ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 566 - (show annotations)
Wed Nov 3 18:32:55 2010 UTC (4 years, 11 months ago) by ph10
File size: 194016 byte(s)
Error occurred while calculating annotation data.
Fix broken /g code in pcretest and -g code in pcredemo.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.11 10-Oct-2010
5 ------------------------
7 1. (*THEN) was not working properly if there were untried alternatives prior
8 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9 backtracked to try for "b" instead of moving to the next alternative branch
10 at the same level (in this case, to look for "c"). The Perl documentation
11 is clear that when (*THEN) is backtracked onto, it goes to the "next
12 alternative in the innermost enclosing group".
14 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
15 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
16 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
17 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
18 (*THEN).
20 3. If \s appeared in a character class, it removed the VT character from
21 the class, even if it had been included by some previous item, for example
22 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
23 of \s, but is part of the POSIX "space" class.)
25 4. A partial match never returns an empty string (because you can always
26 match an empty string at the end of the subject); however the checking for
27 an empty string was starting at the "start of match" point. This has been
28 changed to the "earliest inspected character" point, because the returned
29 data for a partial match starts at this character. This means that, for
30 example, /(?<=abc)def/ gives a partial match for the subject "abc"
31 (previously it gave "no match").
33 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
34 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
35 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
36 has an implication that the given string is incomplete (because a partial
37 match is preferred over a full match). For this reason, these items now
38 give a partial match in this situation. [Aside: previously, the one case
39 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
40 match rather than a full match, which was wrong by the old rules, but is
41 now correct.]
43 6. There was a bug in the handling of #-introduced comments, recognized when
44 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
45 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
46 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
47 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
48 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
49 places in pcre_compile().
51 7. Related to (6) above, when pcre_compile() was skipping #-introduced
52 comments when looking ahead for named forward references to subpatterns,
53 the only newline sequence it recognized was NL. It now handles newlines
54 according to the set newline convention.
56 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
57 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
58 cater for a lack of strerror(). These oversights have been fixed.
60 9. Added --match-limit and --recursion-limit to pcregrep.
62 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
64 11. When the -o option was used, pcregrep was setting a return code of 1, even
65 when matches were found, and --line-buffered was not being honoured.
67 12. Added an optional parentheses number to the -o and --only-matching options
68 of pcregrep.
70 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
71 can match an empty string. The code to do it in pcretest and pcredemo
72 needed fixing:
74 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
75 only one byte after an empty string match just before CRLF (this case
76 just got forgotten; "any" and "anycrlf" were OK).
78 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
79 mode when an empty string match preceded an ASCII character followed by
80 a non-ASCII character. (The code for advancing by one character rather
81 than one byte was nonsense.)
83 (c) The pcredemo.c sample program did not have any code at all to handle
84 the cases when CRLF is a valid newline sequence.
87 Version 8.10 25-Jun-2010
88 ------------------------
90 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
91 THEN.
93 2. (*ACCEPT) was not working when inside an atomic group.
95 3. Inside a character class, \B is treated as a literal by default, but
96 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
97 causes the error). The code is unchanged, but I tidied the documentation.
99 4. Inside a character class, PCRE always treated \R and \X as literals,
100 whereas Perl faults them if its -w option is set. I have changed PCRE so
101 that it faults them when PCRE_EXTRA is set.
103 5. Added support for \N, which always matches any character other than
104 newline. (It is the same as "." when PCRE_DOTALL is not set.)
106 6. When compiling pcregrep with newer versions of gcc which may have
107 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
108 declared with attribute warn_unused_result" were given. Just casting the
109 result to (void) does not stop the warnings; a more elaborate fudge is
110 needed. I've used a macro to implement this.
112 7. Minor change to pcretest.c to avoid a compiler warning.
114 8. Added four artifical Unicode properties to help with an option to make
115 \s etc use properties (see next item). The new properties are: Xan
116 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
118 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
119 use Unicode properties. (*UCP) at the start of a pattern can be used to set
120 this option. Modified pcretest to add /W to test this facility. Added
121 REG_UCP to make it available via the POSIX interface.
123 10. Added --line-buffered to pcregrep.
125 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
126 studied, and the match started with a letter with a code point greater than
127 127 whose first byte was different to the first byte of the other case of
128 the letter, the other case of this starting letter was not recognized
129 (#976).
131 12. If a pattern that was studied started with a repeated Unicode property
132 test, for example, \p{Nd}+, there was the theoretical possibility of
133 setting up an incorrect bitmap of starting bytes, but fortunately it could
134 not have actually happened in practice until change 8 above was made (it
135 added property types that matched character-matching opcodes).
137 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
138 possible starting bytes for non-anchored patterns.
140 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
141 \R, and also a number of cases that involve Unicode properties, both
142 explicit and implicit when PCRE_UCP is set.
144 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
145 input, it could crash or give wrong results if characters with values
146 greater than 0xc0 were present in the subject string. (Detail: it assumed
147 UTF-8 input when processing these items.)
149 16. Added a lot of (int) casts to avoid compiler warnings in systems where
150 size_t is 64-bit (#991).
152 17. Added a check for running out of memory when PCRE is compiled with
153 --disable-stack-for-recursion (#990).
155 18. If the last data line in a file for pcretest does not have a newline on
156 the end, a newline was missing in the output.
158 19. The default pcre_chartables.c file recognizes only ASCII characters (values
159 less than 128) in its various bitmaps. However, there is a facility for
160 generating tables according to the current locale when PCRE is compiled. It
161 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
162 characters, are recognized by isspace() and therefore were getting set in
163 these tables, and indeed these tables seem to approximate to ISO 8859. This
164 caused a problem in UTF-8 mode when pcre_study() was used to create a list
165 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
166 which of course cannot start UTF-8 characters. I have changed the code so
167 that only real ASCII characters (less than 128) and the correct starting
168 bytes for UTF-8 encodings are set for characters greater than 127 when in
169 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
170 altogether.)
172 20. Added the /T option to pcretest so as to be able to run tests with non-
173 standard character tables, thus making it possible to include the tests
174 used for 19 above in the standard set of tests.
176 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
177 reference to a subpattern the other side of a comment that contains an
178 opening parenthesis caused either an internal compiling error, or a
179 reference to the wrong subpattern.
182 Version 8.02 19-Mar-2010
183 ------------------------
185 1. The Unicode data tables have been updated to Unicode 5.2.0.
187 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
188 configured.
190 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
191 original author of that file, following a query about its status.
193 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
194 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
196 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
197 quantifier applied to a forward-referencing subroutine call, could compile
198 incorrect code or give the error "internal error: previously-checked
199 referenced subpattern not found".
201 6. Both MS Visual Studio and Symbian OS have problems with initializing
202 variables to point to external functions. For these systems, therefore,
203 pcre_malloc etc. are now initialized to local functions that call the
204 relevant global functions.
206 7. There were two entries missing in the vectors called coptable and poptable
207 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
208 I've fixed the data, and added a kludgy way of testing at compile time that
209 the lengths are correct (equal to the number of opcodes).
211 8. Following on from 7, I added a similar kludge to check the length of the
212 eint vector in pcreposix.c.
214 9. Error texts for pcre_compile() are held as one long string to avoid too
215 much relocation at load time. To find a text, the string is searched,
216 counting zeros. There was no check for running off the end of the string,
217 which could happen if a new error number was added without updating the
218 string.
220 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
222 11. \K was not working if it appeared in an atomic group or in a group that
223 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
224 \K is "not well defined" if used in an assertion. PCRE now accepts it if
225 the assertion is positive, but not if it is negative.
227 12. Change 11 fortuitously reduced the size of the stack frame used in the
228 "match()" function of pcre_exec.c by one pointer. Forthcoming
229 implementation of support for (*MARK) will need an extra pointer on the
230 stack; I have reserved it now, so that the stack frame size does not
231 decrease.
233 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
234 item in branch that calls a recursion is a subroutine call - as in the
235 second branch in the above example - was incorrectly given the compile-
236 time error "recursive call could loop indefinitely" because pcre_compile()
237 was not correctly checking the subroutine for matching a non-empty string.
239 14. The checks for overrunning compiling workspace could trigger after an
240 overrun had occurred. This is a "should never occur" error, but it can be
241 triggered by pathological patterns such as hundreds of nested parentheses.
242 The checks now trigger 100 bytes before the end of the workspace.
244 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
247 Version 8.01 19-Jan-2010
248 ------------------------
250 1. If a pattern contained a conditional subpattern with only one branch (in
251 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
252 computed the wrong minimum data length (which is of course zero for such
253 subpatterns). This could cause incorrect "no match" results.
255 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
256 the pattern is reset in the first branch, pcre_compile() failed with
257 "internal error: code overflow at offset...". This happened only when
258 the reset was to the original external option setting. (An optimization
259 abstracts leading options settings into an external setting, which was the
260 cause of this.)
262 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
263 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
264 assertion pattern did not match (meaning that the assertion was true), it
265 was incorrectly treated as false if the SKIP had been reached during the
266 matching. This also applied to assertions used as conditions.
268 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
269 assertion subpattern, including such a pattern used as a condition,
270 unpredictable results occurred, instead of the error return
273 5. The C++ GlobalReplace function was not working like Perl for the special
274 situation when an empty string is matched. It now does the fancy magic
275 stuff that is necessary.
277 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
278 removed. (These were left over from very, very early versions of PCRE.)
280 7. Some cosmetic changes to the code to make life easier when compiling it
281 as part of something else:
283 (a) Change DEBUG to PCRE_DEBUG.
285 (b) In pcre_compile(), rename the member of the "branch_chain" structure
286 called "current" as "current_branch", to prevent a collision with the
287 Linux macro when compiled as a kernel module.
289 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
290 prevent a collision with the Linux macro when compiled as a kernel
291 module.
293 8. In pcre_compile() there are some checks for integer overflows that used to
294 cast potentially large values to (double). This has been changed to that
295 when building, a check for int64_t is made, and if it is found, it is used
296 instead, thus avoiding the use of floating point arithmetic. (There is no
297 other use of FP in PCRE.) If int64_t is not found, the fallback is to
298 double.
300 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
301 2005 (difference between two addresses compared to an unsigned value).
303 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
304 custom one, because of the following reported problem in Windows:
306 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
307 under Win32.
308 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
309 therefore missing the function definition.
310 - The compiler thus generates a "C" signature for the test function.
311 - The linker fails to find the "C" function.
312 - PCRE fails to configure if asked to do so against libbz2.
314 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
315 messages were output:
317 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
318 rerunning libtoolize, to keep the correct libtool macros in-tree.
319 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
321 I have done both of these things.
323 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
324 most of the time, it *can* run out if it is given a pattern that contains a
325 runaway infinite recursion. I updated the discussion in the pcrestack man
326 page.
328 13. Now that we have gone to the x.xx style of version numbers, the minor
329 version may start with zero. Using 08 or 09 is a bad idea because users
330 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
331 interpreted as invalid octal numbers. I've updated the previous comment in
332 configure.ac, and also added a check that gives an error if 08 or 09 are
333 used.
335 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
336 causing partial matching to fail when the end of the subject matched \W
337 in a UTF-8 pattern where \W was quantified with a minimum of 3.
339 15. There were some discrepancies between the declarations in pcre_internal.h
340 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
341 their definitions. The declarations used "const uschar *" and the
342 definitions used USPTR. Even though USPTR is normally defined as "const
343 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
344 reported that: "This difference in casting confuses some C++ compilers, for
345 example, SunCC recognizes above declarations as different functions and
346 generates broken code for hbpcre." I have changed the declarations to use
347 USPTR.
349 16. GNU libtool is named differently on some systems. The autogen.sh script now
350 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
351 (FreeBSD).
353 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
354 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
355 comment: "Figure out how to create a longlong from a string: strtoll and
356 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
357 instance, but it only takes 2 args instead of 3!"
359 18. A subtle bug concerned with back references has been fixed by a change of
360 specification, with a corresponding code fix. A pattern such as
361 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
362 refers, was giving matches when it shouldn't. For example, xa=xaaa would
363 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
364 same bug. Such groups have to be quantified to be useful, or contained
365 inside another quantified group. (If there's no repetition, the reference
366 can never match.) The problem arises because, having left the group and
367 moved on to the rest of the pattern, a later failure that backtracks into
368 the group uses the captured value from the final iteration of the group
369 rather than the correct earlier one. I have fixed this in PCRE by forcing
370 any group that contains a reference to itself to be an atomic group; that
371 is, there cannot be any backtracking into it once it has completed. This is
372 similar to recursive and subroutine calls.
375 Version 8.00 19-Oct-09
376 ----------------------
378 1. The table for translating pcre_compile() error codes into POSIX error codes
379 was out-of-date, and there was no check on the pcre_compile() error code
380 being within the table. This could lead to an OK return being given in
381 error.
383 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
384 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
385 in a Windows environment.
387 3. The pcregrep --count option prints the count for each file even when it is
388 zero, as does GNU grep. However, pcregrep was also printing all files when
389 --files-with-matches was added. Now, when both options are given, it prints
390 counts only for those files that have at least one match. (GNU grep just
391 prints the file name in this circumstance, but including the count seems
392 more useful - otherwise, why use --count?) Also ensured that the
393 combination -clh just lists non-zero counts, with no names.
395 4. The long form of the pcregrep -F option was incorrectly implemented as
396 --fixed_strings instead of --fixed-strings. This is an incompatible change,
397 but it seems right to fix it, and I didn't think it was worth preserving
398 the old behaviour.
400 5. The command line items --regex=pattern and --regexp=pattern were not
401 recognized by pcregrep, which required --regex pattern or --regexp pattern
402 (with a space rather than an '='). The man page documented the '=' forms,
403 which are compatible with GNU grep; these now work.
405 6. No libpcreposix.pc file was created for pkg-config; there was just
406 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
408 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
409 when UCP support is not needed, by modifying the Python script that
410 generates it from Unicode data files. This should not matter if the module
411 is correctly used as a library, but I received one complaint about 50K of
412 unwanted data. My guess is that the person linked everything into his
413 program rather than using a library. Anyway, it does no harm.
415 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
416 was a minimum greater than 1 for a wide character in a possessive
417 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
418 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
419 character. Chaos in the form of incorrect output or a compiling loop could
420 result.
422 9. The restrictions on what a pattern can contain when partial matching is
423 requested for pcre_exec() have been removed. All patterns can now be
424 partially matched by this function. In addition, if there are at least two
425 slots in the offset vector, the offset of the earliest inspected character
426 for the match and the offset of the end of the subject are set in them when
427 PCRE_ERROR_PARTIAL is returned.
429 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
430 synonymous with PCRE_PARTIAL, for backwards compatibility, and
431 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
432 and may be more useful for multi-segment matching.
434 11. Partial matching with pcre_exec() is now more intuitive. A partial match
435 used to be given if ever the end of the subject was reached; now it is
436 given only if matching could not proceed because another character was
437 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
438 string "Z", which now yields "no match" instead of "partial match". In the
439 case of pcre_dfa_exec(), "no match" is given if every matching path for the
440 final character ended with (*FAIL).
442 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
443 if the pattern had a "must contain" character that was already found in the
444 earlier partial match, unless partial matching was again requested. For
445 example, with the pattern /dog.(body)?/, the "must contain" character is
446 "g". If the first part-match was for the string "dog", restarting with
447 "sbody" failed. This bug has been fixed.
449 13. The string returned by pcre_dfa_exec() after a partial match has been
450 changed so that it starts at the first inspected character rather than the
451 first character of the match. This makes a difference only if the pattern
452 starts with a lookbehind assertion or \b or \B (\K is not supported by
453 pcre_dfa_exec()). It's an incompatible change, but it makes the two
454 matching functions compatible, and I think it's the right thing to do.
456 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
457 so that the demonstration program is easily available in environments where
458 PCRE has not been installed from source.
460 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
461 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
462 library.
464 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
465 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
466 is not the first non-POSIX option to be added. Clearly some people find
467 these options useful.
469 17. If a caller to the POSIX matching function regexec() passes a non-zero
470 value for nmatch with a NULL value for pmatch, the value of
471 nmatch is forced to zero.
473 18. RunGrepTest did not have a test for the availability of the -u option of
474 the diff command, as RunTest does. It now checks in the same way as
475 RunTest, and also checks for the -b option.
477 19. If an odd number of negated classes containing just a single character
478 interposed, within parentheses, between a forward reference to a named
479 subpattern and the definition of the subpattern, compilation crashed with
480 an internal error, complaining that it could not find the referenced
481 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
482 [The bug was that it was starting one character too far in when skipping
483 over the character class, thus treating the ] as data rather than
484 terminating the class. This meant it could skip too much.]
486 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
487 /g option in pcretest when the pattern contains \K, which makes it possible
488 to have an empty string match not at the start, even when the pattern is
489 anchored. Updated pcretest and pcredemo to use this option.
491 21. If the maximum number of capturing subpatterns in a recursion was greater
492 than the maximum at the outer level, the higher number was returned, but
493 with unset values at the outer level. The correct (outer level) value is
494 now given.
496 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
497 PCRE did not set those parentheses (unlike Perl). I have now found a way to
498 make it do so. The string so far is captured, making this feature
499 compatible with Perl.
501 23. The tests have been re-organized, adding tests 11 and 12, to make it
502 possible to check the Perl 5.10 features against Perl 5.10.
504 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
505 pattern matches a fixed length string. PCRE did not allow this; now it
506 does. Neither allows recursion.
508 25. I finally figured out how to implement a request to provide the minimum
509 length of subject string that was needed in order to match a given pattern.
510 (It was back references and recursion that I had previously got hung up
511 on.) This code has now been added to pcre_study(); it finds a lower bound
512 to the length of subject needed. It is not necessarily the greatest lower
513 bound, but using it to avoid searching strings that are too short does give
514 some useful speed-ups. The value is available to calling programs via
515 pcre_fullinfo().
517 26. While implementing 25, I discovered to my embarrassment that pcretest had
518 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
519 study optimizations had never been tested with that matching function.
520 Oops. What is worse, even when it was passed study data, there was a bug in
521 pcre_dfa_exec() that meant it never actually used it. Double oops. There
522 were also very few tests of studied patterns with pcre_dfa_exec().
524 27. If (?| is used to create subpatterns with duplicate numbers, they are now
525 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
526 on the other side of the coin, they are no longer allowed to have different
527 names, because these cannot be distinguished in PCRE, and this has caused
528 confusion. (This is a difference from Perl.)
530 28. When duplicate subpattern names are present (necessarily with different
531 numbers, as required by 27 above), and a test is made by name in a
532 conditional pattern, either for a subpattern having been matched, or for
533 recursion in such a pattern, all the associated numbered subpatterns are
534 tested, and the overall condition is true if the condition is true for any
535 one of them. This is the way Perl works, and is also more like the way
536 testing by number works.
539 Version 7.9 11-Apr-09
540 ---------------------
542 1. When building with support for bzlib/zlib (pcregrep) and/or readline
543 (pcretest), all targets were linked against these libraries. This included
544 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
545 libraries. This caused unwanted dependencies to be created. This problem
546 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
547 pcretest is linked with readline.
549 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
550 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
551 moved outside it again, because FALSE and TRUE are already defined in AIX,
552 but BOOL is not.
554 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
555 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
557 4. The pcregrep documentation said spaces were inserted as well as colons (or
558 hyphens) following file names and line numbers when outputting matching
559 lines. This is not true; no spaces are inserted. I have also clarified the
560 wording for the --colour (or --color) option.
562 5. In pcregrep, when --colour was used with -o, the list of matching strings
563 was not coloured; this is different to GNU grep, so I have changed it to be
564 the same.
566 6. When --colo(u)r was used in pcregrep, only the first matching substring in
567 each matching line was coloured. Now it goes on to look for further matches
568 of any of the test patterns, which is the same behaviour as GNU grep.
570 7. A pattern that could match an empty string could cause pcregrep to loop; it
571 doesn't make sense to accept an empty string match in pcregrep, so I have
572 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
573 seems to be how GNU grep behaves.
575 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
576 start or after a newline", because the conditional assertion was not being
577 correctly handled. The rule now is that both the assertion and what follows
578 in the first alternative must satisfy the test.
580 9. If auto-callout was enabled in a pattern with a conditional group whose
581 condition was an assertion, PCRE could crash during matching, both with
582 pcre_exec() and pcre_dfa_exec().
584 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
585 used for matching.
587 11. Unicode property support in character classes was not working for
588 characters (bytes) greater than 127 when not in UTF-8 mode.
590 12. Added the -M command line option to pcretest.
592 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
594 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
596 16. Added comments and documentation about mis-use of no_arg in the C++
597 wrapper.
599 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
600 from Martin Jerabek that uses macro names for all relevant character and
601 string constants.
603 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
604 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
605 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
606 these, but not everybody uses configure.
608 19. A conditional group that had only one branch was not being correctly
609 recognized as an item that could match an empty string. This meant that an
610 enclosing group might also not be so recognized, causing infinite looping
611 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
612 with the subject "ab", where knowledge that the repeated group can match
613 nothing is needed in order to break the loop.
615 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
616 exec(), but without supplying a callout function, matching went wrong.
618 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
619 leak if the size of the offset vector was greater than 30. When the vector
620 is smaller, the saved offsets during recursion go onto a local stack
621 vector, but for larger vectors malloc() is used. It was failing to free
622 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
623 error, in fact).
625 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
626 heapframe that is used only when UTF-8 support is enabled. This caused no
627 problem, but was untidy.
629 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
630 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
631 included within another project.
633 24. Steven Van Ingelgem's patches to add more options to the CMake support,
634 slightly modified by me:
636 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
637 not building pcregrep.
639 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
640 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
642 25. Forward references, both numeric and by name, in patterns that made use of
643 duplicate group numbers, could behave incorrectly or give incorrect errors,
644 because when scanning forward to find the reference group, PCRE was not
645 taking into account the duplicate group numbers. A pattern such as
646 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
648 26. Changed a few more instances of "const unsigned char *" to USPTR, making
649 the feature of a custom pointer more persuasive (as requested by a user).
651 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
652 pcretest.c, inside #ifndefs, because it seems they are sometimes already
653 pre-defined.
655 28. Added support for (*UTF8) at the start of a pattern.
657 29. Arrange for flags added by the "release type" setting in CMake to be shown
658 in the configuration summary.
661 Version 7.8 05-Sep-08
662 ---------------------
664 1. Replaced UCP searching code with optimized version as implemented for Ad
665 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
666 stage table and inline lookup instead of a function, giving speed ups of 2
667 to 5 times on some simple patterns that I tested. Permission was given to
668 distribute the MultiStage2.py script that generates the tables (it's not in
669 the tarball, but is in the Subversion repository).
671 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
672 scripts.
674 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
675 a group with a zero qualifier. The result of the study could be incorrect,
676 or the function might crash, depending on the pattern.
678 4. Caseless matching was not working for non-ASCII characters in back
679 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
680 It now works when Unicode Property Support is available.
682 5. In pcretest, an escape such as \x{de} in the data was always generating
683 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
684 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
685 truncation.
687 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
689 7. Added two (int) casts to pcregrep when printing the difference of two
690 pointers, in case they are 64-bit values.
692 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
693 test 2 if it fails.
695 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
696 and a #define of that name to empty if it is not externally set. This is to
697 allow users of MSVC to set it if necessary.
699 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
700 the convenience functions in the pcre_get.c source file.
702 11. An option change at the start of a pattern that had top-level alternatives
703 could cause overwriting and/or a crash. This command provoked a crash in
704 some environments:
706 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
708 This potential security problem was recorded as CVE-2008-2371.
710 12. For a pattern where the match had to start at the beginning or immediately
711 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
712 pcre_dfa_exec() could read past the end of the passed subject if there was
713 no match. To help with detecting such bugs (e.g. with valgrind), I modified
714 pcretest so that it places the subject at the end of its malloc-ed buffer.
716 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
717 exec() might read past the end of the data buffer in UTF-8 mode.
719 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
720 the data contained the byte 0x85 as part of a UTF-8 character within its
721 first line. This applied both to normal and DFA matching.
723 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
724 /^[^d]*?$/8 failed to match "abc".
726 16. Added a missing copyright notice to pcrecpp_internal.h.
728 17. Make it more clear in the documentation that values returned from
729 pcre_exec() in ovector are byte offsets, not character counts.
731 18. Tidied a few places to stop certain compilers from issuing warnings.
733 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
734 supplied by Stefan Weber. I made a further small update for 7.8 because
735 there is a change of source arrangements: the pcre_searchfuncs.c module is
736 replaced by pcre_ucd.c.
739 Version 7.7 07-May-08
740 ---------------------
742 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
743 a string to a long long, pretend we don't even have a long long." This is
744 done by checking for the strtoq, strtoll, and _strtoi64 functions.
746 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
747 pre-7.6 versions, which defined a global no_arg variable instead of putting
748 it in the RE class. (See also #8 below.)
750 3. Remove a line of dead code, identified by coverity and reported by Nuno
751 Lopes.
753 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
755 (1) The include/exclude patterns were being applied to the whole pathnames
756 of files, instead of just to the final components.
758 (2) If there was more than one level of directory, the subdirectories were
759 skipped unless they satisfied the include/exclude conditions. This is
760 inconsistent with GNU grep (and could even be seen as contrary to the
761 pcregrep specification - which I improved to make it absolutely clear).
762 The action now is always to scan all levels of directory, and just
763 apply the include/exclude patterns to regular files.
765 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
766 --exclude_dir in the tests to avoid scanning .svn directories.
768 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
769 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
770 doesn't support NULs in patterns.
772 7. Added some missing "const"s to declarations of static tables in
773 pcre_compile.c and pcre_dfa_exec.c.
775 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
776 caused by fix #2 above. (Subsequently also a second patch to fix the
777 first patch. And a third patch - this was a messy problem.)
779 9. Applied Craig's patch to remove the use of push_back().
781 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
782 matching function regexec().
784 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
785 which, however, unlike Perl's \g{...}, are subroutine calls, not back
786 references. PCRE supports relative numbers with this syntax (I don't think
787 Oniguruma does).
789 12. Previously, a group with a zero repeat such as (...){0} was completely
790 omitted from the compiled regex. However, this means that if the group
791 was called as a subroutine from elsewhere in the pattern, things went wrong
792 (an internal error was given). Such groups are now left in the compiled
793 pattern, with a new opcode that causes them to be skipped at execution
794 time.
796 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
797 to the way PCRE behaves:
799 (a) A lone ] character is dis-allowed (Perl treats it as data).
801 (b) A back reference to an unmatched subpattern matches an empty string
802 (Perl fails the current match path).
804 (c) A data ] in a character class must be notated as \] because if the
805 first data character in a class is ], it defines an empty class. (In
806 Perl it is not possible to have an empty class.) The empty class []
807 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
808 The negative empty class [^] matches any one character, independently
809 of the DOTALL setting.
811 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
812 non-existent subpattern following a character class starting with ']' and
813 containing () gave an internal compiling error instead of "reference to
814 non-existent subpattern". Fortunately, when the pattern did exist, the
815 compiled code was correct. (When scanning forwards to check for the
816 existencd of the subpattern, it was treating the data ']' as terminating
817 the class, so got the count wrong. When actually compiling, the reference
818 was subsequently set up correctly.)
820 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
821 it was being rejected as not supported by pcre_dfa_exec(), even though
822 other assertions are supported. I have made pcre_dfa_exec() support
823 (*FAIL).
825 16. The implementation of 13c above involved the invention of a new opcode,
826 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
827 cannot be changed at match time, I realized I could make a small
828 improvement to matching performance by compiling OP_ALLANY instead of
829 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
830 on the OP_ANY path.
832 17. Compiling pcretest on Windows with readline support failed without the
833 following two fixes: (1) Make the unistd.h include conditional on
834 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
836 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
837 ncurses library to be included for pcretest when ReadLine support is
838 requested, but also to allow for it to be overridden. This patch came from
839 Daniel Bergström.
841 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
842 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
843 any errors with the current Unicode tables. Thanks to Peter Kankowski for
844 spotting this.
847 Version 7.6 28-Jan-08
848 ---------------------
850 1. A character class containing a very large number of characters with
851 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
852 overflow.
854 2. Patch to cut out the "long long" test in pcrecpp_unittest when
855 HAVE_LONG_LONG is not defined.
857 3. Applied Christian Ehrlicher's patch to update the CMake build files to
858 bring them up to date and include new features. This patch includes:
860 - Fixed PH's badly added libz and libbz2 support.
861 - Fixed a problem with static linking.
862 - Added pcredemo. [But later removed - see 7 below.]
863 - Fixed dftables problem and added an option.
864 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
866 - Added readline support for pcretest.
867 - Added an listing of the option settings after cmake has run.
869 4. A user submitted a patch to Makefile that makes it easy to create
870 "pcre.dll" under mingw when using Configure/Make. I added stuff to
871 Makefile.am that cause it to include this special target, without
872 affecting anything else. Note that the same mingw target plus all
873 the other distribution libraries and programs are now supported
874 when configuring with CMake (see 6 below) instead of with
875 Configure/Make.
877 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
878 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
879 exported in the Windows port". It has not yet been confirmed that the patch
880 solves the problem, but it does no harm.
882 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
883 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
884 with CMake, and also correct the comment about stack recursion.
886 7. Remove the automatic building of pcredemo from the ./configure system and
887 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
888 of a program that users should build themselves after PCRE is installed, so
889 building it automatically is not really right. What is more, it gave
890 trouble in some build environments.
892 8. Further tidies to CMakeLists.txt from Sheri and Christian.
895 Version 7.5 10-Jan-08
896 ---------------------
898 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
899 values in parens when parsing an RE using the C++ wrapper."
901 2. Negative specials like \S did not work in character classes in UTF-8 mode.
902 Characters greater than 255 were excluded from the class instead of being
903 included.
905 3. The same bug as (2) above applied to negated POSIX classes such as
906 [:^space:].
908 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
909 defined or documented. It seems to have been a typo for PCRE_STATIC, so
910 I have changed it.
912 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
913 first named subpattern) and a construct such as (?&a) would reference the
914 first named subpattern whose name started with "a" (in other words, the
915 length check was missing). Both these problems are fixed. "Subpattern name
916 expected" is now given for (?&) (a zero-length name), and this patch also
917 makes it give the same error for \k'' (previously it complained that that
918 was a reference to a non-existent subpattern).
920 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
921 this is right because (?- can be followed by option settings as well as by
922 digits. I have, however, made the messages clearer.
924 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
925 than the number used in the conditional) now cause a compile-time error.
926 This is actually not compatible with Perl, which accepts such patterns, but
927 treats the conditional as always being FALSE (as PCRE used to), but it
928 seems to me that giving a diagnostic is better.
930 8. Change "alphameric" to the more common word "alphanumeric" in comments
931 and messages.
933 9. Fix two occurrences of "backslash" in comments that should have been
934 "backspace".
936 10. Remove two redundant lines of code that can never be obeyed (their function
937 was moved elsewhere).
939 11. The program that makes PCRE's Unicode character property table had a bug
940 which caused it to generate incorrect table entries for sequences of
941 characters that have the same character type, but are in different scripts.
942 It amalgamated them into a single range, with the script of the first of
943 them. In other words, some characters were in the wrong script. There were
944 thirteen such cases, affecting characters in the following ranges:
946 U+002b0 - U+002c1
947 U+0060c - U+0060d
948 U+0061e - U+00612
949 U+0064b - U+0065e
950 U+0074d - U+0076d
951 U+01800 - U+01805
952 U+01d00 - U+01d77
953 U+01d9b - U+01dbf
954 U+0200b - U+0200f
955 U+030fc - U+030fe
956 U+03260 - U+0327f
957 U+0fb46 - U+0fbb1
958 U+10450 - U+1049d
960 12. The -o option (show only the matching part of a line) for pcregrep was not
961 compatible with GNU grep in that, if there was more than one match in a
962 line, it showed only the first of them. It now behaves in the same way as
963 GNU grep.
965 13. If the -o and -v options were combined for pcregrep, it printed a blank
966 line for every non-matching line. GNU grep prints nothing, and pcregrep now
967 does the same. The return code can be used to tell if there were any
968 non-matching lines.
970 14. Added --file-offsets and --line-offsets to pcregrep.
972 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
973 infinitely looping recursion. The bug was that positive lookaheads were not
974 being skipped when checking for a possible empty match (negative lookaheads
975 and both kinds of lookbehind were skipped).
977 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
978 inclusion of <windows.h> to before rather than after the definition of
979 INVALID_FILE_ATTRIBUTES (patch from David Byron).
981 17. Specifying a possessive quantifier with a specific limit for a Unicode
982 character property caused pcre_compile() to compile bad code, which led at
983 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
984 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
985 caused the error; without that there was no problem.
987 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
989 19. Added --enable-pcretest-libreadline.
991 20. In pcrecpp.cc, the variable 'count' was incremented twice in
992 RE::GlobalReplace(). As a result, the number of replacements returned was
993 double what it should be. I removed one of the increments, but Craig sent a
994 later patch that removed the other one (the right fix) and added unit tests
995 that check the return values (which was not done before).
997 21. Several CMake things:
999 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1000 the names libpcre and libpcreposix, not just pcre and pcreposix.
1002 (2) The above change means that pcretest and pcregrep are now correctly
1003 linked with the newly-built libraries, not previously installed ones.
1007 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1008 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1009 UTF-8 newline character). The key issue is that the pattern starts .*;
1010 this means that the match must be either at the beginning, or after a
1011 newline. The bug was in the code for advancing after a failed match and
1012 checking that the new position followed a newline. It was not taking
1013 account of UTF-8 characters correctly.
1015 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1016 character classes. PCRE was not treating the sequence [:...:] as a
1017 character class unless the ... were all letters. Perl, however, seems to
1018 allow any characters between [: and :], though of course it rejects as
1019 unknown any "names" that contain non-letters, because all the known class
1020 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1021 for example, whereas PCRE did not - it did not recognize a POSIX character
1022 class. This seemed a bit dangerous, so the code has been changed to be
1023 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1024 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1025 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1026 Perl does, and where it didn't before.
1028 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1029 Windows environments %n is disabled by default.
1032 Version 7.4 21-Sep-07
1033 ---------------------
1035 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1036 means that a class such as [\s] counted as "explicit reference to CR or
1037 LF". That isn't really right - the whole point of the change was to try to
1038 help when there was an actual mention of one of the two characters. So now
1039 the change happens only if \r or \n (or a literal CR or LF) character is
1040 encountered.
1042 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1043 of both had grown to the point where there were only 3 bits left.
1044 Fortunately, there was spare space in the data structure, and so I have
1045 moved the internal flags into a new 16-bit field to free up more option
1046 bits.
1048 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1049 but did not set the internal JCHANGED flag - either of these is enough to
1050 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1051 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1052 start sets both bits.
1054 4. Added options (at build time, compile time, exec time) to change \R from
1055 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1057 5. doc/pcresyntax.html was missing from the distribution.
1059 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1060 compatibility, even though it is no longer used.
1062 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1063 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1064 windows.h file is present (where different names are used). [This was
1065 reversed later after testing - see 16 below.]
1067 8. Changed all #include <config.h> to #include "config.h". There were also
1068 some further <pcre.h> cases that I changed to "pcre.h".
1070 9. When pcregrep was used with the --colour option, it missed the line ending
1071 sequence off the lines that it output.
1073 10. It was pointed out to me that arrays of string pointers cause lots of
1074 relocations when a shared library is dynamically loaded. A technique of
1075 using a single long string with a table of offsets can drastically reduce
1076 these. I have refactored PCRE in four places to do this. The result is
1077 dramatic:
1079 Originally: 290
1080 After changing UCP table: 187
1081 After changing error message table: 43
1082 After changing table of "verbs" 36
1083 After changing table of Posix names 22
1085 Thanks to the folks working on Gregex for glib for this insight.
1087 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1088 unicode-properties was also set.
1090 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1092 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1093 checked only for CRLF.
1095 14. Added casts to pcretest.c to avoid compiler warnings.
1097 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1099 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1100 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1101 entirely. This removes changes made in 7 above.
1103 17. The CMake files have been updated, and there is now more information about
1104 building with CMake in the NON-UNIX-USE document.
1107 Version 7.3 28-Aug-07
1108 ---------------------
1110 1. In the rejigging of the build system that eventually resulted in 7.1, the
1111 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1112 brackets there is not right, since it causes compilers to look for an
1113 installed pcre.h, not the version that is in the source that is being
1114 compiled (which of course may be different). I have changed it back to:
1116 #include "pcre.h"
1118 I have a vague recollection that the change was concerned with compiling in
1119 different directories, but in the new build system, that is taken care of
1120 by the VPATH setting the Makefile.
1122 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1123 when the subject happened to end in the byte 0x85 (e.g. if the last
1124 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1125 characters but of course it shouldn't be taken as a newline when it is part
1126 of another character. The bug was that, for an unlimited repeat of . in
1127 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1128 characters when looking for a newline.
1130 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1132 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1133 in debug output.
1135 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1136 long printing in the pcrecpp unittest when running under MinGW.
1138 6. ESC_K was left out of the EBCDIC table.
1140 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1141 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1142 limit also applies to "virtual nesting" when a pattern is recursive, and in
1143 this case 1000 isn't so big. I have been able to remove this limit at the
1144 expense of backing off one optimization in certain circumstances. Normally,
1145 when pcre_exec() would call its internal match() function recursively and
1146 immediately return the result unconditionally, it uses a "tail recursion"
1147 feature to save stack. However, when a subpattern that can match an empty
1148 string has an unlimited repetition quantifier, it no longer makes this
1149 optimization. That gives it a stack frame in which to save the data for
1150 checking that an empty string has been matched. Previously this was taken
1151 from the 1000-entry workspace that had been reserved. So now there is no
1152 explicit limit, but more stack is used.
1154 8. Applied Daniel's patches to solve problems with the import/export magic
1155 syntax that is required for Windows, and which was going wrong for the
1156 pcreposix and pcrecpp parts of the library. These were overlooked when this
1157 problem was solved for the main library.
1159 9. There were some crude static tests to avoid integer overflow when computing
1160 the size of patterns that contain repeated groups with explicit upper
1161 limits. As the maximum quantifier is 65535, the maximum group length was
1162 set at 30,000 so that the product of these two numbers did not overflow a
1163 32-bit integer. However, it turns out that people want to use groups that
1164 are longer than 30,000 bytes (though not repeat them that many times).
1165 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1166 made it possible to implement the integer overflow checks in a much more
1167 dynamic way, which I have now done. The artificial limitation on group
1168 length has been removed - we now have only the limit on the total length of
1169 the compiled pattern, which depends on the LINK_SIZE setting.
1171 10. Fixed a bug in the documentation for get/copy named substring when
1172 duplicate names are permitted. If none of the named substrings are set, the
1173 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1174 empty string.
1176 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1177 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1178 because the ] is interpreted as the first data character and the
1179 terminating ] is not found. PCRE has been made compatible with Perl in this
1180 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1181 cause memory overwriting.
1183 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1184 string has been matched (to stop an infinite loop). It was not recognizing
1185 a conditional subpattern that could match an empty string if that
1186 subpattern was within another subpattern. For example, it looped when
1187 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1188 condition was not nested. This bug has been fixed.
1190 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1191 past the start of the subject in the presence of bytes with the top bit
1192 set, for example "\x8aBCD".
1194 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1195 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1197 14. Optimized (?!) to (*FAIL).
1199 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1200 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1201 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1202 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1203 does: it's just the validity check that is more restrictive.
1205 16. Inserted checks for integer overflows during escape sequence (backslash)
1206 processing, and also fixed erroneous offset values for syntax errors during
1207 backslash processing.
1209 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1210 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1212 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1213 caused an overrun.
1215 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1216 something other than just ASCII characters) inside a group that had an
1217 unlimited repeat caused a loop at compile time (while checking to see
1218 whether the group could match an empty string).
1220 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1221 [\P{Any}] did so. (Error in the code for printing property names.)
1223 21. An orphan \E inside a character class could cause a crash.
1225 22. A repeated capturing bracket such as (A)? could cause a wild memory
1226 reference during compilation.
1228 23. There are several functions in pcre_compile() that scan along a compiled
1229 expression for various reasons (e.g. to see if it's fixed length for look
1230 behind). There were bugs in these functions when a repeated \p or \P was
1231 present in the pattern. These operators have additional parameters compared
1232 with \d, etc, and these were not being taken into account when moving along
1233 the compiled data. Specifically:
1235 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1236 length.
1238 (b) An item such as \pL+ within a repeated group could cause crashes or
1239 loops.
1241 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1242 "reference to non-existent subpattern" error.
1244 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1246 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1247 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1249 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1251 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1252 character were causing crashes (broken optimization).
1254 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1255 \p or \P) caused a compile-time loop.
1257 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1258 break. For example, the unstudied pattern [\r\n]A does not match the string
1259 "\r\nA" because change 7.0/46 below moves the current point on by two
1260 characters after failing to match at the start. However, the pattern \nA
1261 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1262 the same is true. There doesn't seem any very clean way out of this, but
1263 what I have chosen to do makes the common cases work: PCRE now takes note
1264 of whether there can be an explicit match for \r or \n anywhere in the
1265 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1266 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1267 pattern has explicit CR or LF references.
1269 29. Added (*CR) etc for changing newline setting at start of pattern.
1272 Version 7.2 19-Jun-07
1273 ---------------------
1275 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1276 which is apparently normally available under Windows.
1278 2. Re-jig the pcregrep tests with different newline settings in an attempt
1279 to make them independent of the local environment's newline setting.
1281 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1283 4. Some of the "internals" tests were previously cut out when the link size
1284 was not 2, because the output contained actual offsets. The recent new
1285 "Z" feature of pcretest means that these can be cut out, making the tests
1286 usable with all link sizes.
1288 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1289 stack recursion. This gives a massive performance boost under BSD, but just
1290 a small improvement under Linux. However, it saves one field in the frame
1291 in all cases.
1293 6. Added more features from the forthcoming Perl 5.10:
1295 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1296 recursion call. It refers to the nth most recently opened parentheses.
1298 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1299 to be opened parentheses.
1301 (c) Conditions that refer to capturing parentheses can be specified
1302 relatively, for example, (?(-2)... or (?(+3)...
1304 (d) \K resets the start of the current match so that everything before
1305 is not part of it.
1307 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1309 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1310 reference syntax.
1312 (g) (?| introduces a group in which the numbering of parentheses in each
1313 alternative starts with the same number.
1315 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1317 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1320 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1321 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1322 for detecting groups that can match an empty string.
1324 9. A pattern with a very large number of alternatives (more than several
1325 hundred) was running out of internal workspace during the pre-compile
1326 phase, where pcre_compile() figures out how much memory will be needed. A
1327 bit of new cunning has reduced the workspace needed for groups with
1328 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1329 workspace instead of running out of the 4096 that are available.
1331 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1333 11. Applied patch from Google to remove an optimization that didn't quite work.
1334 The report of the bug said:
1336 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1337 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1338 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1340 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1341 it matched the wrong number of bytes.
1344 Version 7.1 24-Apr-07
1345 ---------------------
1347 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1348 that is more "standard", making use of automake and other Autotools. There
1349 is some re-arrangement of the files and adjustment of comments consequent
1350 on this.
1352 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1353 for recursive directory scanning broke on some systems because the files
1354 are not scanned in any specific order and on different systems the order
1355 was different. A call to "sort" has been inserted into RunGrepTest for the
1356 approprate test as a short-term fix. In the longer term there may be an
1357 alternative.
1359 3. I had an email from Eric Raymond about problems translating some of PCRE's
1360 man pages to HTML (despite the fact that I distribute HTML pages, some
1361 people do their own conversions for various reasons). The problems
1362 concerned the use of low-level troff macros .br and .in. I have therefore
1363 removed all such uses from the man pages (some were redundant, some could
1364 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1365 HTML has been updated to handle .nf/.fi and to complain if it encounters
1366 .br or .in.
1368 4. Updated comments in configure.ac that get placed in config.h.in and also
1369 arranged for config.h to be included in the distribution, with the name
1370 config.h.generic, for the benefit of those who have to compile without
1371 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1373 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1374 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1375 makevp.bat for the current PCRE, using the additional files
1376 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1378 6. A Windows user reported a minor discrepancy with test 2, which turned out
1379 to be caused by a trailing space on an input line that had got lost in his
1380 copy. The trailing space was an accident, so I've just removed it.
1382 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1383 that is needed.
1385 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1386 as "const" (a) because they are and (b) because it helps the PHP
1387 maintainers who have recently made a script to detect big data structures
1388 in the php code that should be moved to the .rodata section. I remembered
1389 to update Builducptable as well, so it won't revert if ucptable.h is ever
1390 re-created.
1392 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1393 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1394 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1395 support is not required. This saves 1.5-2K of code, which is important in
1396 some applications.
1398 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1399 so as not to refer to the tables, even though these functions will never be
1400 called when UTF-8 support is disabled. Otherwise there are problems with a
1401 shared library.
1403 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1405 (a) It was defining its arguments as char * instead of void *.
1407 (b) It was assuming that all moves were upwards in memory; this was true
1408 a long time ago when I wrote it, but is no longer the case.
1410 The emulated memove() is provided for those environments that have neither
1411 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1412 is clearly not the case, as these two bugs were recently reported.
1414 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1415 and Detrail to create the HTML documentation, the .txt form of the man
1416 pages, and it removes trailing spaces from listed files. It also creates
1417 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1418 case, it wraps all the #defines with #ifndefs. This script should be run
1419 before "make dist".
1421 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1422 with Unicode property support.
1424 (a) For a maximizing quantifier, if the two different cases of the
1425 character were of different lengths in their UTF-8 codings (there are
1426 some cases like this - I found 11), and the matching function had to
1427 back up over a mixture of the two cases, it incorrectly assumed they
1428 were both the same length.
1430 (b) When PCRE was configured to use the heap rather than the stack for
1431 recursion during matching, it was not correctly preserving the data for
1432 the other case of a UTF-8 character when checking ahead for a match
1433 while processing a minimizing repeat. If the check also involved
1434 matching a wide character, but failed, corruption could cause an
1435 erroneous result when trying to check for a repeat of the original
1436 character.
1438 13. Some tidying changes to the testing mechanism:
1440 (a) The RunTest script now detects the internal link size and whether there
1441 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1442 values substituted by "configure". (The RunGrepTest script already did
1443 this for UTF-8.) The configure.ac script no longer substitutes the
1444 relevant variables.
1446 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1447 with length and offset values. This means that the output is different
1448 for different internal link sizes. Test 2 is skipped for link sizes
1449 other than 2 because of this, bypassing the problem. Unfortunately,
1450 there was also a test in test 3 (the locale tests) that used /B and
1451 failed for link sizes other than 2. Rather than cut the whole test out,
1452 I have added a new /Z option to pcretest that replaces the length and
1453 offset values with spaces. This is now used to make test 3 independent
1454 of link size. (Test 2 will be tidied up later.)
1456 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1457 segmentation fault instead of returning the appropriate error message.
1459 15. In multiline mode when the newline sequence was set to "any", the pattern
1460 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1461 This doesn't seem right; it now treats the CRLF combination as the line
1462 ending, and so does not match in that case. It's only a pattern such as ^$
1463 that would hit this one: something like ^ABC$ would have failed after \r
1464 and then tried again after \r\n.
1466 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1467 in an attempt to make files that differ only in their line terminators
1468 compare equal. This works on Linux.
1470 17. Under certain error circumstances pcregrep might try to free random memory
1471 as it exited. This is now fixed, thanks to valgrind.
1473 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1474 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1475 was because its rules for how to advance for /g after matching an empty
1476 string at the end of a line did not allow for this case. They now check for
1477 it specially.
1479 20. pcretest is supposed to handle patterns and data of any length, by
1480 extending its buffers when necessary. It was getting this wrong when the
1481 buffer for a data line had to be extended.
1483 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1484 CRLF as a newline sequence.
1486 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1487 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1488 I have nevertheless tidied it up.
1490 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1492 24. Added a man page for pcre-config.
1495 Version 7.0 19-Dec-06
1496 ---------------------
1498 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1499 moving to gcc 4.1.1.
1501 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1502 sys/time.h, which is documented as needed for this function. It doesn't
1503 seem to matter on Linux, but it showed up on some releases of OS X.
1505 3. It seems that there are systems where bytes whose values are greater than
1506 127 match isprint() in the "C" locale. The "C" locale should be the
1507 default when a C program starts up. In most systems, only ASCII printing
1508 characters match isprint(). This difference caused the output from pcretest
1509 to vary, making some of the tests fail. I have changed pcretest so that:
1511 (a) When it is outputting text in the compiled version of a pattern, bytes
1512 other than 32-126 are always shown as hex escapes.
1514 (b) When it is outputting text that is a matched part of a subject string,
1515 it does the same, unless a different locale has been set for the match
1516 (using the /L modifier). In this case, it uses isprint() to decide.
1518 4. Fixed a major bug that caused incorrect computation of the amount of memory
1519 required for a compiled pattern when options that changed within the
1520 pattern affected the logic of the preliminary scan that determines the
1521 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1522 that the computed length was too small. The symptoms of this bug were
1523 either the PCRE error "internal error: code overflow" from pcre_compile(),
1524 or a glibc crash with a message such as "pcretest: free(): invalid next
1525 size (fast)". Examples of patterns that provoked this bug (shown in
1526 pcretest format) are:
1528 /(?-x: )/x
1529 /(?x)(?-x: \s*#\s*)/
1530 /((?i)[\x{c0}])/8
1531 /(?i:[\x{c0}])/8
1533 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1534 is now done differently.
1536 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1537 wrapper classes; (b) implement a new function in the C++ scanner that is
1538 more efficient than the old way of doing things because it avoids levels of
1539 recursion in the regex matching; (c) add a paragraph to the documentation
1540 for the FullMatch() function.
1542 6. The escape sequence \n was being treated as whatever was defined as
1543 "newline". Not only was this contrary to the documentation, which states
1544 that \n is character 10 (hex 0A), but it also went horribly wrong when
1545 "newline" was defined as CRLF. This has been fixed.
1547 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1548 was being set to -1 for the "end of line" case (supposedly a value that no
1549 character can have). Though this value is never used (the check for end of
1550 line is "zero bytes in current character"), it caused compiler complaints.
1551 I've changed it to 0xffffffff.
1553 8. In pcre_version.c, the version string was being built by a sequence of
1554 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1555 string (as it is for production releases) called a macro with an empty
1556 argument. The C standard says the result of this is undefined. The gcc
1557 compiler treats it as an empty string (which was what was wanted) but it is
1558 reported that Visual C gives an error. The source has been hacked around to
1559 avoid this problem.
1561 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1562 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1563 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1564 of them did).
1566 10. Originally, pcretest opened its input and output without "b"; then I was
1567 told that "b" was needed in some environments, so it was added for release
1568 5.0 to both the input and output. (It makes no difference on Unix-like
1569 systems.) Later I was told that it is wrong for the input on Windows. I've
1570 now abstracted the modes into two macros, to make it easier to fiddle with
1571 them, and removed "b" from the input mode under Windows.
1573 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1575 12. Added -help and --help to pcretest as an official way of being reminded
1576 of the options.
1578 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1579 and pcrecpp.cc because they annoy compilers at high warning levels.
1581 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1583 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1584 scripts are not C programs :-) and which was not noticed because it works
1585 on Linux.
1587 16. pcretest is supposed to handle any length of pattern and data line (as one
1588 line or as a continued sequence of lines) by extending its input buffer if
1589 necessary. This feature was broken for very long pattern lines, leading to
1590 a string of junk being passed to pcre_compile() if the pattern was longer
1591 than about 50K.
1593 17. I have done a major re-factoring of the way pcre_compile() computes the
1594 amount of memory needed for a compiled pattern. Previously, there was code
1595 that made a preliminary scan of the pattern in order to do this. That was
1596 OK when PCRE was new, but as the facilities have expanded, it has become
1597 harder and harder to keep it in step with the real compile phase, and there
1598 have been a number of bugs (see for example, 4 above). I have now found a
1599 cunning way of running the real compile function in a "fake" mode that
1600 enables it to compute how much memory it would need, while actually only
1601 ever using a few hundred bytes of working memory and without too many
1602 tests of the mode. This should make future maintenance and development
1603 easier. A side effect of this work is that the limit of 200 on the nesting
1604 depth of parentheses has been removed (though this was never a serious
1605 limitation, I suspect). However, there is a downside: pcre_compile() now
1606 runs more slowly than before (30% or more, depending on the pattern). I
1607 hope this isn't a big issue. There is no effect on runtime performance.
1609 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1610 newline (only possible for the last line of a file) and it was a
1611 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1613 19. Added additional timing features to pcretest. (1) The -tm option now times
1614 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1615 separate command line item, by a number that specifies the number of
1616 repeats to use when timing. The default is 50000; this gives better
1617 precision, but takes uncomfortably long for very large patterns.
1619 20. Extended pcre_study() to be more clever in cases where a branch of a
1620 subpattern has no definite first character. For example, (a*|b*)[cd] would
1621 previously give no result from pcre_study(). Now it recognizes that the
1622 first character must be a, b, c, or d.
1624 21. There was an incorrect error "recursive call could loop indefinitely" if
1625 a subpattern (or the entire pattern) that was being tested for matching an
1626 empty string contained only one non-empty item after a nested subpattern.
1627 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1628 incorrectly, because the \d was being skipped in the check.
1630 22. The pcretest program now has a new pattern option /B and a command line
1631 option -b, which is equivalent to adding /B to every pattern. This causes
1632 it to show the compiled bytecode, without the additional information that
1633 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1634 is the same as /B/I).
1636 23. A new optimization is now able automatically to treat some sequences such
1637 as a*b as a*+b. More specifically, if something simple (such as a character
1638 or a simple class like \d) has an unlimited quantifier, and is followed by
1639 something that cannot possibly match the quantified thing, the quantifier
1640 is automatically "possessified".
1642 24. A recursive reference to a subpattern whose number was greater than 39
1643 went wrong under certain circumstances in UTF-8 mode. This bug could also
1644 have affected the operation of pcre_study().
1646 25. Realized that a little bit of performance could be had by replacing
1647 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1649 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1651 27. Possessive quantifiers such as a++ were previously implemented by turning
1652 them into atomic groups such as ($>a+). Now they have their own opcodes,
1653 which improves performance. This includes the automatically created ones
1654 from 23 above.
1656 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1657 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1658 the first matched character to be a colon. This applied both to named and
1659 numbered groups.
1661 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1663 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1664 building PCRE on Windows easier, so I have included it in the distribution.
1666 31. There is now a check in pcretest against a ridiculously large number being
1667 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1668 loop, the loop is abandoned.
1670 32. Forward references to subpatterns in conditions such as (?(2)...) where
1671 subpattern 2 is defined later cause pcre_compile() to search forwards in
1672 the pattern for the relevant set of parentheses. This search went wrong
1673 when there were unescaped parentheses in a character class, parentheses
1674 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1676 33. "Subroutine" calls and backreferences were previously restricted to
1677 referencing subpatterns earlier in the regex. This restriction has now
1678 been removed.
1680 34. Added a number of extra features that are going to be in Perl 5.10. On the
1681 whole, these are just syntactic alternatives for features that PCRE had
1682 previously implemented using the Python syntax or my own invention. The
1683 other formats are all retained for compatibility.
1685 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1686 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1687 also .NET compatible.
1689 (b) A recursion or subroutine call to a named group can now be defined as
1690 (?&name) as well as (?P>name).
1692 (c) A backreference to a named group can now be defined as \k<name> or
1693 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1694 5.10, are also .NET compatible.
1696 (d) A conditional reference to a named group can now use the syntax
1697 (?(<name>) or (?('name') as well as (?(name).
1699 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1700 groups (named and numbered) that are never evaluated inline, but can be
1701 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1702 is always false. There may be only one alternative in such a group.
1704 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1705 as the simple (?(R). The condition is true only if the most recent
1706 recursion is that of the given number or name. It does not search out
1707 through the entire recursion stack.
1709 (g) The escape \gN or \g{N} has been added, where N is a positive or
1710 negative number, specifying an absolute or relative reference.
1712 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1713 some "unreachable code" warnings.
1715 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1716 things, this adds five new scripts.
1718 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1719 There were also incompatibilities regarding the handling of \Q..\E inside
1720 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1721 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1723 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1724 matches an empty string, and forcibly breaks the loop. There were bugs in
1725 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1726 against aaaa the result was just "a" rather than "aaaa", for example. Two
1727 separate and independent bugs (that affected different cases) have been
1728 fixed.
1730 39. Refactored the code to abolish the use of different opcodes for small
1731 capturing bracket numbers. This is a tidy that I avoided doing when I
1732 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1733 The new approach is not only tidier, it makes it possible to reduce the
1734 memory needed to fix the previous bug (38).
1736 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1737 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1738 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1739 mode.
1741 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1742 report.
1744 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1745 copy construction and assignment in the C++ wrapper.
1747 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1748 couple of bugs that could have given wrong results in the "--newline=crlf"
1749 case.
1751 44. Added a number of casts and did some reorganization of signed/unsigned int
1752 variables following suggestions from Dair Grant. Also renamed the variable
1753 "this" as "item" because it is a C++ keyword.
1755 45. Arranged for dftables to add
1757 #include "pcre_internal.h"
1759 to pcre_chartables.c because without it, gcc 4.x may remove the array
1760 definition from the final binary if PCRE is built into a static library and
1761 dead code stripping is activated.
1763 46. For an unanchored pattern, if a match attempt fails at the start of a
1764 newline sequence, and the newline setting is CRLF or ANY, and the next two
1765 characters are CRLF, advance by two characters instead of one.
1768 Version 6.7 04-Jul-06
1769 ---------------------
1771 1. In order to handle tests when input lines are enormously long, pcretest has
1772 been re-factored so that it automatically extends its buffers when
1773 necessary. The code is crude, but this _is_ just a test program. The
1774 default size has been increased from 32K to 50K.
1776 2. The code in pcre_study() was using the value of the re argument before
1777 testing it for NULL. (Of course, in any sensible call of the function, it
1778 won't be NULL.)
1780 3. The memmove() emulation function in pcre_internal.h, which is used on
1781 systems that lack both memmove() and bcopy() - that is, hardly ever -
1782 was missing a "static" storage class specifier.
1784 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1785 containing an extended class (one that cannot be represented by a bitmap
1786 because it contains high-valued characters or Unicode property items, e.g.
1787 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1788 pattern, but PCRE should not loop if you do not (it no longer does).
1789 [Detail: two cases were found: (a) a repeated subpattern containing an
1790 extended class; (b) a recursive reference to a subpattern that followed a
1791 previous extended class. It wasn't skipping over the extended class
1792 correctly when UTF-8 mode was not set.]
1794 5. A negated single-character class was not being recognized as fixed-length
1795 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1796 compile error "lookbehind assertion is not fixed length".
1798 6. The RunPerlTest auxiliary script was showing an unexpected difference
1799 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1800 write a Perl script that can interpret lines of an input file either as
1801 byte characters or as UTF-8, which is what "perltest" was being required to
1802 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1803 can't do is switch easily at run time between having the "use utf8;" pragma
1804 or not. In the end, I fudged it by using the RunPerlTest script to insert
1805 "use utf8;" explicitly for the UTF-8 tests.
1807 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1808 the end of the subject string, contrary to the documentation and to what
1809 Perl does. This was true of both matching functions. Now it matches only at
1810 the start of the subject and immediately after *internal* newlines.
1812 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1813 a pointer to an int instead of a pointer to an unsigned long int. This
1814 caused problems on 64-bit systems.
1816 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1817 instance of the 'standard' template library not being so standard".
1819 10. There was no check on the number of named subpatterns nor the maximum
1820 length of a subpattern name. The product of these values is used to compute
1821 the size of the memory block for a compiled pattern. By supplying a very
1822 long subpattern name and a large number of named subpatterns, the size
1823 computation could be caused to overflow. This is now prevented by limiting
1824 the length of names to 32 characters, and the number of named subpatterns
1825 to 10,000.
1827 11. Subpatterns that are repeated with specific counts have to be replicated in
1828 the compiled pattern. The size of memory for this was computed from the
1829 length of the subpattern and the repeat count. The latter is limited to
1830 65535, but there was no limit on the former, meaning that integer overflow
1831 could in principle occur. The compiled length of a repeated subpattern is
1832 now limited to 30,000 bytes in order to prevent this.
1834 12. Added the optional facility to have named substrings with the same name.
1836 13. Added the ability to use a named substring as a condition, using the
1837 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1838 are numbers (not recommended). Forward references are permitted.
1840 14. Added forward references in named backreferences (if you see what I mean).
1842 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1843 pattern could run off the end of the subject. For example, the pattern
1844 "(?s)(.{1,5})"8 did this with the subject "ab".
1846 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1847 PCRE_CASELESS was set when matching characters that were quantified with ?
1848 or *.
1850 17. A character class other than a single negated character that had a minimum
1851 but no maximum quantifier - for example [ab]{6,} - was not handled
1852 correctly by pce_dfa_exec(). It would match only one character.
1854 18. A valid (though odd) pattern that looked like a POSIX character
1855 class but used an invalid character after [ (for example [[,abc,]]) caused
1856 pcre_compile() to give the error "Failed: internal error: code overflow" or
1857 in some cases to crash with a glibc free() error. This could even happen if
1858 the pattern terminated after [[ but there just happened to be a sequence of
1859 letters, a binary zero, and a closing ] in the memory that followed.
1861 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1862 over the years. Originally (before any Unicode support), just the bottom 8
1863 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1864 output from "man perlunicode" includes this:
1866 The regular expression compiler produces polymorphic opcodes. That
1867 is, the pattern adapts to the data and automatically switches to
1868 the Unicode character scheme when presented with Unicode data--or
1869 instead uses a traditional byte scheme when presented with byte
1870 data.
1872 Sadly, a wide octal escape does not cause a switch, and in a string with
1873 no other multibyte characters, these octal escapes are treated as before.
1874 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1875 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1876 Unicode string.
1878 I have not perpetrated such confusion in PCRE. Up till now, it took just
1879 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1880 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1881 translate to the appropriate multibyte character.
1883 29. Applied some refactoring to reduce the number of warnings from Microsoft
1884 and Borland compilers. This has included removing the fudge introduced
1885 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1886 a warning about an unused variable.
1888 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1889 characters since release 4.0, because Perl (from release 5.004) does not.
1890 [Or at least, is documented not to: some releases seem to be in conflict
1891 with the documentation.] However, when a pattern was studied with
1892 pcre_study() and all its branches started with \s, PCRE still included VT
1893 as a possible starting character. Of course, this did no harm; it just
1894 caused an unnecessary match attempt.
1896 22. Removed a now-redundant internal flag bit that recorded the fact that case
1897 dependency changed within the pattern. This was once needed for "required
1898 byte" processing, but is no longer used. This recovers a now-scarce options
1899 bit. Also moved the least significant internal flag bit to the most-
1900 significant bit of the word, which was not previously used (hangover from
1901 the days when it was an int rather than a uint) to free up another bit for
1902 the future.
1904 23. Added support for CRLF line endings as well as CR and LF. As well as the
1905 default being selectable at build time, it can now be changed at runtime
1906 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1907 specify that it is scanning data with non-default line endings.
1909 24. Changed the definition of CXXLINK to make it agree with the definition of
1910 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1912 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1913 recursions. This makes a big different to stack usage for some patterns.
1915 26. If a subpattern containing a named recursion or subroutine reference such
1916 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1917 the space required for the compiled pattern went wrong and gave too small a
1918 value. Depending on the environment, this could lead to "Failed: internal
1919 error: code overflow at offset 49" or "glibc detected double free or
1920 corruption" errors.
1922 27. Applied patches from Google (a) to support the new newline modes and (b) to
1923 advance over multibyte UTF-8 characters in GlobalReplace.
1925 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1926 difference for some implementation of PCRE in some Windows version.
1928 29. Added some extra testing facilities to pcretest:
1930 \q<number> in a data line sets the "match limit" value
1931 \Q<number> in a data line sets the "match recursion limt" value
1932 -S <number> sets the stack size, where <number> is in megabytes
1934 The -S option isn't available for Windows.
1937 Version 6.6 06-Feb-06
1938 ---------------------
1940 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1941 in pcreposix.h. I have copied the definition from pcre.h.
1943 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1944 because pcre.h is no longer a built file.
1946 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1947 not normally included in the compiled code.
1950 Version 6.5 01-Feb-06
1951 ---------------------
1953 1. When using the partial match feature with pcre_dfa_exec(), it was not
1954 anchoring the second and subsequent partial matches at the new starting
1955 point. This could lead to incorrect results. For example, with the pattern
1956 /1234/, partially matching against "123" and then "a4" gave a match.
1958 2. Changes to pcregrep:
1960 (a) All non-match returns from pcre_exec() were being treated as failures
1961 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1962 error message is output. Some extra information is given for the
1964 probably the only errors that are likely to be caused by users (by
1965 specifying a regex that has nested indefinite repeats, for instance).
1966 If there are more than 20 of these errors, pcregrep is abandoned.
1968 (b) A binary zero was treated as data while matching, but terminated the
1969 output line if it was written out. This has been fixed: binary zeroes
1970 are now no different to any other data bytes.
1972 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1973 used to set a locale for matching. The --locale=xxxx long option has
1974 been added (no short equivalent) to specify a locale explicitly on the
1975 pcregrep command, overriding the environment variables.
1977 (d) When -B was used with -n, some line numbers in the output were one less
1978 than they should have been.
1980 (e) Added the -o (--only-matching) option.
1982 (f) If -A or -C was used with -c (count only), some lines of context were
1983 accidentally printed for the final match.
1985 (g) Added the -H (--with-filename) option.
1987 (h) The combination of options -rh failed to suppress file names for files
1988 that were found from directory arguments.
1990 (i) Added the -D (--devices) and -d (--directories) options.
1992 (j) Added the -F (--fixed-strings) option.
1994 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1996 (l) Added the --colo(u)r option.
1998 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1999 is not present by default.
2001 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2002 items such as (?R) or (?1), when the recursion could match a number of
2003 alternatives. If it matched one of the alternatives, but subsequently,
2004 outside the recursion, there was a failure, the code tried to back up into
2005 the recursion. However, because of the way PCRE is implemented, this is not
2006 possible, and the result was an incorrect result from the match.
2008 In order to prevent this happening, the specification of recursion has
2009 been changed so that all such subpatterns are automatically treated as
2010 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2012 4. I had overlooked the fact that, in some locales, there are characters for
2013 which isalpha() is true but neither isupper() nor islower() are true. In
2014 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2015 and ordfeminine) are like this. This affected the treatment of \w and \W
2016 when they appeared in character classes, but not when they appeared outside
2017 a character class. The bit map for "word" characters is now created
2018 separately from the results of isalnum() instead of just taking it from the
2019 upper, lower, and digit maps. (Plus the underscore character, of course.)
2021 5. The above bug also affected the handling of POSIX character classes such as
2022 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2023 permanent tables. Instead, the bit maps for such a class were previously
2024 created as the appropriate unions of the upper, lower, and digit bitmaps.
2025 Now they are created by subtraction from the [[:word:]] class, which has
2026 its own bitmap.
2028 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2029 It is created by subtracting the vertical space characters (\x09, \x0a,
2030 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2031 subtraction was done in the overall bitmap for a character class, meaning
2032 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2033 be recognized. This bug has been fixed.
2035 7. Patches from the folks at Google:
2037 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2038 real life, but is still worth protecting against".
2040 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2041 regular expressions".
2043 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2044 have it.
2046 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2047 "configure" and the latter not, in order to fix a problem somebody had
2048 with compiling the Arg class on HP-UX.
2050 (e) Improve the error-handling of the C++ wrapper a little bit.
2052 (f) New tests for checking recursion limiting.
2054 8. The pcre_memmove() function, which is used only if the environment does not
2055 have a standard memmove() function (and is therefore rarely compiled),
2056 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2057 returning a result (though PCRE never actually uses the result).
2059 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2060 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2061 returned instead of calling malloc() with an overflowing number that would
2062 most likely cause subsequent chaos.
2064 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2066 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2067 with this option is matched, the nmatch and pmatch options of regexec() are
2068 ignored.
2070 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2071 provided in case anyone wants to the the POSIX interface with UTF-8
2072 strings.
2074 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2075 C++ linking (needed for some HP-UX environments).
2077 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2078 (unused parameter) and in the pcre_printint() function (omitted "default"
2079 switch label when the default is to do nothing).
2081 15. Added some code to make it possible, when PCRE is compiled as a C++
2082 library, to replace subject pointers for pcre_exec() with a smart pointer
2083 class, thus making it possible to process discontinuous strings.
2085 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2086 much the same function. They were added by different people who were trying
2087 to make PCRE easy to compile on non-Unix systems. It has been suggested
2088 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2089 for compiling on Windows systems. I have therefore replaced it with
2090 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2091 defaults to "extern" for C or "extern C" for C++, which works fine on
2092 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2093 SCOPE with something explicit in config.h. In addition:
2095 (a) pcreposix.h still had just "extern" instead of either of these macros;
2096 I have replaced it with PCRE_DATA_SCOPE.
2098 (b) Functions such as _pcre_xclass(), which are internal to the library,
2099 but external in the C sense, all had PCRE_EXPORT in their definitions.
2100 This is apparently wrong for the Windows case, so I have removed it.
2101 (It makes no difference on Unix-like systems.)
2103 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2104 of recursive calls to match(). This is different to MATCH_LIMIT because
2105 that limits the total number of calls to match(), not all of which increase
2106 the depth of recursion. Limiting the recursion depth limits the amount of
2107 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2108 when PCRE is compiled, and changed at run time. A patch from Google adds
2109 this functionality to the C++ interface.
2111 18. Changes to the handling of Unicode character properties:
2113 (a) Updated the table to Unicode 4.1.0.
2115 (b) Recognize characters that are not in the table as "Cn" (undefined).
2117 (c) I revised the way the table is implemented to a much improved format
2118 which includes recognition of ranges. It now supports the ranges that
2119 are defined in UnicodeData.txt, and it also amalgamates other
2120 characters into ranges. This has reduced the number of entries in the
2121 table from around 16,000 to around 3,000, thus reducing its size
2122 considerably. I realized I did not need to use a tree structure after
2123 all - a binary chop search is just as efficient. Having reduced the
2124 number of entries, I extended their size from 6 bytes to 8 bytes to
2125 allow for more data.
2127 (d) Added support for Unicode script names via properties such as \p{Han}.
2129 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2130 matching that character.
2132 20. When matching a repeated Unicode property with a minimum greater than zero,
2133 (for example \pL{2,}), PCRE could look past the end of the subject if it
2134 reached it while seeking the minimum number of characters. This could
2135 happen only if some of the characters were more than one byte long, because
2136 there is a check for at least the minimum number of bytes.
2138 21. Refactored the implementation of \p and \P so as to be more general, to
2139 allow for more different types of property in future. This has changed the
2140 compiled form incompatibly. Anybody with saved compiled patterns that use
2141 \p or \P will have to recompile them.
2143 22. Added "Any" and "L&" to the supported property types.
2145 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2146 but give a compile time error if the value is greater than 0xff.
2148 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2149 accidentally not being installed or uninstalled.
2151 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2152 made were to insert the current release number. This seemed silly, because
2153 it made things harder for people building PCRE on systems that don't run
2154 "configure". I have turned pcre.h into a distributed file, no longer built
2155 by "configure", with the version identification directly included. There is
2156 no longer a pcre.h.in file.
2158 However, this change necessitated a change to the pcre-config script as
2159 well. It is built from pcre-config.in, and one of the substitutions was the
2160 release number. I have updated configure.ac so that ./configure now finds
2161 the release number by grepping pcre.h.
2163 26. Added the ability to run the tests under valgrind.
2166 Version 6.4 05-Sep-05
2167 ---------------------
2169 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2170 "--" to be printed when multiple files were scanned, even when none of the
2171 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2172 consider it to be a bug, and have restored the previous behaviour.
2174 2. A couple of code tidies to get rid of compiler warnings.
2176 3. The pcretest program used to cheat by referring to symbols in the library
2177 whose names begin with _pcre_. These are internal symbols that are not
2178 really supposed to be visible externally, and in some environments it is
2179 possible to suppress them. The cheating is now confined to including
2180 certain files from the library's source, which is a bit cleaner.
2182 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2183 file's purpose clearer.
2185 5. Reorganized pcre_ucp_findchar().
2188 Version 6.3 15-Aug-05
2189 ---------------------
2191 1. The file libpcre.pc.in did not have general read permission in the tarball.
2193 2. There were some problems when building without C++ support:
2195 (a) If C++ support was not built, "make install" and "make test" still
2196 tried to test it.
2198 (b) There were problems when the value of CXX was explicitly set. Some
2199 changes have been made to try to fix these, and ...
2201 (c) --disable-cpp can now be used to explicitly disable C++ support.
2203 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2204 backslash in a target when C++ was disabled. This confuses some
2205 versions of "make", apparently. Using an intermediate variable solves
2206 this. (Same for CPP_LOBJ.)
2208 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2209 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2210 necessary on certain architectures.
2212 4. Added a setting of -export-symbols-regex to the link command to remove
2213 those symbols that are exported in the C sense, but actually are local
2214 within the library, and not documented. Their names all begin with
2215 "_pcre_". This is not a perfect job, because (a) we have to except some
2216 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2217 available (and never for static libraries). I have made a note to try to
2218 find a way round (a) in the future.
2221 Version 6.2 01-Aug-05
2222 ---------------------
2224 1. There was no test for integer overflow of quantifier values. A construction
2225 such as {1111111111111111} would give undefined results. What is worse, if
2226 a minimum quantifier for a parenthesized subpattern overflowed and became
2227 negative, the calculation of the memory size went wrong. This could have
2228 led to memory overwriting.
2230 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2232 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2233 operating environments where this matters.
2235 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2236 PCRE options from within the C++ wrapper.
2238 5. Named capturing subpatterns were not being correctly counted when a pattern
2239 was compiled. This caused two problems: (a) If there were more than 100
2240 such subpatterns, the calculation of the memory needed for the whole
2241 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2242 back references of the form \12, where the number was greater than 9, were
2243 not recognized as back references, even though there were sufficient
2244 previous subpatterns.
2246 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2247 versions of gcc, e.g. 2.95.4.
2250 Version 6.1 21-Jun-05
2251 ---------------------
2253 1. There was one reference to the variable "posix" in pcretest.c that was not
2254 surrounded by "#if !defined NOPOSIX".
2256 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2257 the cross-check on the old pcre_info() function, for the benefit of the
2258 cut-down version of PCRE that is currently imported into Exim.
2260 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2261 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2262 patterns that start (?i) but I don't think that matters. The use of (?i) is
2263 just an example; this all applies to the other options as well.
2265 4. Since libtool seems to echo the compile commands it is issuing, the output
2266 from "make" can be reduced a bit by putting "@" in front of each libtool
2267 compile command.
2269 5. Patch from the folks at Google for configure.in to be a bit more thorough
2270 in checking for a suitable C++ installation before trying to compile the
2271 C++ stuff. This should fix a reported problem when a compiler was present,
2272 but no suitable headers.
2274 6. The man pages all had just "PCRE" as their title. I have changed them to
2275 be the relevant file name. I have also arranged that these names are
2276 retained in the file doc/pcre.txt, which is a concatenation in text format
2277 of all the man pages except the little individual ones for each function.
2279 7. The NON-UNIX-USE file had not been updated for the different set of source
2280 files that come with release 6. I also added a few comments about the C++
2281 wrapper.
2284 Version 6.0 07-Jun-05
2285 ---------------------
2287 1. Some minor internal re-organization to help with my DFA experiments.
2289 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2290 didn't matter for the library itself when fully configured, but did matter
2291 when compiling without UCP support, or within Exim, where the ucp files are
2292 not imported.
2294 3. Refactoring of the library code to split up the various functions into
2295 different source modules. The addition of the new DFA matching code (see
2296 below) to a single monolithic source would have made it really too
2297 unwieldy, quite apart from causing all the code to be include in a
2298 statically linked application, when only some functions are used. This is
2299 relevant even without the DFA addition now that patterns can be compiled in
2300 one application and matched in another.
2302 The downside of splitting up is that there have to be some external
2303 functions and data tables that are used internally in different modules of
2304 the library but which are not part of the API. These have all had their
2305 names changed to start with "_pcre_" so that they are unlikely to clash
2306 with other external names.
2308 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2309 a different (DFA) algorithm. Although it is slower than the original
2310 function, it does have some advantages for certain types of matching
2311 problem.
2313 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2314 including restarting after a partial match.
2316 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2317 defined when compiling for Windows was sent to me. I have put it into the
2318 code, though I have no means of testing or verifying it.
2320 7. Added the pcre_refcount() auxiliary function.
2322 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2323 match before or at the first newline in the subject string. In pcretest,
2324 the /f option on a pattern can be used to set this.
2326 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2327 would behave wrongly. This has been present in PCRE since release 4.0.
2329 10. A number of changes to the pcregrep command:
2331 (a) Refactored how -x works; insert ^(...)$ instead of setting
2332 PCRE_ANCHORED and checking the length, in preparation for adding
2333 something similar for -w.
2335 (b) Added the -w (match as a word) option.
2337 (c) Refactored the way lines are read and buffered so as to have more
2338 than one at a time available.
2340 (d) Implemented a pcregrep test script.
2342 (e) Added the -M (multiline match) option. This allows patterns to match
2343 over several lines of the subject. The buffering ensures that at least
2344 8K, or the rest of the document (whichever is the shorter) is available
2345 for matching (and similarly the previous 8K for lookbehind assertions).
2347 (f) Changed the --help output so that it now says
2349 -w, --word-regex(p)
2351 instead of two lines, one with "regex" and the other with "regexp"
2352 because that confused at least one person since the short forms are the
2353 same. (This required a bit of code, as the output is generated
2354 automatically from a table. It wasn't just a text change.)
2356 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2357 option but starts with a hyphen. Could be a pattern or a path name
2358 starting with a hyphen, for instance.
2360 (h) "-" can be given as a file name to represent stdin.
2362 (i) When file names are being printed, "(standard input)" is used for
2363 the standard input, for compatibility with GNU grep. Previously
2364 "<stdin>" was used.
2366 (j) The option --label=xxx can be used to supply a name to be used for
2367 stdin when file names are being printed. There is no short form.
2369 (k) Re-factored the options decoding logic because we are going to add
2370 two more options that take data. Such options can now be given in four
2371 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2373 (l) Added the -A, -B, and -C options for requesting that lines of context
2374 around matches be printed.
2376 (m) Added the -L option to print the names of files that do not contain
2377 any matching lines, that is, the complement of -l.
2379 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2380 continue to scan other files.
2382 (o) The -s option was incorrectly implemented. For compatibility with other
2383 greps, it now suppresses the error message for a non-existent or non-
2384 accessible file (but not the return code). There is a new option called
2385 -q that suppresses the output of matching lines, which was what -s was
2386 previously doing.
2388 (p) Added --include and --exclude options to specify files for inclusion
2389 and exclusion when recursing.
2391 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2392 Hopefully, it now does.
2394 12. Missing cast in pcre_study().
2396 13. Added an "uninstall" target to the makefile.
2398 14. Replaced "extern" in the function prototypes in Makefile.in with
2399 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2400 world, but is set differently for Windows.
2402 15. Added a second compiling function called pcre_compile2(). The only
2403 difference is that it has an extra argument, which is a pointer to an
2404 integer error code. When there is a compile-time failure, this is set
2405 non-zero, in addition to the error test pointer being set to point to an
2406 error message. The new argument may be NULL if no error number is required
2407 (but then you may as well call pcre_compile(), which is now just a
2408 wrapper). This facility is provided because some applications need a
2409 numeric error indication, but it has also enabled me to tidy up the way
2410 compile-time errors are handled in the POSIX wrapper.
2412 16. Added VPATH=.libs to the makefile; this should help when building with one
2413 prefix path and installing with another. (Or so I'm told by someone who
2414 knows more about this stuff than I do.)
2416 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2417 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2418 match everything, including newlines. This is not POSIX-compatible, but
2419 somebody wanted the feature. From pcretest it can be activated by using
2420 both the P and the s flags.
2422 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2424 19. libpcre.pc was being incorrectly installed as executable.
2426 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2427 it now also looks for '\r' so that it will work unmodified on Windows.
2429 21. Added Google's contributed C++ wrapper to the distribution.
2431 22. Added some untidy missing memory free() calls in pcretest, to keep
2432 Electric Fence happy when testing.
2436 Version 5.0 13-Sep-04
2437 ---------------------
2439 1. Internal change: literal characters are no longer packed up into items
2440 containing multiple characters in a single byte-string. Each character
2441 is now matched using a separate opcode. However, there may be more than one
2442 byte in the character in UTF-8 mode.
2444 2. The pcre_callout_block structure has two new fields: pattern_position and
2445 next_item_length. These contain the offset in the pattern to the next match
2446 item, and its length, respectively.
2448 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2449 insertion of callouts before each pattern item. Added the /C option to
2450 pcretest to make use of this.
2452 4. On the advice of a Windows user, the lines
2454 #if defined(_WIN32) || defined(WIN32)
2455 _setmode( _fileno( stdout ), 0x8000 );
2456 #endif /* defined(_WIN32) || defined(WIN32) */
2458 have been added to the source of pcretest. This apparently does useful
2459 magic in relation to line terminators.
2461 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2462 for the benefit of those environments where the "b" makes a difference.
2464 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2465 to know about it. I have put a hack into configure.in that adds in code
2466 to set GCC=yes if CC=icc. This seems to end up at a point in the
2467 generated configure script that is early enough to affect the setting of
2468 compiler options, which is what is needed, but I have no means of testing
2469 whether it really works. (The user who reported this had patched the
2470 generated configure script, which of course I cannot do.)
2472 LATER: After change 22 below (new libtool files), the configure script
2473 seems to know about icc (and also ecc). Therefore, I have commented out
2474 this hack in configure.in.
2476 7. Added support for pkg-config (2 patches were sent in).
2478 8. Negated POSIX character classes that used a combination of internal tables
2479 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2480 [[:^ascii]]. Typically, they would match almost any characters. The other
2481 POSIX classes were not broken in this way.
2483 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2484 to find the match, as PCRE was deluded into thinking that the match had to
2485 start at the start point or following a newline. The same bug applied to
2486 patterns with negative forward assertions or any backward assertions
2487 preceding ".*" at the start, unless the pattern required a fixed first
2488 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2490 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2491 starting at the last subject character, bytes beyond the end of the subject
2492 string were read.
2494 11. Renamed the variable "class" as "classbits" to make life easier for C++
2495 users. (Previously there was a macro definition, but it apparently wasn't
2496 enough.)
2498 12. Added the new field "tables" to the extra data so that tables can be passed
2499 in at exec time, or the internal tables can be re-selected. This allows
2500 a compiled regex to be saved and re-used at a later time by a different
2501 program that might have everything at different addresses.
2503 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2504 -R library as well as a -L library.
2506 14. The debugging options of pcretest (-d on the command line or D on a
2507 pattern) showed incorrect output for anything following an extended class
2508 that contained multibyte characters and which was followed by a quantifier.
2510 15. Added optional support for general category Unicode character properties
2511 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2512 support. It adds about 90K to the size of the library. The meanings of the
2513 inbuilt class escapes such as \d and \s have NOT been changed.
2515 16. Updated pcredemo.c to include calls to free() to release the memory for the
2516 compiled pattern.
2518 17. The generated file chartables.c was being created in the source directory
2519 instead of in the building directory. This caused the build to fail if the
2520 source directory was different from the building directory, and was
2521 read-only.
2523 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2524 file. No doubt somebody will tell me if they don't make sense... Also added
2525 Dan Mooney's comments about building on OpenVMS.
2527 19. Added support for partial matching via the PCRE_PARTIAL option for
2528 pcre_exec() and the \P data escape in pcretest.
2530 20. Extended pcretest with 3 new pattern features:
2532 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2533 write the compiled pattern to the file whose name is "rest-of-line".
2534 This is a straight binary dump of the data, with the saved pointer to
2535 the character tables forced to be NULL. The study data, if any, is
2536 written too. After writing, pcretest reads a new pattern.
2538 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2539 compiled pattern from the given file. There must not be any
2540 occurrences of "<" in the file name (pretty unlikely); if there are,
2541 pcretest will instead treat the initial "<" as a pattern delimiter.
2542 After reading in the pattern, pcretest goes on to read data lines as
2543 usual.
2545 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2546 and 16-bit fields in a compiled pattern, to simulate a pattern that
2547 was compiled on a host of opposite endianness.
2549 21. The pcre-exec() function can now cope with patterns that were compiled on
2550 hosts of opposite endianness, with this restriction:
2552 As for any compiled expression that is saved and used later, the tables
2553 pointer field cannot be preserved; the extra_data field in the arguments
2554 to pcre_exec() should be used to pass in a tables address if a value
2555 other than the default internal tables were used at compile time.
2557 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2558 now diagnosed as an error. Previously, most of the time, a negative number
2559 would have been treated as zero, but if in addition "ovector" was passed as
2560 NULL, a crash could occur.
2562 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2563 new versions from the libtool 1.5 distribution (the last one is a copy of
2564 a file called libtool.m4). This seems to have fixed the need to patch
2565 "configure" to support Darwin 1.3 (which I used to do). However, I still
2566 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2567 workstation).
2569 24. Changed the PCRE licence to be the more standard "BSD" licence.
2572 Version 4.5 01-Dec-03
2573 ---------------------
2575 1. There has been some re-arrangement of the code for the match() function so
2576 that it can be compiled in a version that does not call itself recursively.
2577 Instead, it keeps those local variables that need separate instances for
2578 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2579 needs to "recurse". Keeping track of where control must go is done by means
2580 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2581 hide most of the details from the main code, and operates only if
2582 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2583 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2584 operating.
2586 To make it easier for callers to provide specially tailored get/free
2587 functions for this usage, two new functions, pcre_stack_malloc, and
2588 pcre_stack_free, are used. They are always called in strict stacking order,
2589 and the size of block requested is always the same.
2591 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2592 PCRE has been compiled to use the stack or the heap for recursion. The
2593 -C option of pcretest uses this to show which version is compiled.
2595 A new data escape \S, is added to pcretest; it causes the amounts of store
2596 obtained and freed by both kinds of malloc/free at match time to be added
2597 to the output.
2599 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2600 what's available on my current Linux desktop machine.
2602 3. When matching a UTF-8 string, the test for a valid string at the start has
2603 been extended. If start_offset is not zero, PCRE now checks that it points
2604 to a byte that is the start of a UTF-8 character. If not, it returns
2605 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2606 this is necessary because there may be backward assertions in the pattern.
2607 When matching the same subject several times, it may save resources to use
2608 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2610 4. The code for checking the validity of UTF-8 strings has been tightened so
2611 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2612 containing "overlong sequences".
2614 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2615 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2616 should have been "&", but it just so happened that all the cases this let
2617 through by mistake were picked up later in the function.
2619 6. I had used a variable called "isblank" - this is a C99 function, causing
2620 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2622 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2623 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2624 so the version is shown; (c) stop "make test" echoing "./RunTest".
2626 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2628 9. The prototype for memmove() for systems that don't have it was using
2629 size_t, but the inclusion of the header that defines size_t was later. I've
2630 moved the #includes for the C headers earlier to avoid this.
2632 10. Added some adjustments to the code to make it easier to compiler on certain
2633 special systems:
2635 (a) Some "const" qualifiers were missing.
2636 (b) Added the macro EXPORT before all exported functions; by default this
2637 is defined to be empty.
2638 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2639 that it reads its output file name as an argument instead of writing
2640 to the standard output and assuming this can be redirected.
2642 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2643 class containing characters with values greater than 255, PCRE compilation
2644 went into a loop.
2646 12. A recursive reference to a subpattern that was within another subpattern
2647 that had a minimum quantifier of zero caused PCRE to crash. For example,
2648 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2649 recursion. If the recursively-called subpattern itself had a zero repeat,
2650 that was OK.
2652 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2653 buffer into which it was copied (for escape processing) was still set at
2654 1024, so long lines caused crashes.
2656 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2657 "internal error: code overflow...". This applied to any character class
2658 that was followed by a possessive quantifier.
2660 15. Modified the Makefile to add libpcre.la as a prerequisite for
2661 libpcreposix.la because I was told this is needed for a parallel build to
2662 work.
2664 16. If a pattern that contained .* following optional items at the start was
2665 studied, the wrong optimizing data was generated, leading to matching
2666 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2667 matching string must start with a or b or c. The correct conclusion for
2668 this pattern is that a match can start with any character.
2671 Version 4.4 13-Aug-03
2672 ---------------------
2674 1. In UTF-8 mode, a character class containing characters with values between
2675 127 and 255 was not handled correctly if the compiled pattern was studied.
2676 In fixing this, I have also improved the studying algorithm for such
2677 classes (slightly).
2679 2. Three internal functions had redundant arguments passed to them. Removal
2680 might give a very teeny performance improvement.
2682 3. Documentation bug: the value of the capture_top field in a callout is *one
2683 more than* the number of the hightest numbered captured substring.
2685 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2686 in incorrectly linking with a previously installed version. They now link
2687 explicitly with libpcre.la.
2689 5. configure.in no longer needs to recognize Cygwin specially.
2691 6. A problem in pcre.in for Windows platforms is fixed.
2693 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2694 pcretest, it used to include the size of the study block as part of its
2695 output. Unfortunately, the structure contains a field that has a different
2696 size on different hardware architectures. This meant that the tests that
2697 showed this size failed. As the block is currently always of a fixed size,
2698 this information isn't actually particularly useful in pcretest output, so
2699 I have just removed it.
2701 8. Three pre-processor statements accidentally did not start in column 1.
2702 Sadly, there are *still* compilers around that complain, even though
2703 standard C has not required this for well over a decade. Sigh.
2705 9. In pcretest, the code for checking callouts passed small integers in the
2706 callout_data field, which is a void * field. However, some picky compilers
2707 complained about the casts involved for this on 64-bit systems. Now
2708 pcretest passes the address of the small integer instead, which should get
2709 rid of the warnings.
2711 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2712 both compile and run time, and gives an error if an invalid UTF-8 sequence
2713 is found. There is a option for disabling this check in cases where the
2714 string is known to be correct and/or the maximum performance is wanted.
2716 11. In response to a bug report, I changed one line in Makefile.in from
2718 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2719 to
2720 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2722 to look similar to other lines, but I have no way of telling whether this
2723 is the right thing to do, as I do not use Windows. No doubt I'll get told
2724 if it's wrong...
2727 Version 4.3 21-May-03
2728 ---------------------
2730 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2731 Makefile.
2733 2. Some refactoring to improve the quality of the code:
2735 (i) The utf8_table... variables are now declared "const".
2737 (ii) The code for \cx, which used the "case flipping" table to upper case
2738 lower case letters, now just substracts 32. This is ASCII-specific,
2739 but the whole concept of \cx is ASCII-specific, so it seems
2740 reasonable.
2742 (iii) PCRE was using its character types table to recognize decimal and
2743 hexadecimal digits in the pattern. This is silly, because it handles
2744 only 0-9, a-f, and A-F, but the character types table is locale-
2745 specific, which means strange things might happen. A private
2746 table is now used for this - though it costs 256 bytes, a table is
2747 much faster than multiple explicit tests. Of course, the standard
2748 character types table is still used for matching digits in subject
2749 strings against \d.
2751 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2752 ending in _t are). So I've renamed it as ESC_tee.
2754 3. The first argument for regexec() in the POSIX wrapper should have been
2755 defined as "const".
2757 4. Changed pcretest to use malloc() for its buffers so that they can be
2758 Electric Fenced for debugging.
2760 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2761 to read one or more bytes before the start of the subject string. Often this
2762 had no effect on PCRE's behaviour, but in some circumstances it could
2763 provoke a segmentation fault.
2765 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2766 to try to read one or more bytes before the start of the subject string.
2768 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2769 UTF-8 support could misbehave in various ways if the subject string
2770 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2771 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2772 back over UTF-8 characters.)
2775 Version 4.2 14-Apr-03
2776 ---------------------
2778 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2780 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2781 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2782 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2783 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2784 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2786 Note: automatic setting of the BUILD variables is not yet working
2787 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2788 compile-time but not at link-time
2789 [LINK]: use for linking executables only
2790 make different versions for Windows and non-Windows
2791 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2792 libraries
2793 [LINK_FOR_BUILD]: new variable
2794 [OBJEXT]: use throughout
2795 [EXEEXT]: use throughout
2796 <winshared>: new target
2797 <wininstall>: new target
2798 <dftables.o>: use native compiler
2799 <dftables>: use native linker
2800 <install>: handle Windows platform correctly
2801 <clean>: ditto
2802 <check>: ditto
2803 copy DLL to top builddir before testing
2805 As part of these changes, -no-undefined was removed again. This was reported
2806 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2807 in any case.
2809 3. Some tidies to get rid of compiler warnings:
2811 . In the match_data structure, match_limit was an unsigned long int, whereas
2812 match_call_count was an int. I've made them both unsigned long ints.
2814 . In pcretest the fact that a const uschar * doesn't automatically cast to
2815 a void * provoked a warning.
2817 . Turning on some more compiler warnings threw up some "shadow" variables
2818 and a few more missing casts.
2820 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2821 option, a class that contained a single character with a value between 128
2822 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2824 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2825 option, a class that contained several characters, but with at least one
2826 whose value was between 128 and 255 caused PCRE to crash.
2829 Version 4.1 12-Mar-03
2830 ---------------------
2832 1. Compiling with gcc -pedantic found a couple of places where casts were
2833 needed, and a string in dftables.c that was longer than standard compilers are
2834 required to support.
2836 2. Compiling with Sun's compiler found a few more places where the code could
2837 be tidied up in order to avoid warnings.
2839 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2840 first of these names is deprecated in the latest Autoconf in favour of the name
2841 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2842 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2843 analogy I have changed it to CFLAGS_FOR_BUILD.
2845 4. Added -no-undefined to the linking command in the Makefile, because this is
2846 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2847 linking step for the pcreposix library.
2849 5. PCRE was failing to diagnose the case of two named groups with the same
2850 name.
2852 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2853 literal character that is needed in the subject for a match, and scans along to
2854 ensure that it is present before embarking on the full matching process. This
2855 saves time in cases of nested unlimited repeats that are never going to match.
2856 Problem: the scan can take a lot of time if the subject is very long (e.g.
2857 megabytes), thus penalizing straightforward matches. It is now done only if the
2858 amount of subject to be scanned is less than 1000 bytes.
2860 7. A lesser problem with the same optimization is that it was recording the
2861 first character of an anchored pattern as "needed", thus provoking a search
2862 right along the subject, even when the first match of the pattern was going to
2863 fail. The "needed" character is now not set for anchored patterns, unless it
2864 follows something in the pattern that is of non-fixed length. Thus, it still
2865 fulfils its original purpose of finding quick non-matches in cases of nested
2866 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2869 Version 4.0 17-Feb-03
2870 ---------------------
2872 1. If a comment in an extended regex that started immediately after a meta-item
2873 extended to the end of string, PCRE compiled incorrect data. This could lead to
2874 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2876 2. Moved to autoconf 2.53 and libtool 1.4.2.
2878 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2879 the special perltest8 script is no longer needed - all the tests can be run
2880 from a single perltest script.
2882 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2883 by \s. It has now been removed in PCRE. This means it isn't recognized as
2884 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2885 class [:space:] *does* include VT, thereby creating a mess.
2887 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2888 space and tab.
2890 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2891 its new features into the main test script, reducing the number of scripts.
2893 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2894 were backward compatible, and made the (?i) apply to the whole pattern, as if
2895 /i were given. Now it behaves more logically, and applies the option setting
2896 only to what follows. PCRE has been changed to follow suit. However, if it
2897 finds options settings right at the start of the pattern, it extracts them into
2898 the global options, as before. Thus, they show up in the info data.
2900 8. Added support for the \Q...\E escape sequence. Characters in between are
2901 treated as literals. This is slightly different from Perl in that $ and @ are
2902 also handled as literals inside the quotes. In Perl, they will cause variable
2903 interpolation. Note the following examples:
2905 Pattern PCRE matches Perl matches
2907 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2908 \Qabc\$xyz\E abc\$xyz abc\$xyz
2909 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2911 For compatibility with Perl, \Q...\E sequences are recognized inside character
2912 classes as well as outside them.
2914 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2915 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2916 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2917 signed/unsigned warnings.
2919 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2920 option for pcretest, so I've replaced it by a simple function that does just
2921 that job.
2923 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2924 "pcregrep -".
2926 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2927 Java package. This provides some syntactic sugar for simple cases of what my
2928 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2929 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2930 item, you can use this simplified notation. Note that only makes sense with
2931 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2932 greediness, whatever the setting of the PCRE_UNGREEDY option.
2934 13. A change of greediness default within a pattern was not taking effect at
2935 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2936 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2937 was abstracted outside.
2939 14. PCRE now supports the \G assertion. It is true when the current matching
2940 position is at the start point of the match. This differs from \A when the
2941 starting offset is non-zero. Used with the /g option of pcretest (or similar
2942 code), it works in the same way as it does for Perl's /g option. If all
2943 alternatives of a regex begin with \G, the expression is anchored to the start
2944 match position, and the "anchored" flag is set in the compiled expression.
2946 15. Some bugs concerning the handling of certain option changes within patterns
2947 have been fixed. These applied to options other than (?ims). For example,
2948 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2949 been the other way round. Some of this was related to change 7 above.
2951 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2952 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2953 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2954 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2956 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2957 mode. Unlike ".", it always matches newline, whatever the setting of
2958 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2959 assertions. Perl allows it, but it doesn't (in general) work because it can't
2960 calculate the length of the lookbehind. At least, that's the case for Perl
2961 5.8.0 - I've been told they are going to document that it doesn't work in
2962 future.
2964 18. Added an error diagnosis for escapes that PCRE does not support: these are
2965 \L, \l, \N, \P, \p, \U, \u, and \X.
2967 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2968 reading past the end of the pattern in cases such as /[abcd/.
2970 20. PCRE was getting more memory than necessary for patterns with classes that
2971 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2973 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2974 compiling PCRE for use with Virtual Pascal.
2976 22. Small fix to the Makefile to make it work properly if the build is done
2977 outside the source tree.
2979 23. Added a new extension: a condition to go with recursion. If a conditional
2980 subpattern starts with (?(R) the "true" branch is used if recursion has
2981 happened, whereas the "false" branch is used only at the top level.
2983 24. When there was a very long string of literal characters (over 255 bytes
2984 without UTF support, over 250 bytes with UTF support), the computation of how
2985 much memory was required could be incorrect, leading to segfaults or other
2986 strange effects.
2988 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2989 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2990 there was a subsequent back reference to those brackets. This meant that, for
2991 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2992 possible to check for precisely this case. All we can do is abandon the
2993 optimization if .* occurs inside capturing brackets when there are any back
2994 references whatsoever. (See below for a better fix that came later.)
2996 26. The handling of the optimization for finding the first character of a
2997 non-anchored pattern, and for finding a character that is required later in the
2998 match were failing in some cases. This didn't break the matching; it just
2999 failed to optimize when it could. The way this is done has been re-implemented.
3001 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3003 28. Added a new feature that provides some of the functionality that Perl
3004 provides with (?{...}). The facility is termed a "callout". The way it is done
3005 in PCRE is for the caller to provide an optional function, by setting
3006 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3007 global variable. By default it is unset, which disables all calling out. To get
3008 the function called, the regex must include (?C) at appropriate points. This
3009 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3010 This provides a means of identifying different callout points. When PCRE
3011 reaches such a point in the regex, if pcre_callout has been set, the external
3012 function is called. It is provided with data in a structure called
3013 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3014 matching continues; if it returns a non-zero value, the match at the current
3015 point fails. However, backtracking will occur if possible. [This was changed
3016 later and other features added - see item 49 below.]
3018 29. pcretest is upgraded to test the callout functionality. It provides a
3019 callout function that displays information. By default, it shows the start of
3020 the match and the current position in the text. There are some new data escapes
3021 to vary what happens:
3023 \C+ in addition, show current contents of captured substrings
3024 \C- do not supply a callout function
3025 \C!n return 1 when callout number n is reached
3026 \C!n!m return 1 when callout number n is reached for the mth time
3028 30. If pcregrep was called with the -l option and just a single file name, it
3029 output "<stdin>" if a match was found, instead of the file name.
3031 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3032 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3033 pcre_exec(). This saves a malloc/free per call. The default value of
3034 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3035 when configuring.
3037 32. The default maximum size of a compiled pattern is 64K. There have been a
3038 few cases of people hitting this limit. The code now uses macros to handle the
3039 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3040 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3041 configuring. Tests 2 and 5 work only with 2-byte links because they output
3042 debugging information about compiled patterns.
3044 33. Internal code re-arrangements:
3046 (a) Moved the debugging function for printing out a compiled regex into
3047 its own source file (printint.c) and used #include to pull it into
3048 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3049 separate copies.
3051 (b) Defined the list of op-code names for debugging as a macro in
3052 internal.h so that it is next to the definition of the opcodes.
3054 (c) Defined a table of op-code lengths for simpler skipping along compiled
3055 code. This is again a macro in internal.h so that it is next to the
3056 definition of the opcodes.
3058 34. Added support for recursive calls to individual subpatterns, along the
3059 lines of Robin Houston's patch (but implemented somewhat differently).
3061 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3062 allow it to read and process whole directories in Win32. This code was
3063 contributed by Lionel Fourquaux; it has not been tested by me.
3065 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3066 used to name a group. Names consist of alphanumerics and underscores, and must
3067 be unique. Back references use the syntax (?P=name) and recursive calls use
3068 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3069 numbers. The function pcre_fullinfo() can be used after compilation to extract
3070 a name/number map. There are three relevant calls:
3072 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3073 PCRE_INFO_NAMECOUNT yields the number of entries
3074 PCRE_INFO_NAMETABLE yields a pointer to the map.
3076 The map is a vector of fixed-size entries. The size of each entry depends on
3077 the length of the longest name used. The first two bytes of each entry are the
3078 group number, most significant byte first. There follows the corresponding
3079 name, zero terminated. The names are in alphabetical order.
3081 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3082 case instead of 255. Making it the same both with and without UTF-8 support
3083 means that the same test output works with both.
3085 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3086 calling malloc() with a zero argument.
3088 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3089 optimization. I've improved things by keeping a bitmap of backreferences with
3090 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3091 fact referenced, the optimization can be applied. It is unlikely that a
3092 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3093 the match to follow \n) will appear inside brackets with a number greater than
3094 31, but if it does, any back reference > 31 suppresses the optimization.
3096 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3097 of disabling numbered capturing parentheses. Any opening parenthesis that is
3098 not followed by ? behaves as if it were followed by ?: but named parentheses
3099 can still be used for capturing (and they will acquire numbers in the usual
3100 way).
3102 41. Redesigned the return codes from the match() function into yes/no/error so
3103 that errors can be passed back from deep inside the nested calls. A malloc
3104 failure while inside a recursive subpattern call now causes the
3105 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3107 42. It is now possible to set a limit on the number of times the match()
3108 function is called in a call to pcre_exec(). This facility makes it possible to
3109 limit the amount of recursion and backtracking, though not in a directly
3110 obvious way, because the match() function is used in a number of different
3111 circumstances. The count starts from zero for each position in the subject
3112 string (for non-anchored patterns). The default limit is, for compatibility, a
3113 large number, namely 10 000 000. You can change this in two ways:
3115 (a) When configuring PCRE before making, you can use --with-match-limit=n
3116 to set a default value for the compiled library.
3118 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3119 a different value is set. See 45 below.
3121 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3123 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3124 of things that can be changed at compile time. The first argument specifies
3125 what is wanted and the second points to where the information is to be placed.
3126 The current list of available information is:
3130 The output is an integer that is set to one if UTF-8 support is available;
3131 otherwise it is set to zero.
3135 The output is an integer that it set to the value of the code that is used for
3136 newline. It is either LF (10) or CR (13).
3140 The output is an integer that contains the number of bytes used for internal
3141 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3145 The output is an integer that contains the threshold above which the POSIX
3146 interface uses malloc() for output vectors. See item 31 above.
3150 The output is an unsigned integer that contains the default limit of the number
3151 of match() calls in a pcre_exec() execution. See 42 above.
3153 44. pcretest has been upgraded by the addition of the -C option. This causes it
3154 to extract all the available output from the new pcre_config() function, and to
3155 output it. The program then exits immediately.
3157 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3158 order to support additional features. One way would have been to define
3159 pcre_exec2() (for example) with extra arguments, but this would not have been
3160 extensible, and would also have required all calls to the original function to
3161 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3162 is used for passing in "extra" data from pcre_study().
3164 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3165 contains the following fields:
3167 flags a bitmap indicating which of the following fields are set
3168 study_data opaque data from pcre_study()
3169 match_limit a way of specifying a limit on match() calls for a specific
3170 call to pcre_exec()
3171 callout_data data for callouts (see 49 below)
3173 The flag bits are also defined in pcre.h, and are
3179 The pcre_study() function now returns one of these new pcre_extra blocks, with
3180 the actual study data pointed to by the study_data field, and the
3181 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3182 before. That is, this change is entirely upwards-compatible and requires no
3183 change to existing code.
3185 If you want to pass in additional data to pcre_exec(), you can either place it
3186 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3187 block.
3189 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3190 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3191 times with different match limits, until it finds the minimum value needed for
3192 pcre_exec() to complete. The value is then output. This can be instructive; for
3193 most simple matches the number is quite small, but for pathological cases it
3194 gets very large very quickly.
3196 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3197 returns the size of the data block pointed to by the study_data field in a
3198 pcre_extra block, that is, the value that was passed as the argument to
3199 pcre_malloc() when PCRE was getting memory in which to place the information
3200 created by pcre_study(). The fourth argument should point to a size_t variable.
3201 pcretest has been extended so that this information is shown after a successful
3202 pcre_study() call when information about the compiled regex is being displayed.
3204 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3205 because what follows is always an absolute path. (Later: it turns out that this
3206 is more than cosmetic for MinGW, because it doesn't like empty path
3207 components.)
3209 49. Some changes have been made to the callout feature (see 28 above):
3211 (i) A callout function now has three choices for what it returns:
3213 0 => success, carry on matching
3214 > 0 => failure at this point, but backtrack if possible
3215 < 0 => serious error, return this value from pcre_exec()
3217 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3218 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3219 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3220 use by callout functions. It will never be used by PCRE itself.
3222 (ii) The pcre_extra structure (see 45 above) has a void * field called
3223 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3224 pcre_callout_block structure has a field of the same name. The contents of
3225 the field passed in the pcre_extra structure are passed to the callout
3226 function in the corresponding field in the callout block. This makes it
3227 easier to use the same callout-containing regex from multiple threads. For
3228 testing, the pcretest program has a new data escape
3230 \C*n pass the number n (may be negative) as callout_data
3232 If the callout function in pcretest receives a non-zero value as
3233 callout_data, it returns that value.
3235 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3236 there were some redundant $(CFLAGS) in commands that are now specified as
3237 $(LINK), which already includes $(CFLAGS).
3239 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3240 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3241 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3242 one-byte characters throughout. Note that case-insensitive matching applies
3243 only to characters whose values are less than 256. PCRE doesn't support the
3244 notion of cases for higher-valued characters.
3246 (i) A character class whose characters are all within 0-255 is handled as
3247 a bit map, and the map is inverted for negative classes. Previously, a
3248 character > 255 always failed to match such a class; however it should
3249 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3251 (ii) A negated character class with a single character < 255 is coded as
3252 "not this character" (OP_NOT). This wasn't working properly when the test
3253 character was multibyte, either singly or repeated.
3255 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3256 mode, for example: \x{100}{2,3}.
3258 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3259 singly or repeated) now correctly test multibyte characters. However,
3260 PCRE doesn't recognize any characters with values greater than 255 as
3261 digits, spaces, or word characters. Such characters always match \D, \S,
3262 and \W, and never match \d, \s, or \w.
3264 (v) Classes may now contain characters and character ranges with values
3265 greater than 255. For example: [ab\x{100}-\x{400}].
3267 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3268 PCRE in UTF-8 mode.
3270 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3271 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3272 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3273 value.)
3275 53. The single man page has become too large. I have therefore split it up into
3276 a number of separate man pages. These also give rise to individual HTML pages;
3277 these are now put in a separate directory, and there is an index.html page that
3278 lists them all. Some hyperlinking between the pages has been installed.
3280 54. Added convenience functions for handling named capturing parentheses.
3282 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3283 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3284 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3285 are faulted.
3287 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3288 calling configure. These values are used when compiling the dftables.c program
3289 which is run to generate the source of the default character tables. They
3290 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3291 you will need to set these values.
3293 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3296 Version 3.9 02-Jan-02
3297 ---------------------
3299 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3301 2. If --disable-static was given, the building process failed when trying to
3302 build pcretest and pcregrep. (For some reason it was using libtool to compile
3303 them, which is not right, as they aren't part of the library.)
3306 Version 3.8 18-Dec-01
3307 ---------------------
3309 1. The experimental UTF-8 code was completely screwed up. It was packing the
3310 bytes in the wrong order. How dumb can you get?
3313 Version 3.7 29-Oct-01
3314 ---------------------
3316 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3317 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3318 this didn't happen under Solaris 8, where I normally test things.
3320 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3321 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3322 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3323 this fix an hour or so after the initial 3.7 release.)
3326 Version 3.6 23-Oct-01
3327 ---------------------
3329 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3330 offsets passed as NULL with zero offset count.
3332 2. The config.guess and config.sub files had not been updated when I moved to
3333 the latest autoconf.
3336 Version 3.5 15-Aug-01
3337 ---------------------
3339 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3340 had been forgotten.
3342 2. By using declared but undefined structures, we can avoid using "void"
3343 definitions in pcre.h while keeping the internal definitions of the structures
3344 private.
3346 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3347 user point of view, this means that both static and shared libraries are built
3348 by default, but this can be individually controlled. More of the work of
3349 handling this static/shared cases is now inside libtool instead of PCRE's make
3350 file.
3352 4. The pcretest utility is now installed along with pcregrep because it is
3353 useful for users (to test regexs) and by doing this, it automatically gets
3354 relinked by libtool. The documentation has been turned into a man page, so
3355 there are now .1, .txt, and .html versions in /doc.
3357 5. Upgrades to pcregrep:
3358 (i) Added long-form option names like gnu grep.
3359 (ii) Added --help to list all options with an explanatory phrase.
3360 (iii) Added -r, --recursive to recurse into sub-directories.
3361 (iv) Added -f, --file to read patterns from a file.
3363 6. pcre_exec() was referring to its "code" argument before testing that
3364 argument for NULL (and giving an error if it was NULL).
3366 7. Upgraded Makefile.in to allow for compiling in a different directory from
3367 the source directory.
3369 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3370 options bits, the pointer it was passed was to an int instead of to an unsigned
3371 long int. This mattered only on 64-bit systems.
3373 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3374 generated) instead of pcre.in, which it its source. Also made the same change
3375 in several of the .c files.
3377 10. A new release of gcc defines printf() as a macro, which broke pcretest
3378 because it had an ifdef in the middle of a string argument for printf(). Fixed
3379 by using separate calls to printf().
3381 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3382 script, to force use of CR or LF instead of \n in the source. On non-Unix
3383 systems, the value can be set in config.h.
3385 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3386 absolute limit. Changed the text of the error message to make this clear, and
3387 likewise updated the man page.
3389 13. The limit of 99 on the number of capturing subpatterns has been removed.
3390 The new limit is 65535, which I hope will not be a "real" limit.
3393 Version 3.4 22-Aug-00
3394 ---------------------
3396 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3398 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3401 Version 3.3 01-Aug-00
3402 ---------------------
3404 1. If an octal character was given, but the value was greater than \377, it
3405 was not getting masked to the least significant bits, as documented. This could
3406 lead to crashes in some systems.
3408 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3409 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3411 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3412 These just pass their arguments on to (pcre_free)(), but they are provided
3413 because some uses of PCRE bind it to non-C systems that can call its functions,
3414 but cannot call free() or pcre_free() directly.
3416 4. Add "make test" as a synonym for "make check". Corrected some comments in
3417 the Makefile.
3419 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3420 Makefile.
3422 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3423 command called pgrep for grepping around the active processes.
3425 7. Added the beginnings of support for UTF-8 character strings.
3427 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3428 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3429 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3430 out for the ar command.)
3433 Version 3.2 12-May-00
3434 ---------------------
3436 This is purely a bug fixing release.
3438 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3439 of ZA. This was just one example of several cases that could provoke this bug,
3440 which was introduced by change 9 of version 2.00. The code for breaking
3441 infinite loops after an iteration that matches an empty string was't working
3442 correctly.
3444 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3445 when matched against abbab (for example). After matching an empty string, it
3446 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3447 caused it to match further down the string than it should.
3449 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3450 was there because it doesn't seem to be needed, and it causes trouble on some
3451 systems, as it is not a Standard C header. It has been removed.
3453 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3454 were reported on the Macintosh. The changes were from
3456 while ((c = *(++ptr)) != 0 && c != '\n');
3457 to
3458 while ((c = *(++ptr)) != 0 && c != '\n') ;
3460 Totally extraordinary, but if that's what it takes...
3462 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3463 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3464 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3465 assumes the way PCRE uses memmove() (always moving upwards).
3467 6. PCRE is being used in one environment where strchr() is not available. There
3468 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3469 faster code anyway.
3472 Version 3.1 09-Feb-00
3473 ---------------------
3475 The only change in this release is the fixing of some bugs in Makefile.in for
3476 the "install" target:
3478 (1) It was failing to install pcreposix.h.
3480 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3483 Version 3.0 01-Feb-00
3484 ---------------------
3486 1. Add support for the /+ modifier to perltest (to output $` like it does in
3487 pcretest).
3489 2. Add support for the /g modifier to perltest.
3491 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3492 matches null strings.
3494 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3495 pattern. Perl treats empty patterns specially - it reuses the most recent
3496 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3497 effect.
3499 5. The POSIX interface was broken in that it was just handing over the POSIX
3500 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3501 required a bigger vector, with some working space on the end. This means that
3502 the POSIX wrapper now has to get and free some memory, and copy the results.
3504 6. Added some simple autoconf support, placing the test data and the
3505 documentation in separate directories, re-organizing some of the
3506 information files, and making it build pcre-config (a GNU standard). Also added
3507 libtool support for building PCRE as a shared library, which is now the
3508 default.
3510 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3511 09 are not valid octal constants. Single digits will be used for minor values
3512 less than 10.
3514 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3515 existing programs that set these in the POSIX interface can use PCRE without
3516 modification.
3518 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3519 return all that pcre_info() returns, plus additional data. The pcre_info()
3520 function is retained for compatibility, but is considered to be obsolete.
3522 10. Added experimental recursion feature (?R) to handle one common case that
3523 Perl 5.6 will be able to do with (?p{...}).
3525 11. Added support for POSIX character classes like [:alpha:], which Perl is
3526 adopting.
3529 Version 2.08 31-Aug-99
3530 ----------------------
3532 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3533 trying to match at the startoffset position, but instead was moving forward to
3534 the next newline as if a previous match had failed.
3536 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3537 and could get into a loop if a null string was matched other than at the start
3538 of the subject.
3540 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3541 be distinguished at compile time, and for completeness also added PCRE_DATE.
3543 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3544 in GnuWin32 environments.
3547 Version 2.07 29-Jul-99
3548 ----------------------
3550 1. The documentation is now supplied in plain text form and HTML as well as in
3551 the form of man page sources.
3553 2. C++ compilers don't like assigning (void *) values to other pointer types.
3554 In particular this affects malloc(). Although there is no problem in Standard
3555 C, I've put in casts to keep C++ compilers happy.
3557 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3558 should be (const char *).
3560 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3561 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3562 However, I haven't made this a standard facility. The documentation doesn't
3563 mention it, and the Makefile doesn't support it.
3565 5. The Makefile now contains an "install" target, with editable destinations at
3566 the top of the file. The pcretest program is not installed.
3568 6. pgrep -V now gives the PCRE version number and date.
3570 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3571 causing the entire string to be ignored, instead of just the last character.
3573 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3574 non-matching string, it can take a very, very long time, even for strings of
3575 quite modest length, because of the nested recursion. PCRE now does better in
3576 some of these cases. It does this by remembering the last required literal
3577 character in the pattern, and pre-searching the subject to ensure it is present
3578 before running the real match. In other words, it applies a heuristic to detect
3579 some types of certain failure quickly, and in the above example, if presented
3580 with a string that has no trailing " it gives "no match" very quickly.
3582 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3583 other alternatives are tried instead.
3586 Version 2.06 09-Jun-99
3587 ----------------------
3589 1. Change pcretest's output for amount of store used to show just the code
3590 space, because the remainder (the data block) varies in size between 32-bit and
3591 64-bit systems.
3593 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3594 start matching at. This allows lookbehinds to work when searching for multiple
3595 occurrences in a string.
3597 3. Added additional options to pcretest for testing multiple occurrences:
3599 /+ outputs the rest of the string that follows a match
3600 /g loops for multiple occurrences, using the new startoffset argument
3601 /G loops for multiple occurrences by passing an incremented pointer
3603 4. PCRE wasn't doing the "first character" optimization for patterns starting
3604 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3605 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3606 the letter 'x'. On long subject strings, this gives a significant speed-up.
3609 Version 2.05 21-Apr-99
3610 ----------------------
3612 1. Changed the type of magic_number from int to long int so that it works
3613 properly on 16-bit systems.
3615 2. Fixed a bug which caused patterns starting with .* not to work correctly
3616 when the subject string contained newline characters. PCRE was assuming
3617 anchoring for such patterns in all cases, which is not correct because .* will
3618 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3619 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3620 must be retried after every newline in the subject.
3623 Version 2.04 18-Feb-99
3624 ----------------------
3626 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3627 computation of the store needed to hold the pattern was incorrect (too large).
3628 If such patterns were nested a few deep, this could multiply and become a real
3629 problem.
3631 2. Added /M option to pcretest to show the memory requirement of a specific
3632 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3634 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3635 compiled in such a way that the backtracking after subsequent failure was
3636 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3637 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3640 Version 2.03 02-Feb-99
3641 ----------------------
3643 1. Fixed typo and small mistake in man page.
3645 2. Added 4th condition (GPL supersedes if conflict) and created separate
3646 LICENCE file containing the conditions.
3648 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3649 Perl, that is the internal \ allows the delimiter to be included in the
3650 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3651 the final delimiter, add \ to the end of the pattern (to test the error).
3653 4. Added the convenience functions for extracting substrings after a successful
3654 match. Updated pcretest to make it able to test these functions.
3657 Version 2.02 14-Jan-99
3658 ----------------------
3660 1. Initialized the working variables associated with each extraction so that
3661 their saving and restoring doesn't refer to uninitialized store.
3663 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3664 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3665 fix the problem.
3667 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3668 calls, and wasn't printing the correct value for compiling calls. Increased the
3669 default value of LOOPREPEAT, and the number of significant figures in the
3670 times.
3672 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3674 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3675 a building problem on Windows NT with a FAT file system.
3678 Version 2.01 21-Oct-98
3679 ----------------------
3681 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3682 to character tables built by pcre_maketables() in the current locale. If NULL
3683 is passed, the default tables are used.
3686 Version 2.00 24-Sep-98
3687 ----------------------
3689 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3690 it any more.
3692 2. Allow quantification of (?>) groups, and make it work correctly.
3694 3. The first character computation wasn't working for (?>) groups.
3696 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3697 end of the subject) and add 5.005's \z, which really does match only at the
3698 very end of the subject.
3700 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3702 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3703 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3704 localized options. All options to pcre_study() were also removed.
3706 7. Add other new features from 5.005:
3708 $(?<= positive lookbehind
3709 $(?<! negative lookbehind
3710 (?imsx-imsx) added the unsetting capability
3711 such a setting is global if at outer level; local otherwise
3712 (?imsx-imsx:) non-capturing groups with option setting
3713 (?(cond)re|re) conditional pattern matching
3715 A backreference to itself in a repeated group matches the previous
3716 captured string.
3718 8. General tidying up of studying (both automatic and via "study")
3719 consequential on the addition of new assertions.
3721 9. As in 5.005, unlimited repeated groups that could match an empty substring
3722 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3723 runtime if any iteration does actually match an empty substring.
3725 10. Include the RunTest script in the distribution.
3727 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3728 discrepancies, some of which were old and were also with respect to 5.004. They
3729 have now been fixed.
3732 Version 1.09 28-Apr-98
3733 ----------------------
3735 1. A negated single character class followed by a quantifier with a minimum
3736 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3737 program crashes, or just wrong answers. This did not apply to negated classes
3738 containing more than one character, or to minima other than one.
3741 Version 1.08 27-Mar-98
3742 ----------------------
3744 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3746 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3747 latter must appear before anything that relies on it in the pattern.
3750 Version 1.07 16-Feb-98
3751 ----------------------
3753 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3754 repeat of a potentially empty string).
3757 Version 1.06 23-Jan-98
3758 ----------------------
3760 1. Added Markus Oberhumer's little patches for C++.
3762 2. Literal strings longer than 255 characters were broken.
3765 Version 1.05 23-Dec-97
3766 ----------------------
3768 1. Negated character classes containing more than one character were failing if
3769 PCRE_CASELESS was set at run time.
3772 Version 1.04 19-Dec-97
3773 ----------------------
3775 1. Corrected the man page, where some "const" qualifiers had been omitted.
3777 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3778 input syntax.
3780 3. Fixed memory leak which occurred when a regex with back references was
3781 matched with an offsets vector that wasn't big enough. The temporary memory
3782 that is used in this case wasn't being freed if the match failed.
3784 4. Tidied pcretest to ensure it frees memory that it gets.
3786 5. Temporary memory was being obtained in the case where the passed offsets
3787 vector was exactly big enough.
3789 6. Corrected definition of offsetof() from change 5 below.
3791 7. I had screwed up change 6 below and broken the rules for the use of
3792 setjmp(). Now fixed.
3795 Version 1.03 18-Dec-97
3796 ----------------------
3798 1. A erroneous regex with a missing opening parenthesis was correctly
3799 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3800 on some systems.
3802 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3803 it was reported that one broken compiler failed on the former because "code" is
3804 also an independent variable.
3806 3. The erroneous regex a[]b caused an array overrun reference.
3808 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3809 fail on data ending with that character. (It was going on too far, and checking
3810 the next character, typically a binary zero.) This was specific to the
3811 optimized code for single-character negative classes.
3813 5. Added a contributed patch from the TIN world which does the following:
3815 + Add an undef for memmove, in case the the system defines a macro for it.
3817 + Add a definition of offsetof(), in case there isn't one. (I don't know
3818 the reason behind this - offsetof() is part of the ANSI standard - but
3819 it does no harm).
3821 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3822 most of the places where whitespace preceded '#'. I have given up and
3823 allowed the remaining 2 cases to be at the margin.
3825 + Rename some variables in pcre to eliminate shadowing. This seems very
3826 pedantic, but does no harm, of course.
3828 6. Moved the call to setjmp() into its own function, to get rid of warnings
3829 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3831 7. Constructs such as \d{8,} were compiling into the equivalent of
3832 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3833 outcome, but in this particular case used more store than had been allocated,
3834 which caused the bug to be discovered because it threw up an internal error.
3836 8. The debugging code in both pcre and pcretest for outputting the compiled
3837 form of a regex was going wrong in the case of back references followed by
3838 curly-bracketed repeats.
3841 Version 1.02 12-Dec-97
3842 ----------------------
3844 1. Typos in pcre.3 and comments in the source fixed.
3846 2. Applied a contributed patch to get rid of places where it used to remove
3847 'const' from variables, and fixed some signed/unsigned and uninitialized
3848 variable warnings.
3850 3. Added the "runtest" target to Makefile.
3852 4. Set default compiler flag to -O2 rather than just -O.
3855 Version 1.01 19-Nov-97
3856 ----------------------
3858 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3859 like /([ab]*)*/, that is, for classes with more than one character in them.
3861 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3862 as /((?>a*))*/ (a PCRE_EXTRA facility).
3865 Version 1.00 18-Nov-97
3866 ----------------------
3868 1. Added compile-time macros to support systems such as SunOS4 which don't have
3869 memmove() or strerror() but have other things that can be used instead.
3871 2. Arranged that "make clean" removes the executables.
3874 Version 0.99 27-Oct-97
3875 ----------------------
3877 1. Fixed bug in code for optimizing classes with only one character. It was
3878 initializing a 32-byte map regardless, which could cause it to run off the end
3879 of the memory it had got.
3881 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3884 Version 0.98 22-Oct-97
3885 ----------------------
3887 1. Fixed bug in code for handling temporary memory usage when there are more
3888 back references than supplied space in the ovector. This could cause segfaults.
3891 Version 0.97 21-Oct-97
3892 ----------------------
3894 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3896 2. Optimized negated single characters not to use a bit map.
3898 3. Brought error texts together as macro definitions; clarified some of them;
3899 fixed one that was wrong - it said "range out of order" when it meant "invalid
3900 escape sequence".
3902 4. Changed some char * arguments to const char *.
3904 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3906 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3907 pcretest.
3910 Version 0.96 16-Oct-97
3911 ----------------------
3913 1. Added a simple "pgrep" utility to the distribution.
3915 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3916 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3917 where "ddd" means "one or more decimal digits".
3919 3. Fixed serious bug. If a pattern had a back reference, but the call to
3920 pcre_exec() didn't supply a large enough ovector to record the related
3921 identifying subpattern, the match always failed. PCRE now remembers the number
3922 of the largest back reference, and gets some temporary memory in which to save
3923 the offsets during matching if necessary, in order to ensure that
3924 backreferences always work.
3926 4. Increased the compatibility with Perl in a number of ways:
3928 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3929 to request this handling. The option can be set at compile or exec time.
3931 (b) $ matches before a terminating newline by default; an option
3932 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3933 mode). The option can be set at compile or exec time.
3935 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3936 the same as Perl's. If the decimal number it represents is less than 10
3937 or there aren't that many previous left capturing parentheses, an octal
3938 escape is read. Inside a character class, it's always an octal escape,
3939 even if it is a single digit.
3941 (d) An escaped but undefined alphabetic character is taken as a literal,
3942 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3943 escapes.
3945 (e) {0} is now permitted. (The previous item is removed from the compiled
3946 pattern).
3948 5. Changed all the names of code files so that the basic parts are no longer
3949 than 10 characters, and abolished the teeny "globals.c" file.
3951 6. Changed the handling of character classes; they are now done with a 32-byte
3952 bit map always.
3954 7. Added the -d and /D options to pcretest to make it possible to look at the
3955 internals of compilation without having to recompile pcre.
3958 Version 0.95 23-Sep-97
3959 ----------------------
3961 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3962 \x20 at the start of a run of normal characters. These were being treated as
3963 real characters, instead of the source characters being re-checked.
3966 Version 0.94 18-Sep-97
3967 ----------------------
3969 1. The functions are now thread-safe, with the caveat that the global variables
3970 containing pointers to malloc() and free() or alternative functions are the
3971 same for all threads.
3973 2. Get pcre_study() to generate a bitmap of initial characters for non-
3974 anchored patterns when this is possible, and use it if passed to pcre_exec().
3977 Version 0.93 15-Sep-97
3978 ----------------------
3980 1. /(b)|(:+)/ was computing an incorrect first character.
3982 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3983 but not actually doing anything yet.
3985 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3986 as Perl does (e.g. [-az] or [az-]).
3988 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3989 all possible positions.
3991 5. Split up into different modules to avoid including unneeded functions in a
3992 compiled binary. However, compile and exec are still in one module. The "study"
3993 function is split off.
3995 6. The character tables are now in a separate module whose source is generated
3996 by an auxiliary program - but can then be edited by hand if required. There are
3997 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3998 toupper() in the code.
4000 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4001 make them global. Abolish the function for setting them, as the caller can now
4002 set them directly.
4005 Version 0.92 11-Sep-97
4006 ----------------------
4008 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4009 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4011 2. Caseless matching was not working in character classes if the characters in
4012 the pattern were in upper case.
4014 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4016 4. Make PCRE_ANCHORED public and accept as a compile option.
4018 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4019 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4020 pass them.
4022 6. Give an error if bad option bits passed at compile or run time.
4024 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4025 pcretest to cause it to pass that flag.
4027 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4028 options, and the first character, if set.
4030 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4033 Version 0.91 10-Sep-97
4034 ----------------------
4036 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4037 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4039 2. PCRE was looping on encountering an indefinitely repeated back reference to
4040 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4041 Perl does - treats the match as successful.
4043 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5