1744 starting offset points to the beginning of a UTF-8 character was
1745 unnecessarily clumsy. I tidied it up.
1747 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1748 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1750 17. Nobody had reported that the --include_dir option, which was added in
1751 release 7.7 should have been called --include-dir (hyphen, not underscore)
1752 for compatibility with GNU grep. I have changed it to --include-dir, but
1753 left --include_dir as an undocumented synonym, and the same for
1754 --exclude-dir, though that is not available in GNU grep, at least as of
1755 release 2.5.4.
1757 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1758 characters from a string of bytes) have been redefined so as not to use
1759 loops, in order to improve performance in some environments. At the same
1760 time, I abstracted some of the common code into auxiliary macros to save
1761 repetition (this should not affect the compiled code).
1763 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1764 compile-time error is now given if \c is not followed by an ASCII
1765 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1766 different, and any byte value is allowed.)
1768 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1769 START_OPTIMIZE option, which is now allowed at compile time - but just
1770 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1771 to pcregrep and other applications that have no direct access to PCRE
1772 options. The new /Y option in pcretest sets this option when calling
1773 pcre_compile().
1775 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1776 back references. Groups containing recursive back references were forced to
1777 be atomic by that change, but in the case of named groups, the amount of
1778 memory required was incorrectly computed, leading to "Failed: internal
1779 error: code overflow". This has been fixed.
1781 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1782 pcretest.c, to avoid build problems in some Borland environments.
1785 Version 8.10 25-Jun-2010
1786 ------------------------
1788 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1789 THEN.
1791 2. (*ACCEPT) was not working when inside an atomic group.
1793 3. Inside a character class, \B is treated as a literal by default, but
1794 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1795 causes the error). The code is unchanged, but I tidied the documentation.
1797 4. Inside a character class, PCRE always treated \R and \X as literals,
1798 whereas Perl faults them if its -w option is set. I have changed PCRE so
1799 that it faults them when PCRE_EXTRA is set.
1801 5. Added support for \N, which always matches any character other than
1802 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1804 6. When compiling pcregrep with newer versions of gcc which may have
1805 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1806 declared with attribute warn_unused_result" were given. Just casting the
1807 result to (void) does not stop the warnings; a more elaborate fudge is
1808 needed. I've used a macro to implement this.
1810 7. Minor change to pcretest.c to avoid a compiler warning.
1812 8. Added four artifical Unicode properties to help with an option to make
1813 \s etc use properties (see next item). The new properties are: Xan
1814 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1816 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1817 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1818 this option. Modified pcretest to add /W to test this facility. Added
1819 REG_UCP to make it available via the POSIX interface.
1821 10. Added --line-buffered to pcregrep.
1823 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1824 studied, and the match started with a letter with a code point greater than
1825 127 whose first byte was different to the first byte of the other case of
1826 the letter, the other case of this starting letter was not recognized
1827 (#976).
1829 12. If a pattern that was studied started with a repeated Unicode property
1830 test, for example, \p{Nd}+, there was the theoretical possibility of
1831 setting up an incorrect bitmap of starting bytes, but fortunately it could
1832 not have actually happened in practice until change 8 above was made (it
1833 added property types that matched character-matching opcodes).
1835 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1836 possible starting bytes for non-anchored patterns.
1838 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1839 \R, and also a number of cases that involve Unicode properties, both
1840 explicit and implicit when PCRE_UCP is set.
1842 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1843 input, it could crash or give wrong results if characters with values
1844 greater than 0xc0 were present in the subject string. (Detail: it assumed
1845 UTF-8 input when processing these items.)
1847 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1848 size_t is 64-bit (#991).
1850 17. Added a check for running out of memory when PCRE is compiled with
1851 --disable-stack-for-recursion (#990).
1853 18. If the last data line in a file for pcretest does not have a newline on
1854 the end, a newline was missing in the output.
1856 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1857 less than 128) in its various bitmaps. However, there is a facility for
1858 generating tables according to the current locale when PCRE is compiled. It
1859 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1860 characters, are recognized by isspace() and therefore were getting set in
1861 these tables, and indeed these tables seem to approximate to ISO 8859. This
1862 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1863 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1864 which of course cannot start UTF-8 characters. I have changed the code so
1865 that only real ASCII characters (less than 128) and the correct starting
1866 bytes for UTF-8 encodings are set for characters greater than 127 when in
1867 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1868 altogether.)
1870 20. Added the /T option to pcretest so as to be able to run tests with non-
1871 standard character tables, thus making it possible to include the tests
1872 used for 19 above in the standard set of tests.
1874 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1875 reference to a subpattern the other side of a comment that contains an
1876 opening parenthesis caused either an internal compiling error, or a
1877 reference to the wrong subpattern.
1880 Version 8.02 19-Mar-2010
1881 ------------------------
1883 1. The Unicode data tables have been updated to Unicode 5.2.0.
1885 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1886 configured.
1888 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1889 original author of that file, following a query about its status.
1891 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1892 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1894 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1895 quantifier applied to a forward-referencing subroutine call, could compile
1896 incorrect code or give the error "internal error: previously-checked
1897 referenced subpattern not found".
1899 6. Both MS Visual Studio and Symbian OS have problems with initializing
1900 variables to point to external functions. For these systems, therefore,
1901 pcre_malloc etc. are now initialized to local functions that call the
1902 relevant global functions.
1904 7. There were two entries missing in the vectors called coptable and poptable
1905 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1906 I've fixed the data, and added a kludgy way of testing at compile time that
1907 the lengths are correct (equal to the number of opcodes).
1909 8. Following on from 7, I added a similar kludge to check the length of the
1910 eint vector in pcreposix.c.
1912 9. Error texts for pcre_compile() are held as one long string to avoid too
1913 much relocation at load time. To find a text, the string is searched,
1914 counting zeros. There was no check for running off the end of the string,
1915 which could happen if a new error number was added without updating the
1916 string.
1918 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1920 11. \K was not working if it appeared in an atomic group or in a group that
1921 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1922 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1923 the assertion is positive, but not if it is negative.
1925 12. Change 11 fortuitously reduced the size of the stack frame used in the
1926 "match()" function of pcre_exec.c by one pointer. Forthcoming
1927 implementation of support for (*MARK) will need an extra pointer on the
1928 stack; I have reserved it now, so that the stack frame size does not
1929 decrease.
1931 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1932 item in branch that calls a recursion is a subroutine call - as in the
1933 second branch in the above example - was incorrectly given the compile-
1934 time error "recursive call could loop indefinitely" because pcre_compile()
1935 was not correctly checking the subroutine for matching a non-empty string.
1937 14. The checks for overrunning compiling workspace could trigger after an
1938 overrun had occurred. This is a "should never occur" error, but it can be
1939 triggered by pathological patterns such as hundreds of nested parentheses.
1940 The checks now trigger 100 bytes before the end of the workspace.
1942 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1945 Version 8.01 19-Jan-2010
1946 ------------------------
1948 1. If a pattern contained a conditional subpattern with only one branch (in
1949 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1950 computed the wrong minimum data length (which is of course zero for such
1951 subpatterns). This could cause incorrect "no match" results.
1953 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1954 the pattern is reset in the first branch, pcre_compile() failed with
1955 "internal error: code overflow at offset...". This happened only when
1956 the reset was to the original external option setting. (An optimization
1957 abstracts leading options settings into an external setting, which was the
1958 cause of this.)
1960 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1961 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1962 assertion pattern did not match (meaning that the assertion was true), it
1963 was incorrectly treated as false if the SKIP had been reached during the
1964 matching. This also applied to assertions used as conditions.
1966 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1967 assertion subpattern, including such a pattern used as a condition,
1968 unpredictable results occurred, instead of the error return
1971 5. The C++ GlobalReplace function was not working like Perl for the special
1972 situation when an empty string is matched. It now does the fancy magic
1973 stuff that is necessary.
1975 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1976 removed. (These were left over from very, very early versions of PCRE.)
1978 7. Some cosmetic changes to the code to make life easier when compiling it
1979 as part of something else:
1981 (a) Change DEBUG to PCRE_DEBUG.
1983 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1984 called "current" as "current_branch", to prevent a collision with the
1985 Linux macro when compiled as a kernel module.
1987 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1988 prevent a collision with the Linux macro when compiled as a kernel
1989 module.
1991 8. In pcre_compile() there are some checks for integer overflows that used to
1992 cast potentially large values to (double). This has been changed to that
1993 when building, a check for int64_t is made, and if it is found, it is used
1994 instead, thus avoiding the use of floating point arithmetic. (There is no
1995 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1996 double.
1998 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1999 2005 (difference between two addresses compared to an unsigned value).
2001 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
2002 custom one, because of the following reported problem in Windows:
2004 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
2005 under Win32.
2006 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
2007 therefore missing the function definition.
2008 - The compiler thus generates a "C" signature for the test function.
2009 - The linker fails to find the "C" function.
2010 - PCRE fails to configure if asked to do so against libbz2.
2012 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
2013 messages were output:
2015 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
2016 rerunning libtoolize, to keep the correct libtool macros in-tree.
2017 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2019 I have done both of these things.
2021 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2022 most of the time, it *can* run out if it is given a pattern that contains a
2023 runaway infinite recursion. I updated the discussion in the pcrestack man
2024 page.
2026 13. Now that we have gone to the x.xx style of version numbers, the minor
2027 version may start with zero. Using 08 or 09 is a bad idea because users
2028 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2029 interpreted as invalid octal numbers. I've updated the previous comment in
2030 configure.ac, and also added a check that gives an error if 08 or 09 are
2031 used.
2033 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2034 causing partial matching to fail when the end of the subject matched \W
2035 in a UTF-8 pattern where \W was quantified with a minimum of 3.
2037 15. There were some discrepancies between the declarations in pcre_internal.h
2038 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2039 their definitions. The declarations used "const uschar *" and the
2040 definitions used USPTR. Even though USPTR is normally defined as "const
2041 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2042 reported that: "This difference in casting confuses some C++ compilers, for
2043 example, SunCC recognizes above declarations as different functions and
2044 generates broken code for hbpcre." I have changed the declarations to use
2045 USPTR.
2047 16. GNU libtool is named differently on some systems. The autogen.sh script now
2048 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2049 (FreeBSD).
2051 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2052 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2053 comment: "Figure out how to create a longlong from a string: strtoll and
2054 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2055 instance, but it only takes 2 args instead of 3!"
2057 18. A subtle bug concerned with back references has been fixed by a change of
2058 specification, with a corresponding code fix. A pattern such as
2059 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2060 refers, was giving matches when it shouldn't. For example, xa=xaaa would
2061 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2062 same bug. Such groups have to be quantified to be useful, or contained
2063 inside another quantified group. (If there's no repetition, the reference
2064 can never match.) The problem arises because, having left the group and
2065 moved on to the rest of the pattern, a later failure that backtracks into
2066 the group uses the captured value from the final iteration of the group
2067 rather than the correct earlier one. I have fixed this in PCRE by forcing
2068 any group that contains a reference to itself to be an atomic group; that
2069 is, there cannot be any backtracking into it once it has completed. This is
2070 similar to recursive and subroutine calls.
2073 Version 8.00 19-Oct-09
2074 ----------------------
2076 1. The table for translating pcre_compile() error codes into POSIX error codes
2077 was out-of-date, and there was no check on the pcre_compile() error code
2078 being within the table. This could lead to an OK return being given in
2079 error.
2081 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
2082 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2083 in a Windows environment.
2085 3. The pcregrep --count option prints the count for each file even when it is
2086 zero, as does GNU grep. However, pcregrep was also printing all files when
2087 --files-with-matches was added. Now, when both options are given, it prints
2088 counts only for those files that have at least one match. (GNU grep just
2089 prints the file name in this circumstance, but including the count seems
2090 more useful - otherwise, why use --count?) Also ensured that the
2091 combination -clh just lists non-zero counts, with no names.
2093 4. The long form of the pcregrep -F option was incorrectly implemented as
2094 --fixed_strings instead of --fixed-strings. This is an incompatible change,
2095 but it seems right to fix it, and I didn't think it was worth preserving
2096 the old behaviour.
2098 5. The command line items --regex=pattern and --regexp=pattern were not
2099 recognized by pcregrep, which required --regex pattern or --regexp pattern
2100 (with a space rather than an '='). The man page documented the '=' forms,
2101 which are compatible with GNU grep; these now work.
2103 6. No libpcreposix.pc file was created for pkg-config; there was just
2104 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2106 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2107 when UCP support is not needed, by modifying the Python script that
2108 generates it from Unicode data files. This should not matter if the module
2109 is correctly used as a library, but I received one complaint about 50K of
2110 unwanted data. My guess is that the person linked everything into his
2111 program rather than using a library. Anyway, it does no harm.
2113 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2114 was a minimum greater than 1 for a wide character in a possessive
2115 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2116 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2117 character. Chaos in the form of incorrect output or a compiling loop could
2118 result.
2120 9. The restrictions on what a pattern can contain when partial matching is
2121 requested for pcre_exec() have been removed. All patterns can now be
2122 partially matched by this function. In addition, if there are at least two
2123 slots in the offset vector, the offset of the earliest inspected character
2124 for the match and the offset of the end of the subject are set in them when
2125 PCRE_ERROR_PARTIAL is returned.
2127 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2128 synonymous with PCRE_PARTIAL, for backwards compatibility, and
2129 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2130 and may be more useful for multi-segment matching.
2132 11. Partial matching with pcre_exec() is now more intuitive. A partial match
2133 used to be given if ever the end of the subject was reached; now it is
2134 given only if matching could not proceed because another character was
2135 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2136 string "Z", which now yields "no match" instead of "partial match". In the
2137 case of pcre_dfa_exec(), "no match" is given if every matching path for the
2138 final character ended with (*FAIL).
2140 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2141 if the pattern had a "must contain" character that was already found in the
2142 earlier partial match, unless partial matching was again requested. For
2143 example, with the pattern /dog.(body)?/, the "must contain" character is
2144 "g". If the first part-match was for the string "dog", restarting with
2145 "sbody" failed. This bug has been fixed.
2147 13. The string returned by pcre_dfa_exec() after a partial match has been
2148 changed so that it starts at the first inspected character rather than the
2149 first character of the match. This makes a difference only if the pattern
2150 starts with a lookbehind assertion or \b or \B (\K is not supported by
2151 pcre_dfa_exec()). It's an incompatible change, but it makes the two
2152 matching functions compatible, and I think it's the right thing to do.
2154 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2155 so that the demonstration program is easily available in environments where
2156 PCRE has not been installed from source.
2158 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2159 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2160 library.
2162 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2163 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2164 is not the first non-POSIX option to be added. Clearly some people find
2165 these options useful.
2167 17. If a caller to the POSIX matching function regexec() passes a non-zero
2168 value for nmatch with a NULL value for pmatch, the value of
2169 nmatch is forced to zero.
2171 18. RunGrepTest did not have a test for the availability of the -u option of
2172 the diff command, as RunTest does. It now checks in the same way as
2173 RunTest, and also checks for the -b option.
2175 19. If an odd number of negated classes containing just a single character
2176 interposed, within parentheses, between a forward reference to a named
2177 subpattern and the definition of the subpattern, compilation crashed with
2178 an internal error, complaining that it could not find the referenced
2179 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2180 [The bug was that it was starting one character too far in when skipping
2181 over the character class, thus treating the ] as data rather than
2182 terminating the class. This meant it could skip too much.]
2184 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2185 /g option in pcretest when the pattern contains \K, which makes it possible
2186 to have an empty string match not at the start, even when the pattern is
2187 anchored. Updated pcretest and pcredemo to use this option.
2189 21. If the maximum number of capturing subpatterns in a recursion was greater
2190 than the maximum at the outer level, the higher number was returned, but
2191 with unset values at the outer level. The correct (outer level) value is
2192 now given.
2194 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2195 PCRE did not set those parentheses (unlike Perl). I have now found a way to
2196 make it do so. The string so far is captured, making this feature
2197 compatible with Perl.
2199 23. The tests have been re-organized, adding tests 11 and 12, to make it
2200 possible to check the Perl 5.10 features against Perl 5.10.
2202 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2203 pattern matches a fixed length string. PCRE did not allow this; now it
2204 does. Neither allows recursion.
2206 25. I finally figured out how to implement a request to provide the minimum
2207 length of subject string that was needed in order to match a given pattern.
2208 (It was back references and recursion that I had previously got hung up
2209 on.) This code has now been added to pcre_study(); it finds a lower bound
2210 to the length of subject needed. It is not necessarily the greatest lower
2211 bound, but using it to avoid searching strings that are too short does give
2212 some useful speed-ups. The value is available to calling programs via
2213 pcre_fullinfo().
2215 26. While implementing 25, I discovered to my embarrassment that pcretest had
2216 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2217 study optimizations had never been tested with that matching function.
2218 Oops. What is worse, even when it was passed study data, there was a bug in
2219 pcre_dfa_exec() that meant it never actually used it. Double oops. There
2220 were also very few tests of studied patterns with pcre_dfa_exec().
2222 27. If (?| is used to create subpatterns with duplicate numbers, they are now
2223 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2224 on the other side of the coin, they are no longer allowed to have different
2225 names, because these cannot be distinguished in PCRE, and this has caused
2226 confusion. (This is a difference from Perl.)
2228 28. When duplicate subpattern names are present (necessarily with different
2229 numbers, as required by 27 above), and a test is made by name in a
2230 conditional pattern, either for a subpattern having been matched, or for
2231 recursion in such a pattern, all the associated numbered subpatterns are
2232 tested, and the overall condition is true if the condition is true for any
2233 one of them. This is the way Perl works, and is also more like the way
2234 testing by number works.
2237 Version 7.9 11-Apr-09
2238 ---------------------
2240 1. When building with support for bzlib/zlib (pcregrep) and/or readline
2241 (pcretest), all targets were linked against these libraries. This included
2242 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2243 libraries. This caused unwanted dependencies to be created. This problem
2244 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2245 pcretest is linked with readline.
2247 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
2248 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2249 moved outside it again, because FALSE and TRUE are already defined in AIX,
2250 but BOOL is not.
2252 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2253 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2255 4. The pcregrep documentation said spaces were inserted as well as colons (or
2256 hyphens) following file names and line numbers when outputting matching
2257 lines. This is not true; no spaces are inserted. I have also clarified the
2258 wording for the --colour (or --color) option.
2260 5. In pcregrep, when --colour was used with -o, the list of matching strings
2261 was not coloured; this is different to GNU grep, so I have changed it to be
2262 the same.
2264 6. When --colo(u)r was used in pcregrep, only the first matching substring in
2265 each matching line was coloured. Now it goes on to look for further matches
2266 of any of the test patterns, which is the same behaviour as GNU grep.
2268 7. A pattern that could match an empty string could cause pcregrep to loop; it
2269 doesn't make sense to accept an empty string match in pcregrep, so I have
2270 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2271 seems to be how GNU grep behaves. [But see later change 40 for release
2272 8.33.]
2274 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2275 start or after a newline", because the conditional assertion was not being
2276 correctly handled. The rule now is that both the assertion and what follows
2277 in the first alternative must satisfy the test.
2279 9. If auto-callout was enabled in a pattern with a conditional group whose
2280 condition was an assertion, PCRE could crash during matching, both with
2281 pcre_exec() and pcre_dfa_exec().
2283 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2284 used for matching.
2286 11. Unicode property support in character classes was not working for
2287 characters (bytes) greater than 127 when not in UTF-8 mode.
2289 12. Added the -M command line option to pcretest.
2291 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2293 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2295 16. Added comments and documentation about mis-use of no_arg in the C++
2296 wrapper.
2298 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2299 from Martin Jerabek that uses macro names for all relevant character and
2300 string constants.
2302 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2303 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2304 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2305 these, but not everybody uses configure.
2307 19. A conditional group that had only one branch was not being correctly
2308 recognized as an item that could match an empty string. This meant that an
2309 enclosing group might also not be so recognized, causing infinite looping
2310 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2311 with the subject "ab", where knowledge that the repeated group can match
2312 nothing is needed in order to break the loop.
2314 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2315 exec(), but without supplying a callout function, matching went wrong.
2317 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2318 leak if the size of the offset vector was greater than 30. When the vector
2319 is smaller, the saved offsets during recursion go onto a local stack
2320 vector, but for larger vectors malloc() is used. It was failing to free
2321 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2322 error, in fact).
2324 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2325 heapframe that is used only when UTF-8 support is enabled. This caused no
2326 problem, but was untidy.
2328 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2329 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2330 included within another project.
2332 24. Steven Van Ingelgem's patches to add more options to the CMake support,
2333 slightly modified by me:
2335 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2336 not building pcregrep.
2338 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2339 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2341 25. Forward references, both numeric and by name, in patterns that made use of
2342 duplicate group numbers, could behave incorrectly or give incorrect errors,
2343 because when scanning forward to find the reference group, PCRE was not
2344 taking into account the duplicate group numbers. A pattern such as
2345 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2347 26. Changed a few more instances of "const unsigned char *" to USPTR, making
2348 the feature of a custom pointer more persuasive (as requested by a user).
2350 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2351 pcretest.c, inside #ifndefs, because it seems they are sometimes already
2352 pre-defined.
2354 28. Added support for (*UTF8) at the start of a pattern.
2356 29. Arrange for flags added by the "release type" setting in CMake to be shown
2357 in the configuration summary.
2360 Version 7.8 05-Sep-08
2361 ---------------------
2363 1. Replaced UCP searching code with optimized version as implemented for Ad
2364 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2365 stage table and inline lookup instead of a function, giving speed ups of 2
2366 to 5 times on some simple patterns that I tested. Permission was given to
2367 distribute the MultiStage2.py script that generates the tables (it's not in
2368 the tarball, but is in the Subversion repository).
2370 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2371 scripts.
2373 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2374 a group with a zero qualifier. The result of the study could be incorrect,
2375 or the function might crash, depending on the pattern.
2377 4. Caseless matching was not working for non-ASCII characters in back
2378 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2379 It now works when Unicode Property Support is available.
2381 5. In pcretest, an escape such as \x{de} in the data was always generating
2382 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2383 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2384 truncation.
2386 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2388 7. Added two (int) casts to pcregrep when printing the difference of two
2389 pointers, in case they are 64-bit values.
2391 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
2392 test 2 if it fails.
2394 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2395 and a #define of that name to empty if it is not externally set. This is to
2396 allow users of MSVC to set it if necessary.
2398 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2399 the convenience functions in the pcre_get.c source file.
2401 11. An option change at the start of a pattern that had top-level alternatives
2402 could cause overwriting and/or a crash. This command provoked a crash in
2403 some environments:
2405 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2407 This potential security problem was recorded as CVE-2008-2371.
2409 12. For a pattern where the match had to start at the beginning or immediately
2410 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2411 pcre_dfa_exec() could read past the end of the passed subject if there was
2412 no match. To help with detecting such bugs (e.g. with valgrind), I modified
2413 pcretest so that it places the subject at the end of its malloc-ed buffer.
2415 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2416 exec() might read past the end of the data buffer in UTF-8 mode.
2418 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2419 the data contained the byte 0x85 as part of a UTF-8 character within its
2420 first line. This applied both to normal and DFA matching.
2422 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2423 /^[^d]*?$/8 failed to match "abc".
2425 16. Added a missing copyright notice to pcrecpp_internal.h.
2427 17. Make it more clear in the documentation that values returned from
2428 pcre_exec() in ovector are byte offsets, not character counts.
2430 18. Tidied a few places to stop certain compilers from issuing warnings.
2432 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2433 supplied by Stefan Weber. I made a further small update for 7.8 because
2434 there is a change of source arrangements: the pcre_searchfuncs.c module is
2435 replaced by pcre_ucd.c.
2438 Version 7.7 07-May-08
2439 ---------------------
2441 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
2442 a string to a long long, pretend we don't even have a long long." This is
2443 done by checking for the strtoq, strtoll, and _strtoi64 functions.
2445 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2446 pre-7.6 versions, which defined a global no_arg variable instead of putting
2447 it in the RE class. (See also #8 below.)
2449 3. Remove a line of dead code, identified by coverity and reported by Nuno
2450 Lopes.
2452 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
2454 (1) The include/exclude patterns were being applied to the whole pathnames
2455 of files, instead of just to the final components.
2457 (2) If there was more than one level of directory, the subdirectories were
2458 skipped unless they satisfied the include/exclude conditions. This is
2459 inconsistent with GNU grep (and could even be seen as contrary to the
2460 pcregrep specification - which I improved to make it absolutely clear).
2461 The action now is always to scan all levels of directory, and just
2462 apply the include/exclude patterns to regular files.
2464 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2465 --exclude_dir in the tests to avoid scanning .svn directories.
2467 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
2468 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2469 doesn't support NULs in patterns.
2471 7. Added some missing "const"s to declarations of static tables in
2472 pcre_compile.c and pcre_dfa_exec.c.
2474 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2475 caused by fix #2 above. (Subsequently also a second patch to fix the
2476 first patch. And a third patch - this was a messy problem.)
2478 9. Applied Craig's patch to remove the use of push_back().
2480 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2481 matching function regexec().
2483 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2484 which, however, unlike Perl's \g{...}, are subroutine calls, not back
2485 references. PCRE supports relative numbers with this syntax (I don't think
2486 Oniguruma does).
2488 12. Previously, a group with a zero repeat such as (...){0} was completely
2489 omitted from the compiled regex. However, this means that if the group
2490 was called as a subroutine from elsewhere in the pattern, things went wrong
2491 (an internal error was given). Such groups are now left in the compiled
2492 pattern, with a new opcode that causes them to be skipped at execution
2493 time.
2495 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2496 to the way PCRE behaves:
2498 (a) A lone ] character is dis-allowed (Perl treats it as data).
2500 (b) A back reference to an unmatched subpattern matches an empty string
2501 (Perl fails the current match path).
2503 (c) A data ] in a character class must be notated as \] because if the
2504 first data character in a class is ], it defines an empty class. (In
2505 Perl it is not possible to have an empty class.) The empty class []
2506 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2507 The negative empty class [^] matches any one character, independently
2508 of the DOTALL setting.
2510 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2511 non-existent subpattern following a character class starting with ']' and
2512 containing () gave an internal compiling error instead of "reference to
2513 non-existent subpattern". Fortunately, when the pattern did exist, the
2514 compiled code was correct. (When scanning forwards to check for the
2515 existence of the subpattern, it was treating the data ']' as terminating
2516 the class, so got the count wrong. When actually compiling, the reference
2517 was subsequently set up correctly.)
2519 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2520 it was being rejected as not supported by pcre_dfa_exec(), even though
2521 other assertions are supported. I have made pcre_dfa_exec() support
2522 (*FAIL).
2524 16. The implementation of 13c above involved the invention of a new opcode,
2525 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2526 cannot be changed at match time, I realized I could make a small
2527 improvement to matching performance by compiling OP_ALLANY instead of
2528 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2529 on the OP_ANY path.
2531 17. Compiling pcretest on Windows with readline support failed without the
2532 following two fixes: (1) Make the unistd.h include conditional on
2533 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2535 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2536 ncurses library to be included for pcretest when ReadLine support is
2537 requested, but also to allow for it to be overridden. This patch came from
2538 Daniel Bergström.
2540 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2541 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2542 any errors with the current Unicode tables. Thanks to Peter Kankowski for
2543 spotting this.
2546 Version 7.6 28-Jan-08
2547 ---------------------
2549 1. A character class containing a very large number of characters with
2550 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2551 overflow.
2553 2. Patch to cut out the "long long" test in pcrecpp_unittest when
2554 HAVE_LONG_LONG is not defined.
2556 3. Applied Christian Ehrlicher's patch to update the CMake build files to
2557 bring them up to date and include new features. This patch includes:
2559 - Fixed PH's badly added libz and libbz2 support.
2560 - Fixed a problem with static linking.
2561 - Added pcredemo. [But later removed - see 7 below.]
2562 - Fixed dftables problem and added an option.
2563 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2565 - Added readline support for pcretest.
2566 - Added an listing of the option settings after cmake has run.
2568 4. A user submitted a patch to Makefile that makes it easy to create
2569 "pcre.dll" under mingw when using Configure/Make. I added stuff to
2570 Makefile.am that cause it to include this special target, without
2571 affecting anything else. Note that the same mingw target plus all
2572 the other distribution libraries and programs are now supported
2573 when configuring with CMake (see 6 below) instead of with
2574 Configure/Make.
2576 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2577 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2578 exported in the Windows port". It has not yet been confirmed that the patch
2579 solves the problem, but it does no harm.
2581 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2582 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2583 with CMake, and also correct the comment about stack recursion.
2585 7. Remove the automatic building of pcredemo from the ./configure system and
2586 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2587 of a program that users should build themselves after PCRE is installed, so
2588 building it automatically is not really right. What is more, it gave
2589 trouble in some build environments.
2591 8. Further tidies to CMakeLists.txt from Sheri and Christian.
2594 Version 7.5 10-Jan-08
2595 ---------------------
2597 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
2598 values in parens when parsing an RE using the C++ wrapper."
2600 2. Negative specials like \S did not work in character classes in UTF-8 mode.
2601 Characters greater than 255 were excluded from the class instead of being
2602 included.
2604 3. The same bug as (2) above applied to negated POSIX classes such as
2605 [:^space:].
2607 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2608 defined or documented. It seems to have been a typo for PCRE_STATIC, so
2609 I have changed it.
2611 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
2612 first named subpattern) and a construct such as (?&a) would reference the
2613 first named subpattern whose name started with "a" (in other words, the
2614 length check was missing). Both these problems are fixed. "Subpattern name
2615 expected" is now given for (?&) (a zero-length name), and this patch also
2616 makes it give the same error for \k'' (previously it complained that that
2617 was a reference to a non-existent subpattern).
2619 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
2620 this is right because (?- can be followed by option settings as well as by
2621 digits. I have, however, made the messages clearer.
2623 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2624 than the number used in the conditional) now cause a compile-time error.
2625 This is actually not compatible with Perl, which accepts such patterns, but
2626 treats the conditional as always being FALSE (as PCRE used to), but it
2627 seems to me that giving a diagnostic is better.
2629 8. Change "alphameric" to the more common word "alphanumeric" in comments
2630 and messages.
2632 9. Fix two occurrences of "backslash" in comments that should have been
2633 "backspace".
2635 10. Remove two redundant lines of code that can never be obeyed (their function
2636 was moved elsewhere).
2638 11. The program that makes PCRE's Unicode character property table had a bug
2639 which caused it to generate incorrect table entries for sequences of
2640 characters that have the same character type, but are in different scripts.
2641 It amalgamated them into a single range, with the script of the first of
2642 them. In other words, some characters were in the wrong script. There were
2643 thirteen such cases, affecting characters in the following ranges:
2645 U+002b0 - U+002c1
2646 U+0060c - U+0060d
2647 U+0061e - U+00612
2648 U+0064b - U+0065e
2649 U+0074d - U+0076d
2650 U+01800 - U+01805
2651 U+01d00 - U+01d77
2652 U+01d9b - U+01dbf
2653 U+0200b - U+0200f
2654 U+030fc - U+030fe
2655 U+03260 - U+0327f
2656 U+0fb46 - U+0fbb1
2657 U+10450 - U+1049d
2659 12. The -o option (show only the matching part of a line) for pcregrep was not
2660 compatible with GNU grep in that, if there was more than one match in a
2661 line, it showed only the first of them. It now behaves in the same way as
2662 GNU grep.
2664 13. If the -o and -v options were combined for pcregrep, it printed a blank
2665 line for every non-matching line. GNU grep prints nothing, and pcregrep now
2666 does the same. The return code can be used to tell if there were any
2667 non-matching lines.
2669 14. Added --file-offsets and --line-offsets to pcregrep.
2671 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2672 infinitely looping recursion. The bug was that positive lookaheads were not
2673 being skipped when checking for a possible empty match (negative lookaheads
2674 and both kinds of lookbehind were skipped).
2676 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2677 inclusion of <windows.h> to before rather than after the definition of
2678 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2680 17. Specifying a possessive quantifier with a specific limit for a Unicode
2681 character property caused pcre_compile() to compile bad code, which led at
2682 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2683 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2684 caused the error; without that there was no problem.
2686 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2688 19. Added --enable-pcretest-libreadline.
2690 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2691 RE::GlobalReplace(). As a result, the number of replacements returned was
2692 double what it should be. I removed one of the increments, but Craig sent a
2693 later patch that removed the other one (the right fix) and added unit tests
2694 that check the return values (which was not done before).
2696 21. Several CMake things:
2698 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2699 the names libpcre and libpcreposix, not just pcre and pcreposix.
2701 (2) The above change means that pcretest and pcregrep are now correctly
2702 linked with the newly-built libraries, not previously installed ones.
2706 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2707 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2708 UTF-8 newline character). The key issue is that the pattern starts .*;
2709 this means that the match must be either at the beginning, or after a
2710 newline. The bug was in the code for advancing after a failed match and
2711 checking that the new position followed a newline. It was not taking
2712 account of UTF-8 characters correctly.
2714 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2715 character classes. PCRE was not treating the sequence [:...:] as a
2716 character class unless the ... were all letters. Perl, however, seems to
2717 allow any characters between [: and :], though of course it rejects as
2718 unknown any "names" that contain non-letters, because all the known class
2719 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2720 for example, whereas PCRE did not - it did not recognize a POSIX character
2721 class. This seemed a bit dangerous, so the code has been changed to be
2722 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2723 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2724 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2725 Perl does, and where it didn't before.
2727 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2728 Windows environments %n is disabled by default.
2731 Version 7.4 21-Sep-07
2732 ---------------------
2734 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2735 means that a class such as [\s] counted as "explicit reference to CR or
2736 LF". That isn't really right - the whole point of the change was to try to
2737 help when there was an actual mention of one of the two characters. So now
2738 the change happens only if \r or \n (or a literal CR or LF) character is
2739 encountered.
2741 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2742 of both had grown to the point where there were only 3 bits left.
2743 Fortunately, there was spare space in the data structure, and so I have
2744 moved the internal flags into a new 16-bit field to free up more option
2745 bits.
2747 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2748 but did not set the internal JCHANGED flag - either of these is enough to
2749 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2750 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2751 start sets both bits.
2753 4. Added options (at build time, compile time, exec time) to change \R from
2754 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2756 5. doc/pcresyntax.html was missing from the distribution.
2758 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2759 compatibility, even though it is no longer used.
2761 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2762 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2763 windows.h file is present (where different names are used). [This was
2764 reversed later after testing - see 16 below.]
2766 8. Changed all #include <config.h> to #include "config.h". There were also
2767 some further <pcre.h> cases that I changed to "pcre.h".
2769 9. When pcregrep was used with the --colour option, it missed the line ending
2770 sequence off the lines that it output.
2772 10. It was pointed out to me that arrays of string pointers cause lots of
2773 relocations when a shared library is dynamically loaded. A technique of
2774 using a single long string with a table of offsets can drastically reduce
2775 these. I have refactored PCRE in four places to do this. The result is
2776 dramatic:
2778 Originally: 290
2779 After changing UCP table: 187
2780 After changing error message table: 43
2781 After changing table of "verbs" 36
2782 After changing table of Posix names 22
2784 Thanks to the folks working on Gregex for glib for this insight.
2786 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2787 unicode-properties was also set.
2789 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2791 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2792 checked only for CRLF.
2794 14. Added casts to pcretest.c to avoid compiler warnings.
2796 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2798 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2799 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2800 entirely. This removes changes made in 7 above.
2802 17. The CMake files have been updated, and there is now more information about
2803 building with CMake in the NON-UNIX-USE document.
2806 Version 7.3 28-Aug-07
2807 ---------------------
2809 1. In the rejigging of the build system that eventually resulted in 7.1, the
2810 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2811 brackets there is not right, since it causes compilers to look for an
2812 installed pcre.h, not the version that is in the source that is being
2813 compiled (which of course may be different). I have changed it back to:
2815 #include "pcre.h"
2817 I have a vague recollection that the change was concerned with compiling in
2818 different directories, but in the new build system, that is taken care of
2819 by the VPATH setting the Makefile.
2821 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2822 when the subject happened to end in the byte 0x85 (e.g. if the last
2823 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2824 characters but of course it shouldn't be taken as a newline when it is part
2825 of another character. The bug was that, for an unlimited repeat of . in
2826 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2827 characters when looking for a newline.
2829 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2831 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2832 in debug output.
2834 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2835 long printing in the pcrecpp unittest when running under MinGW.
2837 6. ESC_K was left out of the EBCDIC table.
2839 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2840 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2841 limit also applies to "virtual nesting" when a pattern is recursive, and in
2842 this case 1000 isn't so big. I have been able to remove this limit at the
2843 expense of backing off one optimization in certain circumstances. Normally,
2844 when pcre_exec() would call its internal match() function recursively and
2845 immediately return the result unconditionally, it uses a "tail recursion"
2846 feature to save stack. However, when a subpattern that can match an empty
2847 string has an unlimited repetition quantifier, it no longer makes this
2848 optimization. That gives it a stack frame in which to save the data for
2849 checking that an empty string has been matched. Previously this was taken
2850 from the 1000-entry workspace that had been reserved. So now there is no
2851 explicit limit, but more stack is used.
2853 8. Applied Daniel's patches to solve problems with the import/export magic
2854 syntax that is required for Windows, and which was going wrong for the
2855 pcreposix and pcrecpp parts of the library. These were overlooked when this
2856 problem was solved for the main library.
2858 9. There were some crude static tests to avoid integer overflow when computing
2859 the size of patterns that contain repeated groups with explicit upper
2860 limits. As the maximum quantifier is 65535, the maximum group length was
2861 set at 30,000 so that the product of these two numbers did not overflow a
2862 32-bit integer. However, it turns out that people want to use groups that
2863 are longer than 30,000 bytes (though not repeat them that many times).
2864 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2865 made it possible to implement the integer overflow checks in a much more
2866 dynamic way, which I have now done. The artificial limitation on group
2867 length has been removed - we now have only the limit on the total length of
2868 the compiled pattern, which depends on the LINK_SIZE setting.
2870 10. Fixed a bug in the documentation for get/copy named substring when
2871 duplicate names are permitted. If none of the named substrings are set, the
2872 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2873 empty string.
2875 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2876 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2877 because the ] is interpreted as the first data character and the
2878 terminating ] is not found. PCRE has been made compatible with Perl in this
2879 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2880 cause memory overwriting.
2882 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2883 string has been matched (to stop an infinite loop). It was not recognizing
2884 a conditional subpattern that could match an empty string if that
2885 subpattern was within another subpattern. For example, it looped when
2886 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2887 condition was not nested. This bug has been fixed.
2889 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2890 past the start of the subject in the presence of bytes with the top bit
2891 set, for example "\x8aBCD".
2893 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2894 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2896 14. Optimized (?!) to (*FAIL).
2898 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2899 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2900 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2901 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2902 does: it's just the validity check that is more restrictive.
2904 16. Inserted checks for integer overflows during escape sequence (backslash)
2905 processing, and also fixed erroneous offset values for syntax errors during
2906 backslash processing.
2908 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2909 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2911 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2912 caused an overrun.
2914 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2915 something other than just ASCII characters) inside a group that had an
2916 unlimited repeat caused a loop at compile time (while checking to see
2917 whether the group could match an empty string).
2919 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2920 [\P{Any}] did so. (Error in the code for printing property names.)
2922 21. An orphan \E inside a character class could cause a crash.
2924 22. A repeated capturing bracket such as (A)? could cause a wild memory
2925 reference during compilation.
2927 23. There are several functions in pcre_compile() that scan along a compiled
2928 expression for various reasons (e.g. to see if it's fixed length for look
2929 behind). There were bugs in these functions when a repeated \p or \P was
2930 present in the pattern. These operators have additional parameters compared
2931 with \d, etc, and these were not being taken into account when moving along
2932 the compiled data. Specifically:
2934 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2935 length.
2937 (b) An item such as \pL+ within a repeated group could cause crashes or
2938 loops.
2940 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2941 "reference to non-existent subpattern" error.
2943 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2945 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2946 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2948 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2950 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2951 character were causing crashes (broken optimization).
2953 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2954 \p or \P) caused a compile-time loop.
2956 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2957 break. For example, the unstudied pattern [\r\n]A does not match the string
2958 "\r\nA" because change 7.0/46 below moves the current point on by two
2959 characters after failing to match at the start. However, the pattern \nA
2960 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2961 the same is true. There doesn't seem any very clean way out of this, but
2962 what I have chosen to do makes the common cases work: PCRE now takes note
2963 of whether there can be an explicit match for \r or \n anywhere in the
2964 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2965 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2966 pattern has explicit CR or LF references.
2968 29. Added (*CR) etc for changing newline setting at start of pattern.
2971 Version 7.2 19-Jun-07
2972 ---------------------
2974 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2975 which is apparently normally available under Windows.
2977 2. Re-jig the pcregrep tests with different newline settings in an attempt
2978 to make them independent of the local environment's newline setting.
2980 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2982 4. Some of the "internals" tests were previously cut out when the link size
2983 was not 2, because the output contained actual offsets. The recent new
2984 "Z" feature of pcretest means that these can be cut out, making the tests
2985 usable with all link sizes.
2987 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2988 stack recursion. This gives a massive performance boost under BSD, but just
2989 a small improvement under Linux. However, it saves one field in the frame
2990 in all cases.
2992 6. Added more features from the forthcoming Perl 5.10:
2994 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2995 recursion call. It refers to the nth most recently opened parentheses.
2997 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2998 to be opened parentheses.
3000 (c) Conditions that refer to capturing parentheses can be specified
3001 relatively, for example, (?(-2)... or (?(+3)...
3003 (d) \K resets the start of the current match so that everything before
3004 is not part of it.
3006 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
3008 (f) \g{name} is another synonym - part of Perl 5.10's unification of
3009 reference syntax.
3011 (g) (?| introduces a group in which the numbering of parentheses in each
3012 alternative starts with the same number.
3014 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
3016 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
3019 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
3020 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
3021 for detecting groups that can match an empty string.
3023 9. A pattern with a very large number of alternatives (more than several
3024 hundred) was running out of internal workspace during the pre-compile
3025 phase, where pcre_compile() figures out how much memory will be needed. A
3026 bit of new cunning has reduced the workspace needed for groups with
3027 alternatives. The 1000-alternative test pattern now uses 12 bytes of
3028 workspace instead of running out of the 4096 that are available.
3030 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
3032 11. Applied patch from Google to remove an optimization that didn't quite work.
3033 The report of the bug said:
3035 pcrecpp::RE("a*").FullMatch("aaa") matches, while
3036 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
3037 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
3039 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
3040 it matched the wrong number of bytes.
3043 Version 7.1 24-Apr-07
3044 ---------------------
3046 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
3047 that is more "standard", making use of automake and other Autotools. There
3048 is some re-arrangement of the files and adjustment of comments consequent
3049 on this.
3051 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
3052 for recursive directory scanning broke on some systems because the files
3053 are not scanned in any specific order and on different systems the order
3054 was different. A call to "sort" has been inserted into RunGrepTest for the
3055 approprate test as a short-term fix. In the longer term there may be an
3056 alternative.
3058 3. I had an email from Eric Raymond about problems translating some of PCRE's
3059 man pages to HTML (despite the fact that I distribute HTML pages, some
3060 people do their own conversions for various reasons). The problems
3061 concerned the use of low-level troff macros .br and .in. I have therefore
3062 removed all such uses from the man pages (some were redundant, some could
3063 be replaced by .nf/.fi pairs). The 132html script that I use to generate
3064 HTML has been updated to handle .nf/.fi and to complain if it encounters
3065 .br or .in.
3067 4. Updated comments in configure.ac that get placed in config.h.in and also
3068 arranged for config.h to be included in the distribution, with the name
3069 config.h.generic, for the benefit of those who have to compile without
3070 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
3072 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
3073 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
3074 makevp.bat for the current PCRE, using the additional files
3075 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
3077 6. A Windows user reported a minor discrepancy with test 2, which turned out
3078 to be caused by a trailing space on an input line that had got lost in his
3079 copy. The trailing space was an accident, so I've just removed it.
3081 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
3082 that is needed.
3084 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
3085 as "const" (a) because they are and (b) because it helps the PHP
3086 maintainers who have recently made a script to detect big data structures
3087 in the php code that should be moved to the .rodata section. I remembered
3088 to update Builducptable as well, so it won't revert if ucptable.h is ever
3089 re-created.
3091 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
3092 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
3093 order to be able to cut out the UTF-8 tables in the latter when UTF-8
3094 support is not required. This saves 1.5-2K of code, which is important in
3095 some applications.
3097 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
3098 so as not to refer to the tables, even though these functions will never be
3099 called when UTF-8 support is disabled. Otherwise there are problems with a
3100 shared library.
3102 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
3104 (a) It was defining its arguments as char * instead of void *.
3106 (b) It was assuming that all moves were upwards in memory; this was true
3107 a long time ago when I wrote it, but is no longer the case.
3109 The emulated memove() is provided for those environments that have neither
3110 memmove() nor bcopy(). I didn't think anyone used it these days, but that
3111 is clearly not the case, as these two bugs were recently reported.
3113 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
3114 and Detrail to create the HTML documentation, the .txt form of the man
3115 pages, and it removes trailing spaces from listed files. It also creates
3116 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
3117 case, it wraps all the #defines with #ifndefs. This script should be run
3118 before "make dist".
3120 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
3121 with Unicode property support.
3123 (a) For a maximizing quantifier, if the two different cases of the
3124 character were of different lengths in their UTF-8 codings (there are
3125 some cases like this - I found 11), and the matching function had to
3126 back up over a mixture of the two cases, it incorrectly assumed they
3127 were both the same length.
3129 (b) When PCRE was configured to use the heap rather than the stack for
3130 recursion during matching, it was not correctly preserving the data for
3131 the other case of a UTF-8 character when checking ahead for a match
3132 while processing a minimizing repeat. If the check also involved
3133 matching a wide character, but failed, corruption could cause an
3134 erroneous result when trying to check for a repeat of the original
3135 character.
3137 13. Some tidying changes to the testing mechanism:
3139 (a) The RunTest script now detects the internal link size and whether there
3140 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
3141 values substituted by "configure". (The RunGrepTest script already did
3142 this for UTF-8.) The configure.ac script no longer substitutes the
3143 relevant variables.
3145 (b) The debugging options /B and /D in pcretest show the compiled bytecode
3146 with length and offset values. This means that the output is different
3147 for different internal link sizes. Test 2 is skipped for link sizes
3148 other than 2 because of this, bypassing the problem. Unfortunately,
3149 there was also a test in test 3 (the locale tests) that used /B and
3150 failed for link sizes other than 2. Rather than cut the whole test out,
3151 I have added a new /Z option to pcretest that replaces the length and
3152 offset values with spaces. This is now used to make test 3 independent
3153 of link size. (Test 2 will be tidied up later.)
3155 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
3156 segmentation fault instead of returning the appropriate error message.
3158 15. In multiline mode when the newline sequence was set to "any", the pattern
3159 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
3160 This doesn't seem right; it now treats the CRLF combination as the line
3161 ending, and so does not match in that case. It's only a pattern such as ^$
3162 that would hit this one: something like ^ABC$ would have failed after \r
3163 and then tried again after \r\n.
3165 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
3166 in an attempt to make files that differ only in their line terminators
3167 compare equal. This works on Linux.
3169 17. Under certain error circumstances pcregrep might try to free random memory
3170 as it exited. This is now fixed, thanks to valgrind.
3172 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
3173 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
3174 was because its rules for how to advance for /g after matching an empty
3175 string at the end of a line did not allow for this case. They now check for
3176 it specially.
3178 20. pcretest is supposed to handle patterns and data of any length, by
3179 extending its buffers when necessary. It was getting this wrong when the
3180 buffer for a data line had to be extended.
3182 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
3183 CRLF as a newline sequence.
3185 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
3186 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
3187 I have nevertheless tidied it up.
3189 23. Added some casts to kill warnings from HP-UX ia64 compiler.
3191 24. Added a man page for pcre-config.
3194 Version 7.0 19-Dec-06
3195 ---------------------
3197 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
3198 moving to gcc 4.1.1.
3200 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
3201 sys/time.h, which is documented as needed for this function. It doesn't
3202 seem to matter on Linux, but it showed up on some releases of OS X.
3204 3. It seems that there are systems where bytes whose values are greater than
3205 127 match isprint() in the "C" locale. The "C" locale should be the
3206 default when a C program starts up. In most systems, only ASCII printing
3207 characters match isprint(). This difference caused the output from pcretest
3208 to vary, making some of the tests fail. I have changed pcretest so that:
3210 (a) When it is outputting text in the compiled version of a pattern, bytes
3211 other than 32-126 are always shown as hex escapes.
3213 (b) When it is outputting text that is a matched part of a subject string,
3214 it does the same, unless a different locale has been set for the match
3215 (using the /L modifier). In this case, it uses isprint() to decide.
3217 4. Fixed a major bug that caused incorrect computation of the amount of memory
3218 required for a compiled pattern when options that changed within the
3219 pattern affected the logic of the preliminary scan that determines the
3220 length. The relevant options are -x, and -i in UTF-8 mode. The result was
3221 that the computed length was too small. The symptoms of this bug were
3222 either the PCRE error "internal error: code overflow" from pcre_compile(),
3223 or a glibc crash with a message such as "pcretest: free(): invalid next
3224 size (fast)". Examples of patterns that provoked this bug (shown in
3225 pcretest format) are:
3227 /(?-x: )/x
3228 /(?x)(?-x: \s*#\s*)/
3229 /((?i)[\x{c0}])/8
3230 /(?i:[\x{c0}])/8
3232 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
3233 is now done differently.
3235 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
3236 wrapper classes; (b) implement a new function in the C++ scanner that is
3237 more efficient than the old way of doing things because it avoids levels of
3238 recursion in the regex matching; (c) add a paragraph to the documentation
3239 for the FullMatch() function.
3241 6. The escape sequence \n was being treated as whatever was defined as
3242 "newline". Not only was this contrary to the documentation, which states
3243 that \n is character 10 (hex 0A), but it also went horribly wrong when
3244 "newline" was defined as CRLF. This has been fixed.
3246 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
3247 was being set to -1 for the "end of line" case (supposedly a value that no
3248 character can have). Though this value is never used (the check for end of
3249 line is "zero bytes in current character"), it caused compiler complaints.
3250 I've changed it to 0xffffffff.
3252 8. In pcre_version.c, the version string was being built by a sequence of
3253 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
3254 string (as it is for production releases) called a macro with an empty
3255 argument. The C standard says the result of this is undefined. The gcc
3256 compiler treats it as an empty string (which was what was wanted) but it is
3257 reported that Visual C gives an error. The source has been hacked around to
3258 avoid this problem.
3260 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
3261 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
3262 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
3263 of them did).
3265 10. Originally, pcretest opened its input and output without "b"; then I was
3266 told that "b" was needed in some environments, so it was added for release
3267 5.0 to both the input and output. (It makes no difference on Unix-like
3268 systems.) Later I was told that it is wrong for the input on Windows. I've
3269 now abstracted the modes into two macros, to make it easier to fiddle with
3270 them, and removed "b" from the input mode under Windows.
3272 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
3274 12. Added -help and --help to pcretest as an official way of being reminded
3275 of the options.
3277 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
3278 and pcrecpp.cc because they annoy compilers at high warning levels.
3280 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
3282 15. Fixed an occurrence of == in configure.ac that should have been = (shell
3283 scripts are not C programs :-) and which was not noticed because it works
3284 on Linux.
3286 16. pcretest is supposed to handle any length of pattern and data line (as one
3287 line or as a continued sequence of lines) by extending its input buffer if
3288 necessary. This feature was broken for very long pattern lines, leading to
3289 a string of junk being passed to pcre_compile() if the pattern was longer
3290 than about 50K.
3292 17. I have done a major re-factoring of the way pcre_compile() computes the
3293 amount of memory needed for a compiled pattern. Previously, there was code
3294 that made a preliminary scan of the pattern in order to do this. That was
3295 OK when PCRE was new, but as the facilities have expanded, it has become
3296 harder and harder to keep it in step with the real compile phase, and there
3297 have been a number of bugs (see for example, 4 above). I have now found a
3298 cunning way of running the real compile function in a "fake" mode that
3299 enables it to compute how much memory it would need, while actually only
3300 ever using a few hundred bytes of working memory and without too many
3301 tests of the mode. This should make future maintenance and development
3302 easier. A side effect of this work is that the limit of 200 on the nesting
3303 depth of parentheses has been removed (though this was never a serious
3304 limitation, I suspect). However, there is a downside: pcre_compile() now
3305 runs more slowly than before (30% or more, depending on the pattern). I
3306 hope this isn't a big issue. There is no effect on runtime performance.
3308 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
3309 newline (only possible for the last line of a file) and it was a
3310 pattern that set a locale (followed by /Lsomething), pcretest crashed.
3312 19. Added additional timing features to pcretest. (1) The -tm option now times
3313 matching only, not compiling. (2) Both -t and -tm can be followed, as a
3314 separate command line item, by a number that specifies the number of
3315 repeats to use when timing. The default is 50000; this gives better
3316 precision, but takes uncomfortably long for very large patterns.
3318 20. Extended pcre_study() to be more clever in cases where a branch of a
3319 subpattern has no definite first character. For example, (a*|b*)[cd] would
3320 previously give no result from pcre_study(). Now it recognizes that the
3321 first character must be a, b, c, or d.
3323 21. There was an incorrect error "recursive call could loop indefinitely" if
3324 a subpattern (or the entire pattern) that was being tested for matching an
3325 empty string contained only one non-empty item after a nested subpattern.
3326 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
3327 incorrectly, because the \d was being skipped in the check.
3329 22. The pcretest program now has a new pattern option /B and a command line
3330 option -b, which is equivalent to adding /B to every pattern. This causes
3331 it to show the compiled bytecode, without the additional information that
3332 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
3333 is the same as /B/I).
3335 23. A new optimization is now able automatically to treat some sequences such
3336 as a*b as a*+b. More specifically, if something simple (such as a character
3337 or a simple class like \d) has an unlimited quantifier, and is followed by
3338 something that cannot possibly match the quantified thing, the quantifier
3339 is automatically "possessified".
3341 24. A recursive reference to a subpattern whose number was greater than 39
3342 went wrong under certain circumstances in UTF-8 mode. This bug could also
3343 have affected the operation of pcre_study().
3345 25. Realized that a little bit of performance could be had by replacing
3346 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
3348 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
3350 27. Possessive quantifiers such as a++ were previously implemented by turning
3351 them into atomic groups such as ($>a+). Now they have their own opcodes,
3352 which improves performance. This includes the automatically created ones
3353 from 23 above.
3355 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
3356 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
3357 the first matched character to be a colon. This applied both to named and
3358 numbered groups.
3360 29. The ucpinternal.h header file was missing its idempotency #ifdef.
3362 30. I was sent a "project" file called libpcre.a.dev which I understand makes
3363 building PCRE on Windows easier, so I have included it in the distribution.
3365 31. There is now a check in pcretest against a ridiculously large number being
3366 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
3367 loop, the loop is abandoned.
3369 32. Forward references to subpatterns in conditions such as (?(2)...) where
3370 subpattern 2 is defined later cause pcre_compile() to search forwards in
3371 the pattern for the relevant set of parentheses. This search went wrong
3372 when there were unescaped parentheses in a character class, parentheses
3373 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
3375 33. "Subroutine" calls and backreferences were previously restricted to
3376 referencing subpatterns earlier in the regex. This restriction has now
3377 been removed.
3379 34. Added a number of extra features that are going to be in Perl 5.10. On the
3380 whole, these are just syntactic alternatives for features that PCRE had
3381 previously implemented using the Python syntax or my own invention. The
3382 other formats are all retained for compatibility.
3384 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
3385 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
3386 also .NET compatible.
3388 (b) A recursion or subroutine call to a named group can now be defined as
3389 (?&name) as well as (?P>name).
3391 (c) A backreference to a named group can now be defined as \k<name> or
3392 \k'name' as well as (?P=name). The new forms, as well as being in Perl
3393 5.10, are also .NET compatible.
3395 (d) A conditional reference to a named group can now use the syntax
3396 (?(<name>) or (?('name') as well as (?(name).
3398 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
3399 groups (named and numbered) that are never evaluated inline, but can be
3400 called as "subroutines" from elsewhere. In effect, the DEFINE condition
3401 is always false. There may be only one alternative in such a group.
3403 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
3404 as the simple (?(R). The condition is true only if the most recent
3405 recursion is that of the given number or name. It does not search out
3406 through the entire recursion stack.
3408 (g) The escape \gN or \g{N} has been added, where N is a positive or
3409 negative number, specifying an absolute or relative reference.
3411 35. Tidied to get rid of some further signed/unsigned compiler warnings and
3412 some "unreachable code" warnings.
3414 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
3415 things, this adds five new scripts.
3417 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
3418 There were also incompatibilities regarding the handling of \Q..\E inside
3419 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
3420 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
3422 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
3423 matches an empty string, and forcibly breaks the loop. There were bugs in
3424 this code in non-simple cases. For a pattern such as ^(a()*)* matched
3425 against aaaa the result was just "a" rather than "aaaa", for example. Two
3426 separate and independent bugs (that affected different cases) have been
3427 fixed.
3429 39. Refactored the code to abolish the use of different opcodes for small
3430 capturing bracket numbers. This is a tidy that I avoided doing when I
3431 removed the limit on the number of capturing brackets for 3.5 back in 2001.
3432 The new approach is not only tidier, it makes it possible to reduce the
3433 memory needed to fix the previous bug (38).
3435 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3436 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3437 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3438 mode.
3440 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3441 report.
3443 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3444 copy construction and assignment in the C++ wrapper.
3446 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3447 couple of bugs that could have given wrong results in the "--newline=crlf"
3448 case.
3450 44. Added a number of casts and did some reorganization of signed/unsigned int
3451 variables following suggestions from Dair Grant. Also renamed the variable
3452 "this" as "item" because it is a C++ keyword.
3454 45. Arranged for dftables to add
3456 #include "pcre_internal.h"
3458 to pcre_chartables.c because without it, gcc 4.x may remove the array
3459 definition from the final binary if PCRE is built into a static library and
3460 dead code stripping is activated.
3462 46. For an unanchored pattern, if a match attempt fails at the start of a
3463 newline sequence, and the newline setting is CRLF or ANY, and the next two
3464 characters are CRLF, advance by two characters instead of one.
3467 Version 6.7 04-Jul-06
3468 ---------------------
3470 1. In order to handle tests when input lines are enormously long, pcretest has
3471 been re-factored so that it automatically extends its buffers when
3472 necessary. The code is crude, but this _is_ just a test program. The
3473 default size has been increased from 32K to 50K.
3475 2. The code in pcre_study() was using the value of the re argument before
3476 testing it for NULL. (Of course, in any sensible call of the function, it
3477 won't be NULL.)
3479 3. The memmove() emulation function in pcre_internal.h, which is used on
3480 systems that lack both memmove() and bcopy() - that is, hardly ever -
3481 was missing a "static" storage class specifier.
3483 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3484 containing an extended class (one that cannot be represented by a bitmap
3485 because it contains high-valued characters or Unicode property items, e.g.
3486 [\pZ]). Almost always one would set UTF-8 mode when processing such a
3487 pattern, but PCRE should not loop if you do not (it no longer does).
3488 [Detail: two cases were found: (a) a repeated subpattern containing an
3489 extended class; (b) a recursive reference to a subpattern that followed a
3490 previous extended class. It wasn't skipping over the extended class
3491 correctly when UTF-8 mode was not set.]
3493 5. A negated single-character class was not being recognized as fixed-length
3494 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3495 compile error "lookbehind assertion is not fixed length".
3497 6. The RunPerlTest auxiliary script was showing an unexpected difference
3498 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3499 write a Perl script that can interpret lines of an input file either as
3500 byte characters or as UTF-8, which is what "perltest" was being required to
3501 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3502 can't do is switch easily at run time between having the "use utf8;" pragma
3503 or not. In the end, I fudged it by using the RunPerlTest script to insert
3504 "use utf8;" explicitly for the UTF-8 tests.
3506 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3507 the end of the subject string, contrary to the documentation and to what
3508 Perl does. This was true of both matching functions. Now it matches only at
3509 the start of the subject and immediately after *internal* newlines.
3511 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3512 a pointer to an int instead of a pointer to an unsigned long int. This
3513 caused problems on 64-bit systems.
3515 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3516 instance of the 'standard' template library not being so standard".
3518 10. There was no check on the number of named subpatterns nor the maximum
3519 length of a subpattern name. The product of these values is used to compute
3520 the size of the memory block for a compiled pattern. By supplying a very
3521 long subpattern name and a large number of named subpatterns, the size
3522 computation could be caused to overflow. This is now prevented by limiting
3523 the length of names to 32 characters, and the number of named subpatterns
3524 to 10,000.
3526 11. Subpatterns that are repeated with specific counts have to be replicated in
3527 the compiled pattern. The size of memory for this was computed from the
3528 length of the subpattern and the repeat count. The latter is limited to
3529 65535, but there was no limit on the former, meaning that integer overflow
3530 could in principle occur. The compiled length of a repeated subpattern is
3531 now limited to 30,000 bytes in order to prevent this.
3533 12. Added the optional facility to have named substrings with the same name.
3535 13. Added the ability to use a named substring as a condition, using the
3536 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3537 are numbers (not recommended). Forward references are permitted.
3539 14. Added forward references in named backreferences (if you see what I mean).
3541 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3542 pattern could run off the end of the subject. For example, the pattern
3543 "(?s)(.{1,5})"8 did this with the subject "ab".
3545 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3546 PCRE_CASELESS was set when matching characters that were quantified with ?
3547 or *.
3549 17. A character class other than a single negated character that had a minimum
3550 but no maximum quantifier - for example [ab]{6,} - was not handled
3551 correctly by pce_dfa_exec(). It would match only one character.
3553 18. A valid (though odd) pattern that looked like a POSIX character
3554 class but used an invalid character after [ (for example [[,abc,]]) caused
3555 pcre_compile() to give the error "Failed: internal error: code overflow" or
3556 in some cases to crash with a glibc free() error. This could even happen if
3557 the pattern terminated after [[ but there just happened to be a sequence of
3558 letters, a binary zero, and a closing ] in the memory that followed.
3560 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3561 over the years. Originally (before any Unicode support), just the bottom 8
3562 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3563 output from "man perlunicode" includes this:
3565 The regular expression compiler produces polymorphic opcodes. That
3566 is, the pattern adapts to the data and automatically switches to
3567 the Unicode character scheme when presented with Unicode data--or
3568 instead uses a traditional byte scheme when presented with byte
3569 data.
3571 Sadly, a wide octal escape does not cause a switch, and in a string with
3572 no other multibyte characters, these octal escapes are treated as before.
3573 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
3574 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3575 Unicode string.
3577 I have not perpetrated such confusion in PCRE. Up till now, it took just
3578 the bottom 8 bits, as in old Perl. I have now made octal escapes with
3579 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3580 translate to the appropriate multibyte character.
3582 29. Applied some refactoring to reduce the number of warnings from Microsoft
3583 and Borland compilers. This has included removing the fudge introduced
3584 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3585 a warning about an unused variable.
3587 21. PCRE has not included VT (character 0x0b) in the set of whitespace
3588 characters since release 4.0, because Perl (from release 5.004) does not.
3589 [Or at least, is documented not to: some releases seem to be in conflict
3590 with the documentation.] However, when a pattern was studied with
3591 pcre_study() and all its branches started with \s, PCRE still included VT
3592 as a possible starting character. Of course, this did no harm; it just
3593 caused an unnecessary match attempt.
3595 22. Removed a now-redundant internal flag bit that recorded the fact that case
3596 dependency changed within the pattern. This was once needed for "required
3597 byte" processing, but is no longer used. This recovers a now-scarce options
3598 bit. Also moved the least significant internal flag bit to the most-
3599 significant bit of the word, which was not previously used (hangover from
3600 the days when it was an int rather than a uint) to free up another bit for
3601 the future.
3603 23. Added support for CRLF line endings as well as CR and LF. As well as the
3604 default being selectable at build time, it can now be changed at runtime
3605 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3606 specify that it is scanning data with non-default line endings.
3608 24. Changed the definition of CXXLINK to make it agree with the definition of
3609 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3611 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3612 recursions. This makes a big different to stack usage for some patterns.
3614 26. If a subpattern containing a named recursion or subroutine reference such
3615 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3616 the space required for the compiled pattern went wrong and gave too small a
3617 value. Depending on the environment, this could lead to "Failed: internal
3618 error: code overflow at offset 49" or "glibc detected double free or
3619 corruption" errors.
3621 27. Applied patches from Google (a) to support the new newline modes and (b) to
3622 advance over multibyte UTF-8 characters in GlobalReplace.
3624 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3625 difference for some implementation of PCRE in some Windows version.
3627 29. Added some extra testing facilities to pcretest:
3629 \q<number> in a data line sets the "match limit" value
3630 \Q<number> in a data line sets the "match recursion limt" value
3631 -S <number> sets the stack size, where <number> is in megabytes
3633 The -S option isn't available for Windows.
3636 Version 6.6 06-Feb-06
3637 ---------------------
3639 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3640 in pcreposix.h. I have copied the definition from pcre.h.
3642 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3643 because pcre.h is no longer a built file.
3645 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3646 not normally included in the compiled code.
3649 Version 6.5 01-Feb-06
3650 ---------------------
3652 1. When using the partial match feature with pcre_dfa_exec(), it was not
3653 anchoring the second and subsequent partial matches at the new starting
3654 point. This could lead to incorrect results. For example, with the pattern
3655 /1234/, partially matching against "123" and then "a4" gave a match.
3657 2. Changes to pcregrep:
3659 (a) All non-match returns from pcre_exec() were being treated as failures
3660 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3661 error message is output. Some extra information is given for the
3663 probably the only errors that are likely to be caused by users (by
3664 specifying a regex that has nested indefinite repeats, for instance).
3665 If there are more than 20 of these errors, pcregrep is abandoned.
3667 (b) A binary zero was treated as data while matching, but terminated the
3668 output line if it was written out. This has been fixed: binary zeroes
3669 are now no different to any other data bytes.
3671 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3672 used to set a locale for matching. The --locale=xxxx long option has
3673 been added (no short equivalent) to specify a locale explicitly on the
3674 pcregrep command, overriding the environment variables.
3676 (d) When -B was used with -n, some line numbers in the output were one less
3677 than they should have been.
3679 (e) Added the -o (--only-matching) option.
3681 (f) If -A or -C was used with -c (count only), some lines of context were
3682 accidentally printed for the final match.
3684 (g) Added the -H (--with-filename) option.
3686 (h) The combination of options -rh failed to suppress file names for files
3687 that were found from directory arguments.
3689 (i) Added the -D (--devices) and -d (--directories) options.
3691 (j) Added the -F (--fixed-strings) option.
3693 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3695 (l) Added the --colo(u)r option.
3697 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3698 is not present by default.
3700 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3701 items such as (?R) or (?1), when the recursion could match a number of
3702 alternatives. If it matched one of the alternatives, but subsequently,
3703 outside the recursion, there was a failure, the code tried to back up into
3704 the recursion. However, because of the way PCRE is implemented, this is not
3705 possible, and the result was an incorrect result from the match.
3707 In order to prevent this happening, the specification of recursion has
3708 been changed so that all such subpatterns are automatically treated as
3709 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3711 4. I had overlooked the fact that, in some locales, there are characters for
3712 which isalpha() is true but neither isupper() nor islower() are true. In
3713 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3714 and ordfeminine) are like this. This affected the treatment of \w and \W
3715 when they appeared in character classes, but not when they appeared outside
3716 a character class. The bit map for "word" characters is now created
3717 separately from the results of isalnum() instead of just taking it from the
3718 upper, lower, and digit maps. (Plus the underscore character, of course.)
3720 5. The above bug also affected the handling of POSIX character classes such as
3721 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3722 permanent tables. Instead, the bit maps for such a class were previously
3723 created as the appropriate unions of the upper, lower, and digit bitmaps.
3724 Now they are created by subtraction from the [[:word:]] class, which has
3725 its own bitmap.
3727 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3728 It is created by subtracting the vertical space characters (\x09, \x0a,
3729 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3730 subtraction was done in the overall bitmap for a character class, meaning
3731 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3732 be recognized. This bug has been fixed.
3734 7. Patches from the folks at Google:
3736 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3737 real life, but is still worth protecting against".
3739 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3740 regular expressions".
3742 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3743 have it.
3745 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3746 "configure" and the latter not, in order to fix a problem somebody had
3747 with compiling the Arg class on HP-UX.
3749 (e) Improve the error-handling of the C++ wrapper a little bit.
3751 (f) New tests for checking recursion limiting.
3753 8. The pcre_memmove() function, which is used only if the environment does not
3754 have a standard memmove() function (and is therefore rarely compiled),
3755 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3756 returning a result (though PCRE never actually uses the result).
3758 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3759 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3760 returned instead of calling malloc() with an overflowing number that would
3761 most likely cause subsequent chaos.
3763 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3765 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3766 with this option is matched, the nmatch and pmatch options of regexec() are
3767 ignored.
3769 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3770 provided in case anyone wants to the the POSIX interface with UTF-8
3771 strings.
3773 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3774 C++ linking (needed for some HP-UX environments).
3776 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3777 (unused parameter) and in the pcre_printint() function (omitted "default"
3778 switch label when the default is to do nothing).
3780 15. Added some code to make it possible, when PCRE is compiled as a C++
3781 library, to replace subject pointers for pcre_exec() with a smart pointer
3782 class, thus making it possible to process discontinuous strings.
3784 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3785 much the same function. They were added by different people who were trying
3786 to make PCRE easy to compile on non-Unix systems. It has been suggested
3787 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3788 for compiling on Windows systems. I have therefore replaced it with
3789 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3790 defaults to "extern" for C or "extern C" for C++, which works fine on
3791 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3792 SCOPE with something explicit in config.h. In addition:
3794 (a) pcreposix.h still had just "extern" instead of either of these macros;
3795 I have replaced it with PCRE_DATA_SCOPE.
3797 (b) Functions such as _pcre_xclass(), which are internal to the library,
3798 but external in the C sense, all had PCRE_EXPORT in their definitions.
3799 This is apparently wrong for the Windows case, so I have removed it.
3800 (It makes no difference on Unix-like systems.)
3802 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3803 of recursive calls to match(). This is different to MATCH_LIMIT because
3804 that limits the total number of calls to match(), not all of which increase
3805 the depth of recursion. Limiting the recursion depth limits the amount of
3806 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3807 when PCRE is compiled, and changed at run time. A patch from Google adds
3808 this functionality to the C++ interface.
3810 18. Changes to the handling of Unicode character properties:
3812 (a) Updated the table to Unicode 4.1.0.
3814 (b) Recognize characters that are not in the table as "Cn" (undefined).
3816 (c) I revised the way the table is implemented to a much improved format
3817 which includes recognition of ranges. It now supports the ranges that
3818 are defined in UnicodeData.txt, and it also amalgamates other
3819 characters into ranges. This has reduced the number of entries in the
3820 table from around 16,000 to around 3,000, thus reducing its size
3821 considerably. I realized I did not need to use a tree structure after
3822 all - a binary chop search is just as efficient. Having reduced the
3823 number of entries, I extended their size from 6 bytes to 8 bytes to
3824 allow for more data.
3826 (d) Added support for Unicode script names via properties such as \p{Han}.
3828 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3829 matching that character.
3831 20. When matching a repeated Unicode property with a minimum greater than zero,
3832 (for example \pL{2,}), PCRE could look past the end of the subject if it
3833 reached it while seeking the minimum number of characters. This could
3834 happen only if some of the characters were more than one byte long, because
3835 there is a check for at least the minimum number of bytes.
3837 21. Refactored the implementation of \p and \P so as to be more general, to
3838 allow for more different types of property in future. This has changed the
3839 compiled form incompatibly. Anybody with saved compiled patterns that use
3840 \p or \P will have to recompile them.
3842 22. Added "Any" and "L&" to the supported property types.
3844 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3845 but give a compile time error if the value is greater than 0xff.
3847 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3848 accidentally not being installed or uninstalled.
3850 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3851 made were to insert the current release number. This seemed silly, because
3852 it made things harder for people building PCRE on systems that don't run
3853 "configure". I have turned pcre.h into a distributed file, no longer built
3854 by "configure", with the version identification directly included. There is
3855 no longer a pcre.h.in file.
3857 However, this change necessitated a change to the pcre-config script as
3858 well. It is built from pcre-config.in, and one of the substitutions was the
3859 release number. I have updated configure.ac so that ./configure now finds
3860 the release number by grepping pcre.h.
3862 26. Added the ability to run the tests under valgrind.
3865 Version 6.4 05-Sep-05
3866 ---------------------
3868 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3869 "--" to be printed when multiple files were scanned, even when none of the
3870 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3871 consider it to be a bug, and have restored the previous behaviour.
3873 2. A couple of code tidies to get rid of compiler warnings.
3875 3. The pcretest program used to cheat by referring to symbols in the library
3876 whose names begin with _pcre_. These are internal symbols that are not
3877 really supposed to be visible externally, and in some environments it is
3878 possible to suppress them. The cheating is now confined to including
3879 certain files from the library's source, which is a bit cleaner.
3881 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3882 file's purpose clearer.
3884 5. Reorganized pcre_ucp_findchar().
3887 Version 6.3 15-Aug-05
3888 ---------------------
3890 1. The file libpcre.pc.in did not have general read permission in the tarball.
3892 2. There were some problems when building without C++ support:
3894 (a) If C++ support was not built, "make install" and "make test" still
3895 tried to test it.
3897 (b) There were problems when the value of CXX was explicitly set. Some
3898 changes have been made to try to fix these, and ...
3900 (c) --disable-cpp can now be used to explicitly disable C++ support.
3902 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3903 backslash in a target when C++ was disabled. This confuses some
3904 versions of "make", apparently. Using an intermediate variable solves
3905 this. (Same for CPP_LOBJ.)
3907 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3908 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3909 necessary on certain architectures.
3911 4. Added a setting of -export-symbols-regex to the link command to remove
3912 those symbols that are exported in the C sense, but actually are local
3913 within the library, and not documented. Their names all begin with
3914 "_pcre_". This is not a perfect job, because (a) we have to except some
3915 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3916 available (and never for static libraries). I have made a note to try to
3917 find a way round (a) in the future.
3920 Version 6.2 01-Aug-05
3921 ---------------------
3923 1. There was no test for integer overflow of quantifier values. A construction
3924 such as {1111111111111111} would give undefined results. What is worse, if
3925 a minimum quantifier for a parenthesized subpattern overflowed and became
3926 negative, the calculation of the memory size went wrong. This could have
3927 led to memory overwriting.
3929 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3931 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3932 operating environments where this matters.
3934 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3935 PCRE options from within the C++ wrapper.
3937 5. Named capturing subpatterns were not being correctly counted when a pattern
3938 was compiled. This caused two problems: (a) If there were more than 100
3939 such subpatterns, the calculation of the memory needed for the whole
3940 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3941 back references of the form \12, where the number was greater than 9, were
3942 not recognized as back references, even though there were sufficient
3943 previous subpatterns.
3945 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3946 versions of gcc, e.g. 2.95.4.
3949 Version 6.1 21-Jun-05
3950 ---------------------
3952 1. There was one reference to the variable "posix" in pcretest.c that was not
3953 surrounded by "#if !defined NOPOSIX".
3955 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3956 the cross-check on the old pcre_info() function, for the benefit of the
3957 cut-down version of PCRE that is currently imported into Exim.
3959 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3960 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3961 patterns that start (?i) but I don't think that matters. The use of (?i) is
3962 just an example; this all applies to the other options as well.
3964 4. Since libtool seems to echo the compile commands it is issuing, the output
3965 from "make" can be reduced a bit by putting "@" in front of each libtool
3966 compile command.
3968 5. Patch from the folks at Google for configure.in to be a bit more thorough
3969 in checking for a suitable C++ installation before trying to compile the
3970 C++ stuff. This should fix a reported problem when a compiler was present,
3971 but no suitable headers.
3973 6. The man pages all had just "PCRE" as their title. I have changed them to
3974 be the relevant file name. I have also arranged that these names are
3975 retained in the file doc/pcre.txt, which is a concatenation in text format
3976 of all the man pages except the little individual ones for each function.
3978 7. The NON-UNIX-USE file had not been updated for the different set of source
3979 files that come with release 6. I also added a few comments about the C++
3980 wrapper.
3983 Version 6.0 07-Jun-05
3984 ---------------------
3986 1. Some minor internal re-organization to help with my DFA experiments.
3988 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3989 didn't matter for the library itself when fully configured, but did matter
3990 when compiling without UCP support, or within Exim, where the ucp files are
3991 not imported.
3993 3. Refactoring of the library code to split up the various functions into
3994 different source modules. The addition of the new DFA matching code (see
3995 below) to a single monolithic source would have made it really too
3996 unwieldy, quite apart from causing all the code to be include in a
3997 statically linked application, when only some functions are used. This is
3998 relevant even without the DFA addition now that patterns can be compiled in
3999 one application and matched in another.
4001 The downside of splitting up is that there have to be some external
4002 functions and data tables that are used internally in different modules of
4003 the library but which are not part of the API. These have all had their
4004 names changed to start with "_pcre_" so that they are unlikely to clash
4005 with other external names.
4007 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
4008 a different (DFA) algorithm. Although it is slower than the original
4009 function, it does have some advantages for certain types of matching
4010 problem.
4012 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
4013 including restarting after a partial match.
4015 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
4016 defined when compiling for Windows was sent to me. I have put it into the
4017 code, though I have no means of testing or verifying it.
4019 7. Added the pcre_refcount() auxiliary function.
4021 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
4022 match before or at the first newline in the subject string. In pcretest,
4023 the /f option on a pattern can be used to set this.
4025 9. A repeated \w when used in UTF-8 mode with characters greater than 256
4026 would behave wrongly. This has been present in PCRE since release 4.0.
4028 10. A number of changes to the pcregrep command:
4030 (a) Refactored how -x works; insert ^(...)$ instead of setting
4031 PCRE_ANCHORED and checking the length, in preparation for adding
4032 something similar for -w.
4034 (b) Added the -w (match as a word) option.
4036 (c) Refactored the way lines are read and buffered so as to have more
4037 than one at a time available.
4039 (d) Implemented a pcregrep test script.
4041 (e) Added the -M (multiline match) option. This allows patterns to match
4042 over several lines of the subject. The buffering ensures that at least
4043 8K, or the rest of the document (whichever is the shorter) is available
4044 for matching (and similarly the previous 8K for lookbehind assertions).
4046 (f) Changed the --help output so that it now says
4048 -w, --word-regex(p)
4050 instead of two lines, one with "regex" and the other with "regexp"
4051 because that confused at least one person since the short forms are the
4052 same. (This required a bit of code, as the output is generated
4053 automatically from a table. It wasn't just a text change.)
4055 (g) -- can be used to terminate pcregrep options if the next thing isn't an
4056 option but starts with a hyphen. Could be a pattern or a path name
4057 starting with a hyphen, for instance.
4059 (h) "-" can be given as a file name to represent stdin.
4061 (i) When file names are being printed, "(standard input)" is used for
4062 the standard input, for compatibility with GNU grep. Previously
4063 "<stdin>" was used.
4065 (j) The option --label=xxx can be used to supply a name to be used for
4066 stdin when file names are being printed. There is no short form.
4068 (k) Re-factored the options decoding logic because we are going to add
4069 two more options that take data. Such options can now be given in four
4070 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
4072 (l) Added the -A, -B, and -C options for requesting that lines of context
4073 around matches be printed.
4075 (m) Added the -L option to print the names of files that do not contain
4076 any matching lines, that is, the complement of -l.
4078 (n) The return code is 2 if any file cannot be opened, but pcregrep does
4079 continue to scan other files.
4081 (o) The -s option was incorrectly implemented. For compatibility with other
4082 greps, it now suppresses the error message for a non-existent or non-
4083 accessible file (but not the return code). There is a new option called
4084 -q that suppresses the output of matching lines, which was what -s was
4085 previously doing.
4087 (p) Added --include and --exclude options to specify files for inclusion
4088 and exclusion when recursing.
4090 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
4091 Hopefully, it now does.
4093 12. Missing cast in pcre_study().
4095 13. Added an "uninstall" target to the makefile.
4097 14. Replaced "extern" in the function prototypes in Makefile.in with
4098 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
4099 world, but is set differently for Windows.
4101 15. Added a second compiling function called pcre_compile2(). The only
4102 difference is that it has an extra argument, which is a pointer to an
4103 integer error code. When there is a compile-time failure, this is set
4104 non-zero, in addition to the error test pointer being set to point to an
4105 error message. The new argument may be NULL if no error number is required
4106 (but then you may as well call pcre_compile(), which is now just a
4107 wrapper). This facility is provided because some applications need a
4108 numeric error indication, but it has also enabled me to tidy up the way
4109 compile-time errors are handled in the POSIX wrapper.
4111 16. Added VPATH=.libs to the makefile; this should help when building with one
4112 prefix path and installing with another. (Or so I'm told by someone who
4113 knows more about this stuff than I do.)
4115 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
4116 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
4117 match everything, including newlines. This is not POSIX-compatible, but
4118 somebody wanted the feature. From pcretest it can be activated by using
4119 both the P and the s flags.
4121 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
4123 19. libpcre.pc was being incorrectly installed as executable.
4125 20. A couple of places in pcretest check for end-of-line by looking for '\n';
4126 it now also looks for '\r' so that it will work unmodified on Windows.
4128 21. Added Google's contributed C++ wrapper to the distribution.
4130 22. Added some untidy missing memory free() calls in pcretest, to keep
4131 Electric Fence happy when testing.
4135 Version 5.0 13-Sep-04
4136 ---------------------
4138 1. Internal change: literal characters are no longer packed up into items
4139 containing multiple characters in a single byte-string. Each character
4140 is now matched using a separate opcode. However, there may be more than one
4141 byte in the character in UTF-8 mode.
4143 2. The pcre_callout_block structure has two new fields: pattern_position and
4144 next_item_length. These contain the offset in the pattern to the next match
4145 item, and its length, respectively.
4147 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
4148 insertion of callouts before each pattern item. Added the /C option to
4149 pcretest to make use of this.
4151 4. On the advice of a Windows user, the lines
4153 #if defined(_WIN32) || defined(WIN32)
4154 _setmode( _fileno( stdout ), 0x8000 );
4155 #endif /* defined(_WIN32) || defined(WIN32) */
4157 have been added to the source of pcretest. This apparently does useful
4158 magic in relation to line terminators.
4160 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
4161 for the benefit of those environments where the "b" makes a difference.
4163 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
4164 to know about it. I have put a hack into configure.in that adds in code
4165 to set GCC=yes if CC=icc. This seems to end up at a point in the
4166 generated configure script that is early enough to affect the setting of
4167 compiler options, which is what is needed, but I have no means of testing
4168 whether it really works. (The user who reported this had patched the
4169 generated configure script, which of course I cannot do.)
4171 LATER: After change 22 below (new libtool files), the configure script
4172 seems to know about icc (and also ecc). Therefore, I have commented out
4173 this hack in configure.in.
4175 7. Added support for pkg-config (2 patches were sent in).
4177 8. Negated POSIX character classes that used a combination of internal tables
4178 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
4179 [[:^ascii]]. Typically, they would match almost any characters. The other
4180 POSIX classes were not broken in this way.
4182 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
4183 to find the match, as PCRE was deluded into thinking that the match had to
4184 start at the start point or following a newline. The same bug applied to
4185 patterns with negative forward assertions or any backward assertions
4186 preceding ".*" at the start, unless the pattern required a fixed first
4187 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
4189 10. In UTF-8 mode, when moving forwards in the subject after a failed match
4190 starting at the last subject character, bytes beyond the end of the subject
4191 string were read.
4193 11. Renamed the variable "class" as "classbits" to make life easier for C++
4194 users. (Previously there was a macro definition, but it apparently wasn't
4195 enough.)
4197 12. Added the new field "tables" to the extra data so that tables can be passed
4198 in at exec time, or the internal tables can be re-selected. This allows
4199 a compiled regex to be saved and re-used at a later time by a different
4200 program that might have everything at different addresses.
4202 13. Modified the pcre-config script so that, when run on Solaris, it shows a
4203 -R library as well as a -L library.
4205 14. The debugging options of pcretest (-d on the command line or D on a
4206 pattern) showed incorrect output for anything following an extended class
4207 that contained multibyte characters and which was followed by a quantifier.
4209 15. Added optional support for general category Unicode character properties
4210 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
4211 support. It adds about 90K to the size of the library. The meanings of the
4212 inbuilt class escapes such as \d and \s have NOT been changed.
4214 16. Updated pcredemo.c to include calls to free() to release the memory for the
4215 compiled pattern.
4217 17. The generated file chartables.c was being created in the source directory
4218 instead of in the building directory. This caused the build to fail if the
4219 source directory was different from the building directory, and was
4220 read-only.
4222 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
4223 file. No doubt somebody will tell me if they don't make sense... Also added
4224 Dan Mooney's comments about building on OpenVMS.
4226 19. Added support for partial matching via the PCRE_PARTIAL option for
4227 pcre_exec() and the \P data escape in pcretest.
4229 20. Extended pcretest with 3 new pattern features:
4231 (i) A pattern option of the form ">rest-of-line" causes pcretest to
4232 write the compiled pattern to the file whose name is "rest-of-line".
4233 This is a straight binary dump of the data, with the saved pointer to
4234 the character tables forced to be NULL. The study data, if any, is
4235 written too. After writing, pcretest reads a new pattern.
4237 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
4238 compiled pattern from the given file. There must not be any
4239 occurrences of "<" in the file name (pretty unlikely); if there are,
4240 pcretest will instead treat the initial "<" as a pattern delimiter.
4241 After reading in the pattern, pcretest goes on to read data lines as
4242 usual.
4244 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
4245 and 16-bit fields in a compiled pattern, to simulate a pattern that
4246 was compiled on a host of opposite endianness.
4248 21. The pcre-exec() function can now cope with patterns that were compiled on
4249 hosts of opposite endianness, with this restriction:
4251 As for any compiled expression that is saved and used later, the tables
4252 pointer field cannot be preserved; the extra_data field in the arguments
4253 to pcre_exec() should be used to pass in a tables address if a value
4254 other than the default internal tables were used at compile time.
4256 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
4257 now diagnosed as an error. Previously, most of the time, a negative number
4258 would have been treated as zero, but if in addition "ovector" was passed as
4259 NULL, a crash could occur.
4261 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
4262 new versions from the libtool 1.5 distribution (the last one is a copy of
4263 a file called libtool.m4). This seems to have fixed the need to patch
4264 "configure" to support Darwin 1.3 (which I used to do). However, I still
4265 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
4266 workstation).
4268 24. Changed the PCRE licence to be the more standard "BSD" licence.
4271 Version 4.5 01-Dec-03
4272 ---------------------
4274 1. There has been some re-arrangement of the code for the match() function so
4275 that it can be compiled in a version that does not call itself recursively.
4276 Instead, it keeps those local variables that need separate instances for
4277 each "recursion" in a frame on the heap, and gets/frees frames whenever it
4278 needs to "recurse". Keeping track of where control must go is done by means
4279 of setjmp/longjmp. The whole thing is implemented by a set of macros that
4280 hide most of the details from the main code, and operates only if
4281 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
4282 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
4283 operating.
4285 To make it easier for callers to provide specially tailored get/free
4286 functions for this usage, two new functions, pcre_stack_malloc, and
4287 pcre_stack_free, are used. They are always called in strict stacking order,
4288 and the size of block requested is always the same.
4290 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
4291 PCRE has been compiled to use the stack or the heap for recursion. The
4292 -C option of pcretest uses this to show which version is compiled.
4294 A new data escape \S, is added to pcretest; it causes the amounts of store
4295 obtained and freed by both kinds of malloc/free at match time to be added
4296 to the output.
4298 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
4299 what's available on my current Linux desktop machine.
4301 3. When matching a UTF-8 string, the test for a valid string at the start has
4302 been extended. If start_offset is not zero, PCRE now checks that it points
4303 to a byte that is the start of a UTF-8 character. If not, it returns
4304 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
4305 this is necessary because there may be backward assertions in the pattern.
4306 When matching the same subject several times, it may save resources to use
4307 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
4309 4. The code for checking the validity of UTF-8 strings has been tightened so
4310 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
4311 containing "overlong sequences".
4313 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
4314 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
4315 should have been "&", but it just so happened that all the cases this let
4316 through by mistake were picked up later in the function.
4318 6. I had used a variable called "isblank" - this is a C99 function, causing
4319 some compilers to warn. To avoid this, I renamed it (as "blankclass").
4321 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
4322 prompting; (b) run "./pcretest /dev/null" at the start of the test script
4323 so the version is shown; (c) stop "make test" echoing "./RunTest".
4325 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
4327 9. The prototype for memmove() for systems that don't have it was using
4328 size_t, but the inclusion of the header that defines size_t was later. I've
4329 moved the #includes for the C headers earlier to avoid this.
4331 10. Added some adjustments to the code to make it easier to compiler on certain
4332 special systems:
4334 (a) Some "const" qualifiers were missing.
4335 (b) Added the macro EXPORT before all exported functions; by default this
4336 is defined to be empty.
4337 (c) Changed the dftables auxiliary program (that builds chartables.c) so
4338 that it reads its output file name as an argument instead of writing
4339 to the standard output and assuming this can be redirected.
4341 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
4342 class containing characters with values greater than 255, PCRE compilation
4343 went into a loop.
4345 12. A recursive reference to a subpattern that was within another subpattern
4346 that had a minimum quantifier of zero caused PCRE to crash. For example,
4347 (x(y(?2))z)? provoked this bug with a subject that got as far as the
4348 recursion. If the recursively-called subpattern itself had a zero repeat,
4349 that was OK.
4351 13. In pcretest, the buffer for reading a data line was set at 30K, but the
4352 buffer into which it was copied (for escape processing) was still set at
4353 1024, so long lines caused crashes.
4355 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
4356 "internal error: code overflow...". This applied to any character class
4357 that was followed by a possessive quantifier.
4359 15. Modified the Makefile to add libpcre.la as a prerequisite for
4360 libpcreposix.la because I was told this is needed for a parallel build to
4361 work.
4363 16. If a pattern that contained .* following optional items at the start was
4364 studied, the wrong optimizing data was generated, leading to matching
4365 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
4366 matching string must start with a or b or c. The correct conclusion for
4367 this pattern is that a match can start with any character.
4370 Version 4.4 13-Aug-03
4371 ---------------------
4373 1. In UTF-8 mode, a character class containing characters with values between
4374 127 and 255 was not handled correctly if the compiled pattern was studied.
4375 In fixing this, I have also improved the studying algorithm for such
4376 classes (slightly).
4378 2. Three internal functions had redundant arguments passed to them. Removal
4379 might give a very teeny performance improvement.
4381 3. Documentation bug: the value of the capture_top field in a callout is *one
4382 more than* the number of the hightest numbered captured substring.
4384 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
4385 in incorrectly linking with a previously installed version. They now link
4386 explicitly with libpcre.la.
4388 5. configure.in no longer needs to recognize Cygwin specially.
4390 6. A problem in pcre.in for Windows platforms is fixed.
4392 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
4393 pcretest, it used to include the size of the study block as part of its
4394 output. Unfortunately, the structure contains a field that has a different
4395 size on different hardware architectures. This meant that the tests that
4396 showed this size failed. As the block is currently always of a fixed size,
4397 this information isn't actually particularly useful in pcretest output, so
4398 I have just removed it.
4400 8. Three pre-processor statements accidentally did not start in column 1.
4401 Sadly, there are *still* compilers around that complain, even though
4402 standard C has not required this for well over a decade. Sigh.
4404 9. In pcretest, the code for checking callouts passed small integers in the
4405 callout_data field, which is a void * field. However, some picky compilers
4406 complained about the casts involved for this on 64-bit systems. Now
4407 pcretest passes the address of the small integer instead, which should get
4408 rid of the warnings.
4410 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
4411 both compile and run time, and gives an error if an invalid UTF-8 sequence
4412 is found. There is a option for disabling this check in cases where the
4413 string is known to be correct and/or the maximum performance is wanted.
4415 11. In response to a bug report, I changed one line in Makefile.in from
4417 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
4418 to
4419 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
4421 to look similar to other lines, but I have no way of telling whether this
4422 is the right thing to do, as I do not use Windows. No doubt I'll get told
4423 if it's wrong...
4426 Version 4.3 21-May-03
4427 ---------------------
4429 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
4430 Makefile.
4432 2. Some refactoring to improve the quality of the code:
4434 (i) The utf8_table... variables are now declared "const".
4436 (ii) The code for \cx, which used the "case flipping" table to upper case
4437 lower case letters, now just substracts 32. This is ASCII-specific,
4438 but the whole concept of \cx is ASCII-specific, so it seems
4439 reasonable.
4441 (iii) PCRE was using its character types table to recognize decimal and
4442 hexadecimal digits in the pattern. This is silly, because it handles
4443 only 0-9, a-f, and A-F, but the character types table is locale-
4444 specific, which means strange things might happen. A private
4445 table is now used for this - though it costs 256 bytes, a table is
4446 much faster than multiple explicit tests. Of course, the standard
4447 character types table is still used for matching digits in subject
4448 strings against \d.
4450 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
4451 ending in _t are). So I've renamed it as ESC_tee.
4453 3. The first argument for regexec() in the POSIX wrapper should have been
4454 defined as "const".
4456 4. Changed pcretest to use malloc() for its buffers so that they can be
4457 Electric Fenced for debugging.
4459 5. There were several places in the code where, in UTF-8 mode, PCRE would try
4460 to read one or more bytes before the start of the subject string. Often this
4461 had no effect on PCRE's behaviour, but in some circumstances it could
4462 provoke a segmentation fault.
4464 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
4465 to try to read one or more bytes before the start of the subject string.
4467 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
4468 UTF-8 support could misbehave in various ways if the subject string
4469 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
4470 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
4471 back over UTF-8 characters.)
4474 Version 4.2 14-Apr-03
4475 ---------------------
4477 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
4479 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
4480 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
4481 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
4482 [WIN_PREFIX]: new variable, "cyg" for Cygwin
4483 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
4485 Note: automatic setting of the BUILD variables is not yet working
4486 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
4487 compile-time but not at link-time
4488 [LINK]: use for linking executables only
4489 make different versions for Windows and non-Windows
4490 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
4491 libraries
4492 [LINK_FOR_BUILD]: new variable
4493 [OBJEXT]: use throughout
4494 [EXEEXT]: use throughout
4495 <winshared>: new target
4496 <wininstall>: new target
4497 <dftables.o>: use native compiler
4498 <dftables>: use native linker
4499 <install>: handle Windows platform correctly
4500 <clean>: ditto
4501 <check>: ditto
4502 copy DLL to top builddir before testing
4504 As part of these changes, -no-undefined was removed again. This was reported
4505 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
4506 in any case.
4508 3. Some tidies to get rid of compiler warnings:
4510 . In the match_data structure, match_limit was an unsigned long int, whereas
4511 match_call_count was an int. I've made them both unsigned long ints.
4513 . In pcretest the fact that a const uschar * doesn't automatically cast to
4514 a void * provoked a warning.
4516 . Turning on some more compiler warnings threw up some "shadow" variables
4517 and a few more missing casts.
4519 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
4520 option, a class that contained a single character with a value between 128
4521 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
4523 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
4524 option, a class that contained several characters, but with at least one
4525 whose value was between 128 and 255 caused PCRE to crash.
4528 Version 4.1 12-Mar-03
4529 ---------------------
4531 1. Compiling with gcc -pedantic found a couple of places where casts were
4532 needed, and a string in dftables.c that was longer than standard compilers are
4533 required to support.
4535 2. Compiling with Sun's compiler found a few more places where the code could
4536 be tidied up in order to avoid warnings.
4538 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
4539 first of these names is deprecated in the latest Autoconf in favour of the name
4540 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
4541 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
4542 analogy I have changed it to CFLAGS_FOR_BUILD.
4544 4. Added -no-undefined to the linking command in the Makefile, because this is
4545 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
4546 linking step for the pcreposix library.
4548 5. PCRE was failing to diagnose the case of two named groups with the same
4549 name.
4551 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
4552 literal character that is needed in the subject for a match, and scans along to
4553 ensure that it is present before embarking on the full matching process. This
4554 saves time in cases of nested unlimited repeats that are never going to match.
4555 Problem: the scan can take a lot of time if the subject is very long (e.g.
4556 megabytes), thus penalizing straightforward matches. It is now done only if the
4557 amount of subject to be scanned is less than 1000 bytes.
4559 7. A lesser problem with the same optimization is that it was recording the
4560 first character of an anchored pattern as "needed", thus provoking a search
4561 right along the subject, even when the first match of the pattern was going to
4562 fail. The "needed" character is now not set for anchored patterns, unless it
4563 follows something in the pattern that is of non-fixed length. Thus, it still
4564 fulfils its original purpose of finding quick non-matches in cases of nested
4565 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
4568 Version 4.0 17-Feb-03
4569 ---------------------
4571 1. If a comment in an extended regex that started immediately after a meta-item
4572 extended to the end of string, PCRE compiled incorrect data. This could lead to
4573 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
4575 2. Moved to autoconf 2.53 and libtool 1.4.2.
4577 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
4578 the special perltest8 script is no longer needed - all the tests can be run
4579 from a single perltest script.
4581 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
4582 by \s. It has now been removed in PCRE. This means it isn't recognized as
4583 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
4584 class [:space:] *does* include VT, thereby creating a mess.
4586 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
4587 space and tab.
4589 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
4590 its new features into the main test script, reducing the number of scripts.
4592 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
4593 were backward compatible, and made the (?i) apply to the whole pattern, as if
4594 /i were given. Now it behaves more logically, and applies the option setting
4595 only to what follows. PCRE has been changed to follow suit. However, if it
4596 finds options settings right at the start of the pattern, it extracts them into
4597 the global options, as before. Thus, they show up in the info data.
4599 8. Added support for the \Q...\E escape sequence. Characters in between are
4600 treated as literals. This is slightly different from Perl in that $ and @ are
4601 also handled as literals inside the quotes. In Perl, they will cause variable
4602 interpolation. Note the following examples:
4604 Pattern PCRE matches Perl matches
4606 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
4607 \Qabc\$xyz\E abc\$xyz abc\$xyz
4608 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
4610 For compatibility with Perl, \Q...\E sequences are recognized inside character
4611 classes as well as outside them.
4613 9. Re-organized 3 code statements in pcretest to avoid "overflow in
4614 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
4615 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
4616 signed/unsigned warnings.
4618 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
4619 option for pcretest, so I've replaced it by a simple function that does just
4620 that job.
4622 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
4623 "pcregrep -".
4625 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
4626 Java package. This provides some syntactic sugar for simple cases of what my
4627 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
4628 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
4629 item, you can use this simplified notation. Note that only makes sense with
4630 greedy quantifiers. Consequently, the use of the possessive quantifier forces
4631 greediness, whatever the setting of the PCRE_UNGREEDY option.
4633 13. A change of greediness default within a pattern was not taking effect at
4634 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
4635 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
4636 was abstracted outside.
4638 14. PCRE now supports the \G assertion. It is true when the current matching
4639 position is at the start point of the match. This differs from \A when the
4640 starting offset is non-zero. Used with the /g option of pcretest (or similar
4641 code), it works in the same way as it does for Perl's /g option. If all
4642 alternatives of a regex begin with \G, the expression is anchored to the start
4643 match position, and the "anchored" flag is set in the compiled expression.
4645 15. Some bugs concerning the handling of certain option changes within patterns
4646 have been fixed. These applied to options other than (?ims). For example,
4647 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
4648 been the other way round. Some of this was related to change 7 above.
4650 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
4651 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
4652 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
4653 POSIX classes only within a class (e.g. /[[:alpha:]]/).
4655 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
4656 mode. Unlike ".", it always matches newline, whatever the setting of
4657 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
4658 assertions. Perl allows it, but it doesn't (in general) work because it can't
4659 calculate the length of the lookbehind. At least, that's the case for Perl
4660 5.8.0 - I've been told they are going to document that it doesn't work in
4661 future.
4663 18. Added an error diagnosis for escapes that PCRE does not support: these are
4664 \L, \l, \N, \P, \p, \U, \u, and \X.
4666 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4667 reading past the end of the pattern in cases such as /[abcd/.
4669 20. PCRE was getting more memory than necessary for patterns with classes that
4670 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4672 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4673 compiling PCRE for use with Virtual Pascal.
4675 22. Small fix to the Makefile to make it work properly if the build is done
4676 outside the source tree.
4678 23. Added a new extension: a condition to go with recursion. If a conditional
4679 subpattern starts with (?(R) the "true" branch is used if recursion has
4680 happened, whereas the "false" branch is used only at the top level.
4682 24. When there was a very long string of literal characters (over 255 bytes
4683 without UTF support, over 250 bytes with UTF support), the computation of how
4684 much memory was required could be incorrect, leading to segfaults or other
4685 strange effects.
4687 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4688 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4689 there was a subsequent back reference to those brackets. This meant that, for
4690 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4691 possible to check for precisely this case. All we can do is abandon the
4692 optimization if .* occurs inside capturing brackets when there are any back
4693 references whatsoever. (See below for a better fix that came later.)
4695 26. The handling of the optimization for finding the first character of a
4696 non-anchored pattern, and for finding a character that is required later in the
4697 match were failing in some cases. This didn't break the matching; it just
4698 failed to optimize when it could. The way this is done has been re-implemented.
4700 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4702 28. Added a new feature that provides some of the functionality that Perl
4703 provides with (?{...}). The facility is termed a "callout". The way it is done
4704 in PCRE is for the caller to provide an optional function, by setting
4705 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4706 global variable. By default it is unset, which disables all calling out. To get
4707 the function called, the regex must include (?C) at appropriate points. This
4708 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4709 This provides a means of identifying different callout points. When PCRE
4710 reaches such a point in the regex, if pcre_callout has been set, the external
4711 function is called. It is provided with data in a structure called
4712 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4713 matching continues; if it returns a non-zero value, the match at the current
4714 point fails. However, backtracking will occur if possible. [This was changed
4715 later and other features added - see item 49 below.]
4717 29. pcretest is upgraded to test the callout functionality. It provides a
4718 callout function that displays information. By default, it shows the start of
4719 the match and the current position in the text. There are some new data escapes
4720 to vary what happens:
4722 \C+ in addition, show current contents of captured substrings
4723 \C- do not supply a callout function
4724 \C!n return 1 when callout number n is reached
4725 \C!n!m return 1 when callout number n is reached for the mth time
4727 30. If pcregrep was called with the -l option and just a single file name, it
4728 output "<stdin>" if a match was found, instead of the file name.
4730 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4731 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4732 pcre_exec(). This saves a malloc/free per call. The default value of
4733 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4734 when configuring.
4736 32. The default maximum size of a compiled pattern is 64K. There have been a
4737 few cases of people hitting this limit. The code now uses macros to handle the
4738 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4739 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4740 configuring. Tests 2 and 5 work only with 2-byte links because they output
4741 debugging information about compiled patterns.
4743 33. Internal code re-arrangements:
4745 (a) Moved the debugging function for printing out a compiled regex into
4746 its own source file (printint.c) and used #include to pull it into
4747 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4748 separate copies.
4750 (b) Defined the list of op-code names for debugging as a macro in
4751 internal.h so that it is next to the definition of the opcodes.
4753 (c) Defined a table of op-code lengths for simpler skipping along compiled
4754 code. This is again a macro in internal.h so that it is next to the
4755 definition of the opcodes.
4757 34. Added support for recursive calls to individual subpatterns, along the
4758 lines of Robin Houston's patch (but implemented somewhat differently).
4760 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4761 allow it to read and process whole directories in Win32. This code was
4762 contributed by Lionel Fourquaux; it has not been tested by me.
4764 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4765 used to name a group. Names consist of alphanumerics and underscores, and must
4766 be unique. Back references use the syntax (?P=name) and recursive calls use
4767 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4768 numbers. The function pcre_fullinfo() can be used after compilation to extract
4769 a name/number map. There are three relevant calls:
4771 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4772 PCRE_INFO_NAMECOUNT yields the number of entries
4773 PCRE_INFO_NAMETABLE yields a pointer to the map.
4775 The map is a vector of fixed-size entries. The size of each entry depends on
4776 the length of the longest name used. The first two bytes of each entry are the
4777 group number, most significant byte first. There follows the corresponding
4778 name, zero terminated. The names are in alphabetical order.
4780 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4781 case instead of 255. Making it the same both with and without UTF-8 support
4782 means that the same test output works with both.
4784 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4785 calling malloc() with a zero argument.
4787 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4788 optimization. I've improved things by keeping a bitmap of backreferences with
4789 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4790 fact referenced, the optimization can be applied. It is unlikely that a
4791 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4792 the match to follow \n) will appear inside brackets with a number greater than
4793 31, but if it does, any back reference > 31 suppresses the optimization.
4795 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4796 of disabling numbered capturing parentheses. Any opening parenthesis that is
4797 not followed by ? behaves as if it were followed by ?: but named parentheses
4798 can still be used for capturing (and they will acquire numbers in the usual
4799 way).
4801 41. Redesigned the return codes from the match() function into yes/no/error so
4802 that errors can be passed back from deep inside the nested calls. A malloc
4803 failure while inside a recursive subpattern call now causes the
4804 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4806 42. It is now possible to set a limit on the number of times the match()
4807 function is called in a call to pcre_exec(). This facility makes it possible to
4808 limit the amount of recursion and backtracking, though not in a directly
4809 obvious way, because the match() function is used in a number of different
4810 circumstances. The count starts from zero for each position in the subject
4811 string (for non-anchored patterns). The default limit is, for compatibility, a
4812 large number, namely 10 000 000. You can change this in two ways:
4814 (a) When configuring PCRE before making, you can use --with-match-limit=n
4815 to set a default value for the compiled library.
4817 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4818 a different value is set. See 45 below.
4820 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4822 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4823 of things that can be changed at compile time. The first argument specifies
4824 what is wanted and the second points to where the information is to be placed.
4825 The current list of available information is:
4829 The output is an integer that is set to one if UTF-8 support is available;
4830 otherwise it is set to zero.
4834 The output is an integer that it set to the value of the code that is used for
4835 newline. It is either LF (10) or CR (13).
4839 The output is an integer that contains the number of bytes used for internal
4840 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4844 The output is an integer that contains the threshold above which the POSIX
4845 interface uses malloc() for output vectors. See item 31 above.
4849 The output is an unsigned integer that contains the default limit of the number
4850 of match() calls in a pcre_exec() execution. See 42 above.
4852 44. pcretest has been upgraded by the addition of the -C option. This causes it
4853 to extract all the available output from the new pcre_config() function, and to
4854 output it. The program then exits immediately.
4856 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4857 order to support additional features. One way would have been to define
4858 pcre_exec2() (for example) with extra arguments, but this would not have been
4859 extensible, and would also have required all calls to the original function to
4860 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4861 is used for passing in "extra" data from pcre_study().
4863 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4864 contains the following fields:
4866 flags a bitmap indicating which of the following fields are set
4867 study_data opaque data from pcre_study()
4868 match_limit a way of specifying a limit on match() calls for a specific
4869 call to pcre_exec()
4870 callout_data data for callouts (see 49 below)
4872 The flag bits are also defined in pcre.h, and are
4878 The pcre_study() function now returns one of these new pcre_extra blocks, with
4879 the actual study data pointed to by the study_data field, and the
4880 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4881 before. That is, this change is entirely upwards-compatible and requires no
4882 change to existing code.
4884 If you want to pass in additional data to pcre_exec(), you can either place it
4885 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4886 block.
4888 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4889 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4890 times with different match limits, until it finds the minimum value needed for
4891 pcre_exec() to complete. The value is then output. This can be instructive; for
4892 most simple matches the number is quite small, but for pathological cases it
4893 gets very large very quickly.
4895 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4896 returns the size of the data block pointed to by the study_data field in a
4897 pcre_extra block, that is, the value that was passed as the argument to
4898 pcre_malloc() when PCRE was getting memory in which to place the information
4899 created by pcre_study(). The fourth argument should point to a size_t variable.
4900 pcretest has been extended so that this information is shown after a successful
4901 pcre_study() call when information about the compiled regex is being displayed.
4903 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4904 because what follows is always an absolute path. (Later: it turns out that this
4905 is more than cosmetic for MinGW, because it doesn't like empty path
4906 components.)
4908 49. Some changes have been made to the callout feature (see 28 above):
4910 (i) A callout function now has three choices for what it returns:
4912 0 => success, carry on matching
4913 > 0 => failure at this point, but backtrack if possible
4914 < 0 => serious error, return this value from pcre_exec()
4916 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4917 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4918 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4919 use by callout functions. It will never be used by PCRE itself.
4921 (ii) The pcre_extra structure (see 45 above) has a void * field called
4922 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4923 pcre_callout_block structure has a field of the same name. The contents of
4924 the field passed in the pcre_extra structure are passed to the callout
4925 function in the corresponding field in the callout block. This makes it
4926 easier to use the same callout-containing regex from multiple threads. For
4927 testing, the pcretest program has a new data escape
4929 \C*n pass the number n (may be negative) as callout_data
4931 If the callout function in pcretest receives a non-zero value as
4932 callout_data, it returns that value.
4934 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4935 there were some redundant $(CFLAGS) in commands that are now specified as
4936 $(LINK), which already includes $(CFLAGS).
4938 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4939 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4940 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4941 one-byte characters throughout. Note that case-insensitive matching applies
4942 only to characters whose values are less than 256. PCRE doesn't support the
4943 notion of cases for higher-valued characters.
4945 (i) A character class whose characters are all within 0-255 is handled as
4946 a bit map, and the map is inverted for negative classes. Previously, a
4947 character > 255 always failed to match such a class; however it should
4948 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4950 (ii) A negated character class with a single character < 255 is coded as
4951 "not this character" (OP_NOT). This wasn't working properly when the test
4952 character was multibyte, either singly or repeated.
4954 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4955 mode, for example: \x{100}{2,3}.
4957 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4958 singly or repeated) now correctly test multibyte characters. However,
4959 PCRE doesn't recognize any characters with values greater than 255 as
4960 digits, spaces, or word characters. Such characters always match \D, \S,
4961 and \W, and never match \d, \s, or \w.
4963 (v) Classes may now contain characters and character ranges with values
4964 greater than 255. For example: [ab\x{100}-\x{400}].
4966 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4967 PCRE in UTF-8 mode.
4969 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4970 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4971 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4972 value.)
4974 53. The single man page has become too large. I have therefore split it up into
4975 a number of separate man pages. These also give rise to individual HTML pages;
4976 these are now put in a separate directory, and there is an index.html page that
4977 lists them all. Some hyperlinking between the pages has been installed.
4979 54. Added convenience functions for handling named capturing parentheses.
4981 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4982 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4983 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4984 are faulted.
4986 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4987 calling configure. These values are used when compiling the dftables.c program
4988 which is run to generate the source of the default character tables. They
4989 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4990 you will need to set these values.
4992 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4995 Version 3.9 02-Jan-02
4996 ---------------------
4998 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
5000 2. If --disable-static was given, the building process failed when trying to
5001 build pcretest and pcregrep. (For some reason it was using libtool to compile
5002 them, which is not right, as they aren't part of the library.)
5005 Version 3.8 18-Dec-01
5006 ---------------------
5008 1. The experimental UTF-8 code was completely screwed up. It was packing the
5009 bytes in the wrong order. How dumb can you get?
5012 Version 3.7 29-Oct-01
5013 ---------------------
5015 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
5016 This caused pcretest, when used on the test data, to segfault. Unfortunately,
5017 this didn't happen under Solaris 8, where I normally test things.
5019 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
5020 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
5021 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
5022 this fix an hour or so after the initial 3.7 release.)
5025 Version 3.6 23-Oct-01
5026 ---------------------
5028 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
5029 offsets passed as NULL with zero offset count.
5031 2. The config.guess and config.sub files had not been updated when I moved to
5032 the latest autoconf.
5035 Version 3.5 15-Aug-01
5036 ---------------------
5038 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
5039 had been forgotten.
5041 2. By using declared but undefined structures, we can avoid using "void"
5042 definitions in pcre.h while keeping the internal definitions of the structures
5043 private.
5045 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
5046 user point of view, this means that both static and shared libraries are built
5047 by default, but this can be individually controlled. More of the work of
5048 handling this static/shared cases is now inside libtool instead of PCRE's make
5049 file.
5051 4. The pcretest utility is now installed along with pcregrep because it is
5052 useful for users (to test regexs) and by doing this, it automatically gets
5053 relinked by libtool. The documentation has been turned into a man page, so
5054 there are now .1, .txt, and .html versions in /doc.
5056 5. Upgrades to pcregrep:
5057 (i) Added long-form option names like gnu grep.
5058 (ii) Added --help to list all options with an explanatory phrase.
5059 (iii) Added -r, --recursive to recurse into sub-directories.
5060 (iv) Added -f, --file to read patterns from a file.
5062 6. pcre_exec() was referring to its "code" argument before testing that
5063 argument for NULL (and giving an error if it was NULL).
5065 7. Upgraded Makefile.in to allow for compiling in a different directory from
5066 the source directory.
5068 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
5069 options bits, the pointer it was passed was to an int instead of to an unsigned
5070 long int. This mattered only on 64-bit systems.
5072 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
5073 generated) instead of pcre.in, which it its source. Also made the same change
5074 in several of the .c files.
5076 10. A new release of gcc defines printf() as a macro, which broke pcretest
5077 because it had an ifdef in the middle of a string argument for printf(). Fixed
5078 by using separate calls to printf().
5080 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
5081 script, to force use of CR or LF instead of \n in the source. On non-Unix
5082 systems, the value can be set in config.h.
5084 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
5085 absolute limit. Changed the text of the error message to make this clear, and
5086 likewise updated the man page.
5088 13. The limit of 99 on the number of capturing subpatterns has been removed.
5089 The new limit is 65535, which I hope will not be a "real" limit.
5092 Version 3.4 22-Aug-00
5093 ---------------------
5095 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
5097 2. Diagnose condition (?(0) as an error instead of crashing on matching.
5100 Version 3.3 01-Aug-00
5101 ---------------------
5103 1. If an octal character was given, but the value was greater than \377, it
5104 was not getting masked to the least significant bits, as documented. This could
5105 lead to crashes in some systems.
5107 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
5108 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
5110 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
5111 These just pass their arguments on to (pcre_free)(), but they are provided
5112 because some uses of PCRE bind it to non-C systems that can call its functions,
5113 but cannot call free() or pcre_free() directly.
5115 4. Add "make test" as a synonym for "make check". Corrected some comments in
5116 the Makefile.
5118 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
5119 Makefile.
5121 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
5122 command called pgrep for grepping around the active processes.
5124 7. Added the beginnings of support for UTF-8 character strings.
5126 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
5127 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
5128 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
5129 out for the ar command.)
5132 Version 3.2 12-May-00
5133 ---------------------
5135 This is purely a bug fixing release.
5137 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
5138 of ZA. This was just one example of several cases that could provoke this bug,
5139 which was introduced by change 9 of version 2.00. The code for breaking
5140 infinite loops after an iteration that matches an empty string was't working
5141 correctly.
5143 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
5144 when matched against abbab (for example). After matching an empty string, it
5145 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
5146 caused it to match further down the string than it should.
5148 3. The code contained an inclusion of sys/types.h. It isn't clear why this
5149 was there because it doesn't seem to be needed, and it causes trouble on some
5150 systems, as it is not a Standard C header. It has been removed.
5152 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
5153 were reported on the Macintosh. The changes were from
5155 while ((c = *(++ptr)) != 0 && c != '\n');
5156 to
5157 while ((c = *(++ptr)) != 0 && c != '\n') ;
5159 Totally extraordinary, but if that's what it takes...
5161 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
5162 available. Added HAVE_BCOPY and an autoconf test for it; if neither
5163 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
5164 assumes the way PCRE uses memmove() (always moving upwards).
5166 6. PCRE is being used in one environment where strchr() is not available. There
5167 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
5168 faster code anyway.
5171 Version 3.1 09-Feb-00
5172 ---------------------
5174 The only change in this release is the fixing of some bugs in Makefile.in for
5175 the "install" target:
5177 (1) It was failing to install pcreposix.h.
5179 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
5182 Version 3.0 01-Feb-00
5183 ---------------------
5185 1. Add support for the /+ modifier to perltest (to output $` like it does in
5186 pcretest).
5188 2. Add support for the /g modifier to perltest.
5190 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
5191 matches null strings.
5193 4. Fix perltest so that it doesn't do unwanted things when fed an empty
5194 pattern. Perl treats empty patterns specially - it reuses the most recent
5195 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
5196 effect.
5198 5. The POSIX interface was broken in that it was just handing over the POSIX
5199 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
5200 required a bigger vector, with some working space on the end. This means that
5201 the POSIX wrapper now has to get and free some memory, and copy the results.
5203 6. Added some simple autoconf support, placing the test data and the
5204 documentation in separate directories, re-organizing some of the
5205 information files, and making it build pcre-config (a GNU standard). Also added
5206 libtool support for building PCRE as a shared library, which is now the
5207 default.
5209 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
5210 09 are not valid octal constants. Single digits will be used for minor values
5211 less than 10.
5213 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
5214 existing programs that set these in the POSIX interface can use PCRE without
5215 modification.
5217 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
5218 return all that pcre_info() returns, plus additional data. The pcre_info()
5219 function is retained for compatibility, but is considered to be obsolete.
5221 10. Added experimental recursion feature (?R) to handle one common case that
5222 Perl 5.6 will be able to do with (?p{...}).
5224 11. Added support for POSIX character classes like [:alpha:], which Perl is
5225 adopting.
5228 Version 2.08 31-Aug-99
5229 ----------------------
5231 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
5232 trying to match at the startoffset position, but instead was moving forward to
5233 the next newline as if a previous match had failed.
5235 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
5236 and could get into a loop if a null string was matched other than at the start
5237 of the subject.
5239 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
5240 be distinguished at compile time, and for completeness also added PCRE_DATE.
5242 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
5243 in GnuWin32 environments.
5246 Version 2.07 29-Jul-99
5247 ----------------------
5249 1. The documentation is now supplied in plain text form and HTML as well as in
5250 the form of man page sources.
5252 2. C++ compilers don't like assigning (void *) values to other pointer types.
5253 In particular this affects malloc(). Although there is no problem in Standard
5254 C, I've put in casts to keep C++ compilers happy.
5256 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
5257 should be (const char *).
5259 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
5260 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
5261 However, I haven't made this a standard facility. The documentation doesn't
5262 mention it, and the Makefile doesn't support it.
5264 5. The Makefile now contains an "install" target, with editable destinations at
5265 the top of the file. The pcretest program is not installed.
5267 6. pgrep -V now gives the PCRE version number and date.
5269 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
5270 causing the entire string to be ignored, instead of just the last character.
5272 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
5273 non-matching string, it can take a very, very long time, even for strings of
5274 quite modest length, because of the nested recursion. PCRE now does better in
5275 some of these cases. It does this by remembering the last required literal
5276 character in the pattern, and pre-searching the subject to ensure it is present
5277 before running the real match. In other words, it applies a heuristic to detect
5278 some types of certain failure quickly, and in the above example, if presented
5279 with a string that has no trailing " it gives "no match" very quickly.
5281 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
5282 other alternatives are tried instead.
5285 Version 2.06 09-Jun-99
5286 ----------------------
5288 1. Change pcretest's output for amount of store used to show just the code
5289 space, because the remainder (the data block) varies in size between 32-bit and
5290 64-bit systems.
5292 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
5293 start matching at. This allows lookbehinds to work when searching for multiple
5294 occurrences in a string.
5296 3. Added additional options to pcretest for testing multiple occurrences:
5298 /+ outputs the rest of the string that follows a match
5299 /g loops for multiple occurrences, using the new startoffset argument
5300 /G loops for multiple occurrences by passing an incremented pointer
5302 4. PCRE wasn't doing the "first character" optimization for patterns starting
5303 with \b or \B, though it was doing it for other lookbehind assertions. That is,
5304 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
5305 the letter 'x'. On long subject strings, this gives a significant speed-up.
5308 Version 2.05 21-Apr-99
5309 ----------------------
5311 1. Changed the type of magic_number from int to long int so that it works
5312 properly on 16-bit systems.
5314 2. Fixed a bug which caused patterns starting with .* not to work correctly
5315 when the subject string contained newline characters. PCRE was assuming
5316 anchoring for such patterns in all cases, which is not correct because .* will
5317 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
5318 DOTALL is set at top level; otherwise it knows that patterns starting with .*
5319 must be retried after every newline in the subject.
5322 Version 2.04 18-Feb-99
5323 ----------------------
5325 1. For parenthesized subpatterns with repeats whose minimum was zero, the
5326 computation of the store needed to hold the pattern was incorrect (too large).
5327 If such patterns were nested a few deep, this could multiply and become a real
5328 problem.
5330 2. Added /M option to pcretest to show the memory requirement of a specific
5331 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
5333 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
5334 compiled in such a way that the backtracking after subsequent failure was
5335 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
5336 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
5339 Version 2.03 02-Feb-99
5340 ----------------------
5342 1. Fixed typo and small mistake in man page.
5344 2. Added 4th condition (GPL supersedes if conflict) and created separate
5345 LICENCE file containing the conditions.
5347 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
5348 Perl, that is the internal \ allows the delimiter to be included in the
5349 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
5350 the final delimiter, add \ to the end of the pattern (to test the error).
5352 4. Added the convenience functions for extracting substrings after a successful
5353 match. Updated pcretest to make it able to test these functions.
5356 Version 2.02 14-Jan-99
5357 ----------------------
5359 1. Initialized the working variables associated with each extraction so that
5360 their saving and restoring doesn't refer to uninitialized store.
5362 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
5363 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
5364 fix the problem.
5366 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
5367 calls, and wasn't printing the correct value for compiling calls. Increased the
5368 default value of LOOPREPEAT, and the number of significant figures in the
5369 times.
5371 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
5373 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
5374 a building problem on Windows NT with a FAT file system.
5377 Version 2.01 21-Oct-98
5378 ----------------------
5380 1. Changed the API for pcre_compile() to allow for the provision of a pointer
5381 to character tables built by pcre_maketables() in the current locale. If NULL
5382 is passed, the default tables are used.
5385 Version 2.00 24-Sep-98
5386 ----------------------
5388 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
5389 it any more.
5391 2. Allow quantification of (?>) groups, and make it work correctly.
5393 3. The first character computation wasn't working for (?>) groups.
5395 4. Correct the implementation of \Z (it is permitted to match on the \n at the
5396 end of the subject) and add 5.005's \z, which really does match only at the
5397 very end of the subject.
5399 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
5401 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
5402 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
5403 localized options. All options to pcre_study() were also removed.
5405 7. Add other new features from 5.005:
5407 $(?<= positive lookbehind
5408 $(?<! negative lookbehind
5409 (?imsx-imsx) added the unsetting capability
5410 such a setting is global if at outer level; local otherwise
5411 (?imsx-imsx:) non-capturing groups with option setting
5412 (?(cond)re|re) conditional pattern matching
5414 A backreference to itself in a repeated group matches the previous
5415 captured string.
5417 8. General tidying up of studying (both automatic and via "study")
5418 consequential on the addition of new assertions.
5420 9. As in 5.005, unlimited repeated groups that could match an empty substring
5421 are no longer faulted at compile time. Instead, the loop is forcibly broken at
5422 runtime if any iteration does actually match an empty substring.
5424 10. Include the RunTest script in the distribution.
5426 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
5427 discrepancies, some of which were old and were also with respect to 5.004. They
5428 have now been fixed.
5431 Version 1.09 28-Apr-98
5432 ----------------------
5434 1. A negated single character class followed by a quantifier with a minimum
5435 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
5436 program crashes, or just wrong answers. This did not apply to negated classes
5437 containing more than one character, or to minima other than one.
5440 Version 1.08 27-Mar-98
5441 ----------------------
5443 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
5445 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
5446 latter must appear before anything that relies on it in the pattern.
5449 Version 1.07 16-Feb-98
5450 ----------------------
5452 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
5453 repeat of a potentially empty string).
5456 Version 1.06 23-Jan-98
5457 ----------------------
5459 1. Added Markus Oberhumer's little patches for C++.
5461 2. Literal strings longer than 255 characters were broken.
5464 Version 1.05 23-Dec-97
5465 ----------------------
5467 1. Negated character classes containing more than one character were failing if
5468 PCRE_CASELESS was set at run time.
5471 Version 1.04 19-Dec-97
5472 ----------------------
5474 1. Corrected the man page, where some "const" qualifiers had been omitted.
5476 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
5477 input syntax.
5479 3. Fixed memory leak which occurred when a regex with back references was
5480 matched with an offsets vector that wasn't big enough. The temporary memory
5481 that is used in this case wasn't being freed if the match failed.
5483 4. Tidied pcretest to ensure it frees memory that it gets.
5485 5. Temporary memory was being obtained in the case where the passed offsets
5486 vector was exactly big enough.
5488 6. Corrected definition of offsetof() from change 5 below.
5490 7. I had screwed up change 6 below and broken the rules for the use of
5491 setjmp(). Now fixed.
5494 Version 1.03 18-Dec-97
5495 ----------------------
5497 1. A erroneous regex with a missing opening parenthesis was correctly
5498 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
5499 on some systems.
5501 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
5502 it was reported that one broken compiler failed on the former because "code" is
5503 also an independent variable.
5505 3. The erroneous regex a[]b caused an array overrun reference.
5507 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
5508 fail on data ending with that character. (It was going on too far, and checking
5509 the next character, typically a binary zero.) This was specific to the
5510 optimized code for single-character negative classes.
5512 5. Added a contributed patch from the TIN world which does the following:
5514 + Add an undef for memmove, in case the the system defines a macro for it.
5516 + Add a definition of offsetof(), in case there isn't one. (I don't know
5517 the reason behind this - offsetof() is part of the ANSI standard - but
5518 it does no harm).
5520 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
5521 most of the places where whitespace preceded '#'. I have given up and
5522 allowed the remaining 2 cases to be at the margin.
5524 + Rename some variables in pcre to eliminate shadowing. This seems very
5525 pedantic, but does no harm, of course.
5527 6. Moved the call to setjmp() into its own function, to get rid of warnings
5528 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
5530 7. Constructs such as \d{8,} were compiling into the equivalent of
5531 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
5532 outcome, but in this particular case used more store than had been allocated,
5533 which caused the bug to be discovered because it threw up an internal error.
5535 8. The debugging code in both pcre and pcretest for outputting the compiled
5536 form of a regex was going wrong in the case of back references followed by
5537 curly-bracketed repeats.
5540 Version 1.02 12-Dec-97
5541 ----------------------
5543 1. Typos in pcre.3 and comments in the source fixed.
5545 2. Applied a contributed patch to get rid of places where it used to remove
5546 'const' from variables, and fixed some signed/unsigned and uninitialized
5547 variable warnings.
5549 3. Added the "runtest" target to Makefile.
5551 4. Set default compiler flag to -O2 rather than just -O.
5554 Version 1.01 19-Nov-97
5555 ----------------------
5557 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
5558 like /([ab]*)*/, that is, for classes with more than one character in them.
5560 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
5561 as /((?>a*))*/ (a PCRE_EXTRA facility).
5564 Version 1.00 18-Nov-97
5565 ----------------------
5567 1. Added compile-time macros to support systems such as SunOS4 which don't have
5568 memmove() or strerror() but have other things that can be used instead.
5570 2. Arranged that "make clean" removes the executables.
5573 Version 0.99 27-Oct-97
5574 ----------------------
5576 1. Fixed bug in code for optimizing classes with only one character. It was
5577 initializing a 32-byte map regardless, which could cause it to run off the end
5578 of the memory it had got.
5580 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
5583 Version 0.98 22-Oct-97
5584 ----------------------
5586 1. Fixed bug in code for handling temporary memory usage when there are more
5587 back references than supplied space in the ovector. This could cause segfaults.
5590 Version 0.97 21-Oct-97
5591 ----------------------
5593 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
5595 2. Optimized negated single characters not to use a bit map.
5597 3. Brought error texts together as macro definitions; clarified some of them;
5598 fixed one that was wrong - it said "range out of order" when it meant "invalid
5599 escape sequence".
5601 4. Changed some char * arguments to const char *.
5603 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
5605 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
5606 pcretest.
5609 Version 0.96 16-Oct-97
5610 ----------------------
5612 1. Added a simple "pgrep" utility to the distribution.
5614 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
5615 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
5616 where "ddd" means "one or more decimal digits".
5618 3. Fixed serious bug. If a pattern had a back reference, but the call to
5619 pcre_exec() didn't supply a large enough ovector to record the related
5620 identifying subpattern, the match always failed. PCRE now remembers the number
5621 of the largest back reference, and gets some temporary memory in which to save
5622 the offsets during matching if necessary, in order to ensure that
5623 backreferences always work.
5625 4. Increased the compatibility with Perl in a number of ways:
5627 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
5628 to request this handling. The option can be set at compile or exec time.
5630 (b) $ matches before a terminating newline by default; an option
5631 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
5632 mode). The option can be set at compile or exec time.
5634 (c) The handling of \ followed by a digit other than 0 is now supposed to be
5635 the same as Perl's. If the decimal number it represents is less than 10
5636 or there aren't that many previous left capturing parentheses, an octal
5637 escape is read. Inside a character class, it's always an octal escape,
5638 even if it is a single digit.
5640 (d) An escaped but undefined alphabetic character is taken as a literal,
5641 unless PCRE_EXTRA is set. Currently this just reserves the remaining
5642 escapes.
5644 (e) {0} is now permitted. (The previous item is removed from the compiled
5645 pattern).
5647 5. Changed all the names of code files so that the basic parts are no longer
5648 than 10 characters, and abolished the teeny "globals.c" file.
5650 6. Changed the handling of character classes; they are now done with a 32-byte
5651 bit map always.
5653 7. Added the -d and /D options to pcretest to make it possible to look at the
5654 internals of compilation without having to recompile pcre.
5657 Version 0.95 23-Sep-97
5658 ----------------------
5660 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
5661 \x20 at the start of a run of normal characters. These were being treated as
5662 real characters, instead of the source characters being re-checked.
5665 Version 0.94 18-Sep-97
5666 ----------------------
5668 1. The functions are now thread-safe, with the caveat that the global variables
5669 containing pointers to malloc() and free() or alternative functions are the
5670 same for all threads.
5672 2. Get pcre_study() to generate a bitmap of initial characters for non-
5673 anchored patterns when this is possible, and use it if passed to pcre_exec().
5676 Version 0.93 15-Sep-97
5677 ----------------------
5679 1. /(b)|(:+)/ was computing an incorrect first character.
5681 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5682 but not actually doing anything yet.
5684 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5685 as Perl does (e.g. [-az] or [az-]).
5687 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5688 all possible positions.
5690 5. Split up into different modules to avoid including unneeded functions in a
5691 compiled binary. However, compile and exec are still in one module. The "study"
5692 function is split off.
5694 6. The character tables are now in a separate module whose source is generated
5695 by an auxiliary program - but can then be edited by hand if required. There are
5696 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5697 toupper() in the code.
5699 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5700 make them global. Abolish the function for setting them, as the caller can now
5701 set them directly.
5704 Version 0.92 11-Sep-97
5705 ----------------------
5707 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5708 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5710 2. Caseless matching was not working in character classes if the characters in
5711 the pattern were in upper case.
5713 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5715 4. Make PCRE_ANCHORED public and accept as a compile option.
5717 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5718 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5719 pass them.
5721 6. Give an error if bad option bits passed at compile or run time.
5723 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5724 pcretest to cause it to pass that flag.
5726 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5727 options, and the first character, if set.
5729 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5732 Version 0.91 10-Sep-97
5733 ----------------------
5735 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5736 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5738 2. PCRE was looping on encountering an indefinitely repeated back reference to
5739 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5740 Perl does - treats the match as successful.
5742 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5