ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 638 - (show annotations)
Mon Jul 25 09:41:19 2011 UTC (9 years, 4 months ago) by ph10
File size: 208139 byte(s)
Minor code tidy.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.13 30-Apr-2011
5 ------------------------
7 1. The Unicode data tables have been updated to Unicode 6.0.0.
9 2. Two minor typos in pcre_internal.h have been fixed.
11 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
15 4. There were a number of related bugs in the code for matching backrefences
16 caselessly in UTF-8 mode when codes for the characters concerned were
17 different numbers of bytes. For example, U+023A and U+2C65 are an upper
18 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21 2-byte code at the end of the subject (it thought there wasn't enough data
22 left).
24 5. Comprehensive information about what went wrong is now returned by
25 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26 as the output vector has at least 2 elements. The offset of the start of
27 the failing character and a reason code are placed in the vector.
29 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
30 now returned is for the first byte of the failing character, instead of the
31 last byte inspected. This is an incompatible change, but I hope it is small
32 enough not to be a problem. It makes the returned offset consistent with
33 pcre_exec() and pcre_dfa_exec().
35 7. pcretest now gives a text phrase as well as the error number when
36 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37 failure, the offset and reason code are output.
39 8. When \R was used with a maximizing quantifier it failed to skip backwards
40 over a \r\n pair if the subsequent match failed. Instead, it just skipped
41 back over a single character (\n). This seems wrong (because it treated the
42 two characters as a single entity when going forwards), conflicts with the
43 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
44 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
45 has been changed.
47 9. Some internal refactoring has changed the processing so that the handling
48 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
49 time (the PCRE_DOTALL option was changed this way some time ago: version
50 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
51 which was always a bit of a fudge. It also means that there is one less
52 argument for the match() function, which reduces its stack requirements
53 slightly. This change also fixes an incompatibility with Perl: the pattern
54 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
56 10. More internal refactoring has drastically reduced the number of recursive
57 calls to match() for possessively repeated groups such as (abc)++ when
58 using pcre_exec().
60 11. While implementing 10, a number of bugs in the handling of groups were
61 discovered and fixed:
63 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
64 (a|)*(?1) gave a compile-time internal error.
65 ((a|)+)+ did not notice that the outer group could match an empty string.
66 (^a|^)+ was not marked as anchored.
67 (.*a|.*)+ was not marked as matching at start or after a newline.
69 12. Yet more internal refactoring has removed another argument from the match()
70 function. Special calls to this function are now indicated by setting a
71 value in a variable in the "match data" data block.
73 13. Be more explicit in pcre_study() instead of relying on "default" for
74 opcodes that mean there is no starting character; this means that when new
75 ones are added and accidentally left out of pcre_study(), testing should
76 pick them up.
78 14. The -s option of pcretest has been documented for ages as being an old
79 synonym of -m (show memory usage). I have changed it to mean "force study
80 for every regex", that is, assume /S for every regex. This is similar to -i
81 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
82 using it. It makes it easier to run collections of tests with and without
83 study enabled, and thereby test pcre_study() more easily. All the standard
84 tests are now run with and without -s (but some patterns can be marked as
85 "never study" - see 20 below).
87 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
88 restoration of the capturing data to the outer values was not happening
89 correctly.
91 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
92 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
93 pattern had matched an empty string, and so incorrectly returned a no
94 match.
96 17. There was optimizing code for the last branch of non-capturing parentheses,
97 and also for the obeyed branch of a conditional subexpression, which used
98 tail recursion to cut down on stack usage. Unfortunately, not that there is
99 the possibility of (*THEN) occurring in these branches, tail recursion is
100 no longer possible because the return has to be checked for (*THEN). These
101 two optimizations have therefore been removed.
103 18. If a pattern containing \R was studied, it was assumed that \R always
104 matched two bytes, thus causing the minimum subject length to be
105 incorrectly computed because \R can also match just one byte.
107 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
108 was incorrectly computed.
110 20. If /S is present twice on a test pattern in pcretest input, it *disables*
111 studying, thereby overriding the use of -s on the command line. This is
112 necessary for one or two tests to keep the output identical in both cases.
114 21. When (*ACCEPT) was used in an assertion that matched an empty string and
115 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
117 22. When an atomic group that contained a capturing parenthesis was
118 successfully matched, but the branch in which it appeared failed, the
119 capturing was not being forgotten if a higher numbered group was later
120 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
121 group 1 to "a", when in fact it should be unset. This applied to multi-
122 branched capturing and non-capturing groups, repeated or not, and also to
123 positive assertions (capturing in negative assertions is not well defined
124 in PCRE) and also to nested atomic groups.
126 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
127 subject after a captured substring (to make it easier to tell which of a
128 number of identical substrings has been captured).
130 24. The way atomic groups are processed by pcre_exec() has been changed so that
131 if they are repeated, backtracking one repetition now resets captured
132 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
133 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
134 "aaa". Previously, it would have been "a". As part of this code
135 refactoring, the way recursive calls are handled has also been changed.
137 24. If an assertion condition captured any substrings, they were not passed
138 back unless some other capturing happened later. For example, if
139 (?(?=(a))a) was matched against "a", no capturing was returned.
141 25. When studying a pattern that contained subroutine calls or assertions,
142 the code for finding the minimum length of a possible match was handling
143 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
144 group 1 called group 2 while simultaneously a separate group 2 called group
145 1). A stack overflow occurred in this case. I have fixed this by limiting
146 the recursion depth to 10.
148 26. Updated RunTest.bat in the distribution to the version supplied by Tom
149 Fortmann. This supports explicit test numbers on the command line, and has
150 argument validation and error reporting.
152 27. An instance of \X with an unlimited repeat could fail if at any point the
153 first character it looked at was a mark character.
155 28. Some minor code refactoring concerning Unicode properties and scripts
156 should reduce the stack requirement of match() slightly.
158 29. Added the '=' option to pcretest to check the setting of unused capturing
159 slots at the end of the pattern, which are documented as being -1, but are
160 not included in the return count.
162 30. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
163 compiled something random. Now it gives a compile-time error (as does
164 Perl).
166 31. A *MARK encountered during the processing of a positive assertion is now
167 recorded and passed back (compatible with Perl).
169 32. If --only-matching or --colour was set on a pcregrep call whose pattern
170 had alternative anchored branches, the search for a second match in a line
171 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
172 matched the line "0102" twice. The same bug affected patterns that started
173 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
174 twice.
176 33. Previously, PCRE did not allow quantification of assertions. However, Perl
177 does, and because of capturing effects, quantifying parenthesized
178 assertions may at times be useful. Quantifiers are now allowed for
179 parenthesized assertions.
181 34. A minor code tidy in pcre_compile() when checking options for \R usage.
184 Version 8.12 15-Jan-2011
185 ------------------------
187 1. Fixed some typos in the markup of the man pages, and wrote a script that
188 checks for such things as part of the documentation building process.
190 2. On a big-endian 64-bit system, pcregrep did not correctly process the
191 --match-limit and --recursion-limit options (added for 8.11). In
192 particular, this made one of the standard tests fail. (The integer value
193 went into the wrong half of a long int.)
195 3. If the --colour option was given to pcregrep with -v (invert match), it
196 did strange things, either producing crazy output, or crashing. It should,
197 of course, ignore a request for colour when reporting lines that do not
198 match.
200 4. Another pcregrep bug caused similar problems if --colour was specified with
201 -M (multiline) and the pattern match finished with a line ending.
203 5. In pcregrep, when a pattern that ended with a literal newline sequence was
204 matched in multiline mode, the following line was shown as part of the
205 match. This seems wrong, so I have changed it.
207 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
208 the check for further matches in the same line (so they could be coloured)
209 to overrun the end of the current line. If another match was found, it was
210 incorrectly shown (and then shown again when found in the next line).
212 7. If pcregrep was compiled under Windows, there was a reference to the
213 function pcregrep_exit() before it was defined. I am assuming this was
214 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
215 reported by a user. I've moved the definition above the reference.
218 Version 8.11 10-Dec-2010
219 ------------------------
221 1. (*THEN) was not working properly if there were untried alternatives prior
222 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
223 backtracked to try for "b" instead of moving to the next alternative branch
224 at the same level (in this case, to look for "c"). The Perl documentation
225 is clear that when (*THEN) is backtracked onto, it goes to the "next
226 alternative in the innermost enclosing group".
228 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
229 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
230 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
231 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
232 (*THEN).
234 3. If \s appeared in a character class, it removed the VT character from
235 the class, even if it had been included by some previous item, for example
236 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
237 of \s, but is part of the POSIX "space" class.)
239 4. A partial match never returns an empty string (because you can always
240 match an empty string at the end of the subject); however the checking for
241 an empty string was starting at the "start of match" point. This has been
242 changed to the "earliest inspected character" point, because the returned
243 data for a partial match starts at this character. This means that, for
244 example, /(?<=abc)def/ gives a partial match for the subject "abc"
245 (previously it gave "no match").
247 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
248 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
249 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
250 has an implication that the given string is incomplete (because a partial
251 match is preferred over a full match). For this reason, these items now
252 give a partial match in this situation. [Aside: previously, the one case
253 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
254 match rather than a full match, which was wrong by the old rules, but is
255 now correct.]
257 6. There was a bug in the handling of #-introduced comments, recognized when
258 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
259 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
260 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
261 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
262 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
263 places in pcre_compile().
265 7. Related to (6) above, when pcre_compile() was skipping #-introduced
266 comments when looking ahead for named forward references to subpatterns,
267 the only newline sequence it recognized was NL. It now handles newlines
268 according to the set newline convention.
270 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
271 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
272 cater for a lack of strerror(). These oversights have been fixed.
274 9. Added --match-limit and --recursion-limit to pcregrep.
276 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
278 11. When the -o option was used, pcregrep was setting a return code of 1, even
279 when matches were found, and --line-buffered was not being honoured.
281 12. Added an optional parentheses number to the -o and --only-matching options
282 of pcregrep.
284 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
285 can match an empty string. The code to do it in pcretest and pcredemo
286 needed fixing:
288 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
289 only one byte after an empty string match just before CRLF (this case
290 just got forgotten; "any" and "anycrlf" were OK).
292 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
293 mode when an empty string match preceded an ASCII character followed by
294 a non-ASCII character. (The code for advancing by one character rather
295 than one byte was nonsense.)
297 (c) The pcredemo.c sample program did not have any code at all to handle
298 the cases when CRLF is a valid newline sequence.
300 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
301 as a starting offset was within the subject string. There is now a new
302 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
303 negative or greater than the length of the string. In order to test this,
304 pcretest is extended to allow the setting of negative starting offsets.
306 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
307 starting offset points to the beginning of a UTF-8 character was
308 unnecessarily clumsy. I tidied it up.
310 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
311 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
313 17. Nobody had reported that the --include_dir option, which was added in
314 release 7.7 should have been called --include-dir (hyphen, not underscore)
315 for compatibility with GNU grep. I have changed it to --include-dir, but
316 left --include_dir as an undocumented synonym, and the same for
317 --exclude-dir, though that is not available in GNU grep, at least as of
318 release 2.5.4.
320 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
321 characters from a string of bytes) have been redefined so as not to use
322 loops, in order to improve performance in some environments. At the same
323 time, I abstracted some of the common code into auxiliary macros to save
324 repetition (this should not affect the compiled code).
326 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
327 compile-time error is now given if \c is not followed by an ASCII
328 character, that is, a byte less than 128. (In EBCDIC mode, the code is
329 different, and any byte value is allowed.)
331 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
332 START_OPTIMIZE option, which is now allowed at compile time - but just
333 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
334 to pcregrep and other applications that have no direct access to PCRE
335 options. The new /Y option in pcretest sets this option when calling
336 pcre_compile().
338 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
339 back references. Groups containing recursive back references were forced to
340 be atomic by that change, but in the case of named groups, the amount of
341 memory required was incorrectly computed, leading to "Failed: internal
342 error: code overflow". This has been fixed.
344 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
345 pcretest.c, to avoid build problems in some Borland environments.
348 Version 8.10 25-Jun-2010
349 ------------------------
351 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
352 THEN.
354 2. (*ACCEPT) was not working when inside an atomic group.
356 3. Inside a character class, \B is treated as a literal by default, but
357 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
358 causes the error). The code is unchanged, but I tidied the documentation.
360 4. Inside a character class, PCRE always treated \R and \X as literals,
361 whereas Perl faults them if its -w option is set. I have changed PCRE so
362 that it faults them when PCRE_EXTRA is set.
364 5. Added support for \N, which always matches any character other than
365 newline. (It is the same as "." when PCRE_DOTALL is not set.)
367 6. When compiling pcregrep with newer versions of gcc which may have
368 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
369 declared with attribute warn_unused_result" were given. Just casting the
370 result to (void) does not stop the warnings; a more elaborate fudge is
371 needed. I've used a macro to implement this.
373 7. Minor change to pcretest.c to avoid a compiler warning.
375 8. Added four artifical Unicode properties to help with an option to make
376 \s etc use properties (see next item). The new properties are: Xan
377 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
379 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
380 use Unicode properties. (*UCP) at the start of a pattern can be used to set
381 this option. Modified pcretest to add /W to test this facility. Added
382 REG_UCP to make it available via the POSIX interface.
384 10. Added --line-buffered to pcregrep.
386 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
387 studied, and the match started with a letter with a code point greater than
388 127 whose first byte was different to the first byte of the other case of
389 the letter, the other case of this starting letter was not recognized
390 (#976).
392 12. If a pattern that was studied started with a repeated Unicode property
393 test, for example, \p{Nd}+, there was the theoretical possibility of
394 setting up an incorrect bitmap of starting bytes, but fortunately it could
395 not have actually happened in practice until change 8 above was made (it
396 added property types that matched character-matching opcodes).
398 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
399 possible starting bytes for non-anchored patterns.
401 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
402 \R, and also a number of cases that involve Unicode properties, both
403 explicit and implicit when PCRE_UCP is set.
405 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
406 input, it could crash or give wrong results if characters with values
407 greater than 0xc0 were present in the subject string. (Detail: it assumed
408 UTF-8 input when processing these items.)
410 16. Added a lot of (int) casts to avoid compiler warnings in systems where
411 size_t is 64-bit (#991).
413 17. Added a check for running out of memory when PCRE is compiled with
414 --disable-stack-for-recursion (#990).
416 18. If the last data line in a file for pcretest does not have a newline on
417 the end, a newline was missing in the output.
419 19. The default pcre_chartables.c file recognizes only ASCII characters (values
420 less than 128) in its various bitmaps. However, there is a facility for
421 generating tables according to the current locale when PCRE is compiled. It
422 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
423 characters, are recognized by isspace() and therefore were getting set in
424 these tables, and indeed these tables seem to approximate to ISO 8859. This
425 caused a problem in UTF-8 mode when pcre_study() was used to create a list
426 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
427 which of course cannot start UTF-8 characters. I have changed the code so
428 that only real ASCII characters (less than 128) and the correct starting
429 bytes for UTF-8 encodings are set for characters greater than 127 when in
430 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
431 altogether.)
433 20. Added the /T option to pcretest so as to be able to run tests with non-
434 standard character tables, thus making it possible to include the tests
435 used for 19 above in the standard set of tests.
437 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
438 reference to a subpattern the other side of a comment that contains an
439 opening parenthesis caused either an internal compiling error, or a
440 reference to the wrong subpattern.
443 Version 8.02 19-Mar-2010
444 ------------------------
446 1. The Unicode data tables have been updated to Unicode 5.2.0.
448 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
449 configured.
451 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
452 original author of that file, following a query about its status.
454 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
455 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
457 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
458 quantifier applied to a forward-referencing subroutine call, could compile
459 incorrect code or give the error "internal error: previously-checked
460 referenced subpattern not found".
462 6. Both MS Visual Studio and Symbian OS have problems with initializing
463 variables to point to external functions. For these systems, therefore,
464 pcre_malloc etc. are now initialized to local functions that call the
465 relevant global functions.
467 7. There were two entries missing in the vectors called coptable and poptable
468 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
469 I've fixed the data, and added a kludgy way of testing at compile time that
470 the lengths are correct (equal to the number of opcodes).
472 8. Following on from 7, I added a similar kludge to check the length of the
473 eint vector in pcreposix.c.
475 9. Error texts for pcre_compile() are held as one long string to avoid too
476 much relocation at load time. To find a text, the string is searched,
477 counting zeros. There was no check for running off the end of the string,
478 which could happen if a new error number was added without updating the
479 string.
481 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
483 11. \K was not working if it appeared in an atomic group or in a group that
484 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
485 \K is "not well defined" if used in an assertion. PCRE now accepts it if
486 the assertion is positive, but not if it is negative.
488 12. Change 11 fortuitously reduced the size of the stack frame used in the
489 "match()" function of pcre_exec.c by one pointer. Forthcoming
490 implementation of support for (*MARK) will need an extra pointer on the
491 stack; I have reserved it now, so that the stack frame size does not
492 decrease.
494 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
495 item in branch that calls a recursion is a subroutine call - as in the
496 second branch in the above example - was incorrectly given the compile-
497 time error "recursive call could loop indefinitely" because pcre_compile()
498 was not correctly checking the subroutine for matching a non-empty string.
500 14. The checks for overrunning compiling workspace could trigger after an
501 overrun had occurred. This is a "should never occur" error, but it can be
502 triggered by pathological patterns such as hundreds of nested parentheses.
503 The checks now trigger 100 bytes before the end of the workspace.
505 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
508 Version 8.01 19-Jan-2010
509 ------------------------
511 1. If a pattern contained a conditional subpattern with only one branch (in
512 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
513 computed the wrong minimum data length (which is of course zero for such
514 subpatterns). This could cause incorrect "no match" results.
516 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
517 the pattern is reset in the first branch, pcre_compile() failed with
518 "internal error: code overflow at offset...". This happened only when
519 the reset was to the original external option setting. (An optimization
520 abstracts leading options settings into an external setting, which was the
521 cause of this.)
523 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
524 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
525 assertion pattern did not match (meaning that the assertion was true), it
526 was incorrectly treated as false if the SKIP had been reached during the
527 matching. This also applied to assertions used as conditions.
529 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
530 assertion subpattern, including such a pattern used as a condition,
531 unpredictable results occurred, instead of the error return
534 5. The C++ GlobalReplace function was not working like Perl for the special
535 situation when an empty string is matched. It now does the fancy magic
536 stuff that is necessary.
538 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
539 removed. (These were left over from very, very early versions of PCRE.)
541 7. Some cosmetic changes to the code to make life easier when compiling it
542 as part of something else:
544 (a) Change DEBUG to PCRE_DEBUG.
546 (b) In pcre_compile(), rename the member of the "branch_chain" structure
547 called "current" as "current_branch", to prevent a collision with the
548 Linux macro when compiled as a kernel module.
550 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
551 prevent a collision with the Linux macro when compiled as a kernel
552 module.
554 8. In pcre_compile() there are some checks for integer overflows that used to
555 cast potentially large values to (double). This has been changed to that
556 when building, a check for int64_t is made, and if it is found, it is used
557 instead, thus avoiding the use of floating point arithmetic. (There is no
558 other use of FP in PCRE.) If int64_t is not found, the fallback is to
559 double.
561 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
562 2005 (difference between two addresses compared to an unsigned value).
564 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
565 custom one, because of the following reported problem in Windows:
567 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
568 under Win32.
569 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
570 therefore missing the function definition.
571 - The compiler thus generates a "C" signature for the test function.
572 - The linker fails to find the "C" function.
573 - PCRE fails to configure if asked to do so against libbz2.
575 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
576 messages were output:
578 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
579 rerunning libtoolize, to keep the correct libtool macros in-tree.
580 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
582 I have done both of these things.
584 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
585 most of the time, it *can* run out if it is given a pattern that contains a
586 runaway infinite recursion. I updated the discussion in the pcrestack man
587 page.
589 13. Now that we have gone to the x.xx style of version numbers, the minor
590 version may start with zero. Using 08 or 09 is a bad idea because users
591 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
592 interpreted as invalid octal numbers. I've updated the previous comment in
593 configure.ac, and also added a check that gives an error if 08 or 09 are
594 used.
596 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
597 causing partial matching to fail when the end of the subject matched \W
598 in a UTF-8 pattern where \W was quantified with a minimum of 3.
600 15. There were some discrepancies between the declarations in pcre_internal.h
601 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
602 their definitions. The declarations used "const uschar *" and the
603 definitions used USPTR. Even though USPTR is normally defined as "const
604 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
605 reported that: "This difference in casting confuses some C++ compilers, for
606 example, SunCC recognizes above declarations as different functions and
607 generates broken code for hbpcre." I have changed the declarations to use
608 USPTR.
610 16. GNU libtool is named differently on some systems. The autogen.sh script now
611 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
612 (FreeBSD).
614 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
615 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
616 comment: "Figure out how to create a longlong from a string: strtoll and
617 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
618 instance, but it only takes 2 args instead of 3!"
620 18. A subtle bug concerned with back references has been fixed by a change of
621 specification, with a corresponding code fix. A pattern such as
622 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
623 refers, was giving matches when it shouldn't. For example, xa=xaaa would
624 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
625 same bug. Such groups have to be quantified to be useful, or contained
626 inside another quantified group. (If there's no repetition, the reference
627 can never match.) The problem arises because, having left the group and
628 moved on to the rest of the pattern, a later failure that backtracks into
629 the group uses the captured value from the final iteration of the group
630 rather than the correct earlier one. I have fixed this in PCRE by forcing
631 any group that contains a reference to itself to be an atomic group; that
632 is, there cannot be any backtracking into it once it has completed. This is
633 similar to recursive and subroutine calls.
636 Version 8.00 19-Oct-09
637 ----------------------
639 1. The table for translating pcre_compile() error codes into POSIX error codes
640 was out-of-date, and there was no check on the pcre_compile() error code
641 being within the table. This could lead to an OK return being given in
642 error.
644 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
645 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
646 in a Windows environment.
648 3. The pcregrep --count option prints the count for each file even when it is
649 zero, as does GNU grep. However, pcregrep was also printing all files when
650 --files-with-matches was added. Now, when both options are given, it prints
651 counts only for those files that have at least one match. (GNU grep just
652 prints the file name in this circumstance, but including the count seems
653 more useful - otherwise, why use --count?) Also ensured that the
654 combination -clh just lists non-zero counts, with no names.
656 4. The long form of the pcregrep -F option was incorrectly implemented as
657 --fixed_strings instead of --fixed-strings. This is an incompatible change,
658 but it seems right to fix it, and I didn't think it was worth preserving
659 the old behaviour.
661 5. The command line items --regex=pattern and --regexp=pattern were not
662 recognized by pcregrep, which required --regex pattern or --regexp pattern
663 (with a space rather than an '='). The man page documented the '=' forms,
664 which are compatible with GNU grep; these now work.
666 6. No libpcreposix.pc file was created for pkg-config; there was just
667 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
669 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
670 when UCP support is not needed, by modifying the Python script that
671 generates it from Unicode data files. This should not matter if the module
672 is correctly used as a library, but I received one complaint about 50K of
673 unwanted data. My guess is that the person linked everything into his
674 program rather than using a library. Anyway, it does no harm.
676 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
677 was a minimum greater than 1 for a wide character in a possessive
678 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
679 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
680 character. Chaos in the form of incorrect output or a compiling loop could
681 result.
683 9. The restrictions on what a pattern can contain when partial matching is
684 requested for pcre_exec() have been removed. All patterns can now be
685 partially matched by this function. In addition, if there are at least two
686 slots in the offset vector, the offset of the earliest inspected character
687 for the match and the offset of the end of the subject are set in them when
688 PCRE_ERROR_PARTIAL is returned.
690 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
691 synonymous with PCRE_PARTIAL, for backwards compatibility, and
692 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
693 and may be more useful for multi-segment matching.
695 11. Partial matching with pcre_exec() is now more intuitive. A partial match
696 used to be given if ever the end of the subject was reached; now it is
697 given only if matching could not proceed because another character was
698 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
699 string "Z", which now yields "no match" instead of "partial match". In the
700 case of pcre_dfa_exec(), "no match" is given if every matching path for the
701 final character ended with (*FAIL).
703 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
704 if the pattern had a "must contain" character that was already found in the
705 earlier partial match, unless partial matching was again requested. For
706 example, with the pattern /dog.(body)?/, the "must contain" character is
707 "g". If the first part-match was for the string "dog", restarting with
708 "sbody" failed. This bug has been fixed.
710 13. The string returned by pcre_dfa_exec() after a partial match has been
711 changed so that it starts at the first inspected character rather than the
712 first character of the match. This makes a difference only if the pattern
713 starts with a lookbehind assertion or \b or \B (\K is not supported by
714 pcre_dfa_exec()). It's an incompatible change, but it makes the two
715 matching functions compatible, and I think it's the right thing to do.
717 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
718 so that the demonstration program is easily available in environments where
719 PCRE has not been installed from source.
721 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
722 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
723 library.
725 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
726 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
727 is not the first non-POSIX option to be added. Clearly some people find
728 these options useful.
730 17. If a caller to the POSIX matching function regexec() passes a non-zero
731 value for nmatch with a NULL value for pmatch, the value of
732 nmatch is forced to zero.
734 18. RunGrepTest did not have a test for the availability of the -u option of
735 the diff command, as RunTest does. It now checks in the same way as
736 RunTest, and also checks for the -b option.
738 19. If an odd number of negated classes containing just a single character
739 interposed, within parentheses, between a forward reference to a named
740 subpattern and the definition of the subpattern, compilation crashed with
741 an internal error, complaining that it could not find the referenced
742 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
743 [The bug was that it was starting one character too far in when skipping
744 over the character class, thus treating the ] as data rather than
745 terminating the class. This meant it could skip too much.]
747 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
748 /g option in pcretest when the pattern contains \K, which makes it possible
749 to have an empty string match not at the start, even when the pattern is
750 anchored. Updated pcretest and pcredemo to use this option.
752 21. If the maximum number of capturing subpatterns in a recursion was greater
753 than the maximum at the outer level, the higher number was returned, but
754 with unset values at the outer level. The correct (outer level) value is
755 now given.
757 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
758 PCRE did not set those parentheses (unlike Perl). I have now found a way to
759 make it do so. The string so far is captured, making this feature
760 compatible with Perl.
762 23. The tests have been re-organized, adding tests 11 and 12, to make it
763 possible to check the Perl 5.10 features against Perl 5.10.
765 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
766 pattern matches a fixed length string. PCRE did not allow this; now it
767 does. Neither allows recursion.
769 25. I finally figured out how to implement a request to provide the minimum
770 length of subject string that was needed in order to match a given pattern.
771 (It was back references and recursion that I had previously got hung up
772 on.) This code has now been added to pcre_study(); it finds a lower bound
773 to the length of subject needed. It is not necessarily the greatest lower
774 bound, but using it to avoid searching strings that are too short does give
775 some useful speed-ups. The value is available to calling programs via
776 pcre_fullinfo().
778 26. While implementing 25, I discovered to my embarrassment that pcretest had
779 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
780 study optimizations had never been tested with that matching function.
781 Oops. What is worse, even when it was passed study data, there was a bug in
782 pcre_dfa_exec() that meant it never actually used it. Double oops. There
783 were also very few tests of studied patterns with pcre_dfa_exec().
785 27. If (?| is used to create subpatterns with duplicate numbers, they are now
786 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
787 on the other side of the coin, they are no longer allowed to have different
788 names, because these cannot be distinguished in PCRE, and this has caused
789 confusion. (This is a difference from Perl.)
791 28. When duplicate subpattern names are present (necessarily with different
792 numbers, as required by 27 above), and a test is made by name in a
793 conditional pattern, either for a subpattern having been matched, or for
794 recursion in such a pattern, all the associated numbered subpatterns are
795 tested, and the overall condition is true if the condition is true for any
796 one of them. This is the way Perl works, and is also more like the way
797 testing by number works.
800 Version 7.9 11-Apr-09
801 ---------------------
803 1. When building with support for bzlib/zlib (pcregrep) and/or readline
804 (pcretest), all targets were linked against these libraries. This included
805 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
806 libraries. This caused unwanted dependencies to be created. This problem
807 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
808 pcretest is linked with readline.
810 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
811 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
812 moved outside it again, because FALSE and TRUE are already defined in AIX,
813 but BOOL is not.
815 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
816 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
818 4. The pcregrep documentation said spaces were inserted as well as colons (or
819 hyphens) following file names and line numbers when outputting matching
820 lines. This is not true; no spaces are inserted. I have also clarified the
821 wording for the --colour (or --color) option.
823 5. In pcregrep, when --colour was used with -o, the list of matching strings
824 was not coloured; this is different to GNU grep, so I have changed it to be
825 the same.
827 6. When --colo(u)r was used in pcregrep, only the first matching substring in
828 each matching line was coloured. Now it goes on to look for further matches
829 of any of the test patterns, which is the same behaviour as GNU grep.
831 7. A pattern that could match an empty string could cause pcregrep to loop; it
832 doesn't make sense to accept an empty string match in pcregrep, so I have
833 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
834 seems to be how GNU grep behaves.
836 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
837 start or after a newline", because the conditional assertion was not being
838 correctly handled. The rule now is that both the assertion and what follows
839 in the first alternative must satisfy the test.
841 9. If auto-callout was enabled in a pattern with a conditional group whose
842 condition was an assertion, PCRE could crash during matching, both with
843 pcre_exec() and pcre_dfa_exec().
845 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
846 used for matching.
848 11. Unicode property support in character classes was not working for
849 characters (bytes) greater than 127 when not in UTF-8 mode.
851 12. Added the -M command line option to pcretest.
853 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
855 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
857 16. Added comments and documentation about mis-use of no_arg in the C++
858 wrapper.
860 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
861 from Martin Jerabek that uses macro names for all relevant character and
862 string constants.
864 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
865 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
866 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
867 these, but not everybody uses configure.
869 19. A conditional group that had only one branch was not being correctly
870 recognized as an item that could match an empty string. This meant that an
871 enclosing group might also not be so recognized, causing infinite looping
872 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
873 with the subject "ab", where knowledge that the repeated group can match
874 nothing is needed in order to break the loop.
876 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
877 exec(), but without supplying a callout function, matching went wrong.
879 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
880 leak if the size of the offset vector was greater than 30. When the vector
881 is smaller, the saved offsets during recursion go onto a local stack
882 vector, but for larger vectors malloc() is used. It was failing to free
883 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
884 error, in fact).
886 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
887 heapframe that is used only when UTF-8 support is enabled. This caused no
888 problem, but was untidy.
890 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
891 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
892 included within another project.
894 24. Steven Van Ingelgem's patches to add more options to the CMake support,
895 slightly modified by me:
897 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
898 not building pcregrep.
900 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
901 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
903 25. Forward references, both numeric and by name, in patterns that made use of
904 duplicate group numbers, could behave incorrectly or give incorrect errors,
905 because when scanning forward to find the reference group, PCRE was not
906 taking into account the duplicate group numbers. A pattern such as
907 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
909 26. Changed a few more instances of "const unsigned char *" to USPTR, making
910 the feature of a custom pointer more persuasive (as requested by a user).
912 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
913 pcretest.c, inside #ifndefs, because it seems they are sometimes already
914 pre-defined.
916 28. Added support for (*UTF8) at the start of a pattern.
918 29. Arrange for flags added by the "release type" setting in CMake to be shown
919 in the configuration summary.
922 Version 7.8 05-Sep-08
923 ---------------------
925 1. Replaced UCP searching code with optimized version as implemented for Ad
926 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
927 stage table and inline lookup instead of a function, giving speed ups of 2
928 to 5 times on some simple patterns that I tested. Permission was given to
929 distribute the MultiStage2.py script that generates the tables (it's not in
930 the tarball, but is in the Subversion repository).
932 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
933 scripts.
935 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
936 a group with a zero qualifier. The result of the study could be incorrect,
937 or the function might crash, depending on the pattern.
939 4. Caseless matching was not working for non-ASCII characters in back
940 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
941 It now works when Unicode Property Support is available.
943 5. In pcretest, an escape such as \x{de} in the data was always generating
944 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
945 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
946 truncation.
948 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
950 7. Added two (int) casts to pcregrep when printing the difference of two
951 pointers, in case they are 64-bit values.
953 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
954 test 2 if it fails.
956 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
957 and a #define of that name to empty if it is not externally set. This is to
958 allow users of MSVC to set it if necessary.
960 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
961 the convenience functions in the pcre_get.c source file.
963 11. An option change at the start of a pattern that had top-level alternatives
964 could cause overwriting and/or a crash. This command provoked a crash in
965 some environments:
967 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
969 This potential security problem was recorded as CVE-2008-2371.
971 12. For a pattern where the match had to start at the beginning or immediately
972 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
973 pcre_dfa_exec() could read past the end of the passed subject if there was
974 no match. To help with detecting such bugs (e.g. with valgrind), I modified
975 pcretest so that it places the subject at the end of its malloc-ed buffer.
977 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
978 exec() might read past the end of the data buffer in UTF-8 mode.
980 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
981 the data contained the byte 0x85 as part of a UTF-8 character within its
982 first line. This applied both to normal and DFA matching.
984 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
985 /^[^d]*?$/8 failed to match "abc".
987 16. Added a missing copyright notice to pcrecpp_internal.h.
989 17. Make it more clear in the documentation that values returned from
990 pcre_exec() in ovector are byte offsets, not character counts.
992 18. Tidied a few places to stop certain compilers from issuing warnings.
994 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
995 supplied by Stefan Weber. I made a further small update for 7.8 because
996 there is a change of source arrangements: the pcre_searchfuncs.c module is
997 replaced by pcre_ucd.c.
1000 Version 7.7 07-May-08
1001 ---------------------
1003 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1004 a string to a long long, pretend we don't even have a long long." This is
1005 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1007 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1008 pre-7.6 versions, which defined a global no_arg variable instead of putting
1009 it in the RE class. (See also #8 below.)
1011 3. Remove a line of dead code, identified by coverity and reported by Nuno
1012 Lopes.
1014 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1016 (1) The include/exclude patterns were being applied to the whole pathnames
1017 of files, instead of just to the final components.
1019 (2) If there was more than one level of directory, the subdirectories were
1020 skipped unless they satisfied the include/exclude conditions. This is
1021 inconsistent with GNU grep (and could even be seen as contrary to the
1022 pcregrep specification - which I improved to make it absolutely clear).
1023 The action now is always to scan all levels of directory, and just
1024 apply the include/exclude patterns to regular files.
1026 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1027 --exclude_dir in the tests to avoid scanning .svn directories.
1029 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1030 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1031 doesn't support NULs in patterns.
1033 7. Added some missing "const"s to declarations of static tables in
1034 pcre_compile.c and pcre_dfa_exec.c.
1036 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1037 caused by fix #2 above. (Subsequently also a second patch to fix the
1038 first patch. And a third patch - this was a messy problem.)
1040 9. Applied Craig's patch to remove the use of push_back().
1042 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1043 matching function regexec().
1045 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1046 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1047 references. PCRE supports relative numbers with this syntax (I don't think
1048 Oniguruma does).
1050 12. Previously, a group with a zero repeat such as (...){0} was completely
1051 omitted from the compiled regex. However, this means that if the group
1052 was called as a subroutine from elsewhere in the pattern, things went wrong
1053 (an internal error was given). Such groups are now left in the compiled
1054 pattern, with a new opcode that causes them to be skipped at execution
1055 time.
1057 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1058 to the way PCRE behaves:
1060 (a) A lone ] character is dis-allowed (Perl treats it as data).
1062 (b) A back reference to an unmatched subpattern matches an empty string
1063 (Perl fails the current match path).
1065 (c) A data ] in a character class must be notated as \] because if the
1066 first data character in a class is ], it defines an empty class. (In
1067 Perl it is not possible to have an empty class.) The empty class []
1068 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1069 The negative empty class [^] matches any one character, independently
1070 of the DOTALL setting.
1072 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1073 non-existent subpattern following a character class starting with ']' and
1074 containing () gave an internal compiling error instead of "reference to
1075 non-existent subpattern". Fortunately, when the pattern did exist, the
1076 compiled code was correct. (When scanning forwards to check for the
1077 existencd of the subpattern, it was treating the data ']' as terminating
1078 the class, so got the count wrong. When actually compiling, the reference
1079 was subsequently set up correctly.)
1081 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1082 it was being rejected as not supported by pcre_dfa_exec(), even though
1083 other assertions are supported. I have made pcre_dfa_exec() support
1084 (*FAIL).
1086 16. The implementation of 13c above involved the invention of a new opcode,
1087 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1088 cannot be changed at match time, I realized I could make a small
1089 improvement to matching performance by compiling OP_ALLANY instead of
1090 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1091 on the OP_ANY path.
1093 17. Compiling pcretest on Windows with readline support failed without the
1094 following two fixes: (1) Make the unistd.h include conditional on
1095 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1097 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1098 ncurses library to be included for pcretest when ReadLine support is
1099 requested, but also to allow for it to be overridden. This patch came from
1100 Daniel Bergström.
1102 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1103 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1104 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1105 spotting this.
1108 Version 7.6 28-Jan-08
1109 ---------------------
1111 1. A character class containing a very large number of characters with
1112 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1113 overflow.
1115 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1116 HAVE_LONG_LONG is not defined.
1118 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1119 bring them up to date and include new features. This patch includes:
1121 - Fixed PH's badly added libz and libbz2 support.
1122 - Fixed a problem with static linking.
1123 - Added pcredemo. [But later removed - see 7 below.]
1124 - Fixed dftables problem and added an option.
1125 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1127 - Added readline support for pcretest.
1128 - Added an listing of the option settings after cmake has run.
1130 4. A user submitted a patch to Makefile that makes it easy to create
1131 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1132 Makefile.am that cause it to include this special target, without
1133 affecting anything else. Note that the same mingw target plus all
1134 the other distribution libraries and programs are now supported
1135 when configuring with CMake (see 6 below) instead of with
1136 Configure/Make.
1138 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1139 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1140 exported in the Windows port". It has not yet been confirmed that the patch
1141 solves the problem, but it does no harm.
1143 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1144 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1145 with CMake, and also correct the comment about stack recursion.
1147 7. Remove the automatic building of pcredemo from the ./configure system and
1148 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1149 of a program that users should build themselves after PCRE is installed, so
1150 building it automatically is not really right. What is more, it gave
1151 trouble in some build environments.
1153 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1156 Version 7.5 10-Jan-08
1157 ---------------------
1159 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1160 values in parens when parsing an RE using the C++ wrapper."
1162 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1163 Characters greater than 255 were excluded from the class instead of being
1164 included.
1166 3. The same bug as (2) above applied to negated POSIX classes such as
1167 [:^space:].
1169 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1170 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1171 I have changed it.
1173 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1174 first named subpattern) and a construct such as (?&a) would reference the
1175 first named subpattern whose name started with "a" (in other words, the
1176 length check was missing). Both these problems are fixed. "Subpattern name
1177 expected" is now given for (?&) (a zero-length name), and this patch also
1178 makes it give the same error for \k'' (previously it complained that that
1179 was a reference to a non-existent subpattern).
1181 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1182 this is right because (?- can be followed by option settings as well as by
1183 digits. I have, however, made the messages clearer.
1185 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1186 than the number used in the conditional) now cause a compile-time error.
1187 This is actually not compatible with Perl, which accepts such patterns, but
1188 treats the conditional as always being FALSE (as PCRE used to), but it
1189 seems to me that giving a diagnostic is better.
1191 8. Change "alphameric" to the more common word "alphanumeric" in comments
1192 and messages.
1194 9. Fix two occurrences of "backslash" in comments that should have been
1195 "backspace".
1197 10. Remove two redundant lines of code that can never be obeyed (their function
1198 was moved elsewhere).
1200 11. The program that makes PCRE's Unicode character property table had a bug
1201 which caused it to generate incorrect table entries for sequences of
1202 characters that have the same character type, but are in different scripts.
1203 It amalgamated them into a single range, with the script of the first of
1204 them. In other words, some characters were in the wrong script. There were
1205 thirteen such cases, affecting characters in the following ranges:
1207 U+002b0 - U+002c1
1208 U+0060c - U+0060d
1209 U+0061e - U+00612
1210 U+0064b - U+0065e
1211 U+0074d - U+0076d
1212 U+01800 - U+01805
1213 U+01d00 - U+01d77
1214 U+01d9b - U+01dbf
1215 U+0200b - U+0200f
1216 U+030fc - U+030fe
1217 U+03260 - U+0327f
1218 U+0fb46 - U+0fbb1
1219 U+10450 - U+1049d
1221 12. The -o option (show only the matching part of a line) for pcregrep was not
1222 compatible with GNU grep in that, if there was more than one match in a
1223 line, it showed only the first of them. It now behaves in the same way as
1224 GNU grep.
1226 13. If the -o and -v options were combined for pcregrep, it printed a blank
1227 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1228 does the same. The return code can be used to tell if there were any
1229 non-matching lines.
1231 14. Added --file-offsets and --line-offsets to pcregrep.
1233 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1234 infinitely looping recursion. The bug was that positive lookaheads were not
1235 being skipped when checking for a possible empty match (negative lookaheads
1236 and both kinds of lookbehind were skipped).
1238 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1239 inclusion of <windows.h> to before rather than after the definition of
1240 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1242 17. Specifying a possessive quantifier with a specific limit for a Unicode
1243 character property caused pcre_compile() to compile bad code, which led at
1244 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1245 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1246 caused the error; without that there was no problem.
1248 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1250 19. Added --enable-pcretest-libreadline.
1252 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1253 RE::GlobalReplace(). As a result, the number of replacements returned was
1254 double what it should be. I removed one of the increments, but Craig sent a
1255 later patch that removed the other one (the right fix) and added unit tests
1256 that check the return values (which was not done before).
1258 21. Several CMake things:
1260 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1261 the names libpcre and libpcreposix, not just pcre and pcreposix.
1263 (2) The above change means that pcretest and pcregrep are now correctly
1264 linked with the newly-built libraries, not previously installed ones.
1268 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1269 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1270 UTF-8 newline character). The key issue is that the pattern starts .*;
1271 this means that the match must be either at the beginning, or after a
1272 newline. The bug was in the code for advancing after a failed match and
1273 checking that the new position followed a newline. It was not taking
1274 account of UTF-8 characters correctly.
1276 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1277 character classes. PCRE was not treating the sequence [:...:] as a
1278 character class unless the ... were all letters. Perl, however, seems to
1279 allow any characters between [: and :], though of course it rejects as
1280 unknown any "names" that contain non-letters, because all the known class
1281 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1282 for example, whereas PCRE did not - it did not recognize a POSIX character
1283 class. This seemed a bit dangerous, so the code has been changed to be
1284 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1285 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1286 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1287 Perl does, and where it didn't before.
1289 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1290 Windows environments %n is disabled by default.
1293 Version 7.4 21-Sep-07
1294 ---------------------
1296 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1297 means that a class such as [\s] counted as "explicit reference to CR or
1298 LF". That isn't really right - the whole point of the change was to try to
1299 help when there was an actual mention of one of the two characters. So now
1300 the change happens only if \r or \n (or a literal CR or LF) character is
1301 encountered.
1303 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1304 of both had grown to the point where there were only 3 bits left.
1305 Fortunately, there was spare space in the data structure, and so I have
1306 moved the internal flags into a new 16-bit field to free up more option
1307 bits.
1309 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1310 but did not set the internal JCHANGED flag - either of these is enough to
1311 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1312 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1313 start sets both bits.
1315 4. Added options (at build time, compile time, exec time) to change \R from
1316 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1318 5. doc/pcresyntax.html was missing from the distribution.
1320 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1321 compatibility, even though it is no longer used.
1323 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1324 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1325 windows.h file is present (where different names are used). [This was
1326 reversed later after testing - see 16 below.]
1328 8. Changed all #include <config.h> to #include "config.h". There were also
1329 some further <pcre.h> cases that I changed to "pcre.h".
1331 9. When pcregrep was used with the --colour option, it missed the line ending
1332 sequence off the lines that it output.
1334 10. It was pointed out to me that arrays of string pointers cause lots of
1335 relocations when a shared library is dynamically loaded. A technique of
1336 using a single long string with a table of offsets can drastically reduce
1337 these. I have refactored PCRE in four places to do this. The result is
1338 dramatic:
1340 Originally: 290
1341 After changing UCP table: 187
1342 After changing error message table: 43
1343 After changing table of "verbs" 36
1344 After changing table of Posix names 22
1346 Thanks to the folks working on Gregex for glib for this insight.
1348 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1349 unicode-properties was also set.
1351 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1353 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1354 checked only for CRLF.
1356 14. Added casts to pcretest.c to avoid compiler warnings.
1358 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1360 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1361 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1362 entirely. This removes changes made in 7 above.
1364 17. The CMake files have been updated, and there is now more information about
1365 building with CMake in the NON-UNIX-USE document.
1368 Version 7.3 28-Aug-07
1369 ---------------------
1371 1. In the rejigging of the build system that eventually resulted in 7.1, the
1372 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1373 brackets there is not right, since it causes compilers to look for an
1374 installed pcre.h, not the version that is in the source that is being
1375 compiled (which of course may be different). I have changed it back to:
1377 #include "pcre.h"
1379 I have a vague recollection that the change was concerned with compiling in
1380 different directories, but in the new build system, that is taken care of
1381 by the VPATH setting the Makefile.
1383 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1384 when the subject happened to end in the byte 0x85 (e.g. if the last
1385 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1386 characters but of course it shouldn't be taken as a newline when it is part
1387 of another character. The bug was that, for an unlimited repeat of . in
1388 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1389 characters when looking for a newline.
1391 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1393 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1394 in debug output.
1396 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1397 long printing in the pcrecpp unittest when running under MinGW.
1399 6. ESC_K was left out of the EBCDIC table.
1401 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1402 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1403 limit also applies to "virtual nesting" when a pattern is recursive, and in
1404 this case 1000 isn't so big. I have been able to remove this limit at the
1405 expense of backing off one optimization in certain circumstances. Normally,
1406 when pcre_exec() would call its internal match() function recursively and
1407 immediately return the result unconditionally, it uses a "tail recursion"
1408 feature to save stack. However, when a subpattern that can match an empty
1409 string has an unlimited repetition quantifier, it no longer makes this
1410 optimization. That gives it a stack frame in which to save the data for
1411 checking that an empty string has been matched. Previously this was taken
1412 from the 1000-entry workspace that had been reserved. So now there is no
1413 explicit limit, but more stack is used.
1415 8. Applied Daniel's patches to solve problems with the import/export magic
1416 syntax that is required for Windows, and which was going wrong for the
1417 pcreposix and pcrecpp parts of the library. These were overlooked when this
1418 problem was solved for the main library.
1420 9. There were some crude static tests to avoid integer overflow when computing
1421 the size of patterns that contain repeated groups with explicit upper
1422 limits. As the maximum quantifier is 65535, the maximum group length was
1423 set at 30,000 so that the product of these two numbers did not overflow a
1424 32-bit integer. However, it turns out that people want to use groups that
1425 are longer than 30,000 bytes (though not repeat them that many times).
1426 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1427 made it possible to implement the integer overflow checks in a much more
1428 dynamic way, which I have now done. The artificial limitation on group
1429 length has been removed - we now have only the limit on the total length of
1430 the compiled pattern, which depends on the LINK_SIZE setting.
1432 10. Fixed a bug in the documentation for get/copy named substring when
1433 duplicate names are permitted. If none of the named substrings are set, the
1434 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1435 empty string.
1437 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1438 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1439 because the ] is interpreted as the first data character and the
1440 terminating ] is not found. PCRE has been made compatible with Perl in this
1441 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1442 cause memory overwriting.
1444 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1445 string has been matched (to stop an infinite loop). It was not recognizing
1446 a conditional subpattern that could match an empty string if that
1447 subpattern was within another subpattern. For example, it looped when
1448 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1449 condition was not nested. This bug has been fixed.
1451 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1452 past the start of the subject in the presence of bytes with the top bit
1453 set, for example "\x8aBCD".
1455 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1456 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1458 14. Optimized (?!) to (*FAIL).
1460 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1461 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1462 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1463 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1464 does: it's just the validity check that is more restrictive.
1466 16. Inserted checks for integer overflows during escape sequence (backslash)
1467 processing, and also fixed erroneous offset values for syntax errors during
1468 backslash processing.
1470 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1471 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1473 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1474 caused an overrun.
1476 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1477 something other than just ASCII characters) inside a group that had an
1478 unlimited repeat caused a loop at compile time (while checking to see
1479 whether the group could match an empty string).
1481 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1482 [\P{Any}] did so. (Error in the code for printing property names.)
1484 21. An orphan \E inside a character class could cause a crash.
1486 22. A repeated capturing bracket such as (A)? could cause a wild memory
1487 reference during compilation.
1489 23. There are several functions in pcre_compile() that scan along a compiled
1490 expression for various reasons (e.g. to see if it's fixed length for look
1491 behind). There were bugs in these functions when a repeated \p or \P was
1492 present in the pattern. These operators have additional parameters compared
1493 with \d, etc, and these were not being taken into account when moving along
1494 the compiled data. Specifically:
1496 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1497 length.
1499 (b) An item such as \pL+ within a repeated group could cause crashes or
1500 loops.
1502 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1503 "reference to non-existent subpattern" error.
1505 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1507 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1508 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1510 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1512 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1513 character were causing crashes (broken optimization).
1515 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1516 \p or \P) caused a compile-time loop.
1518 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1519 break. For example, the unstudied pattern [\r\n]A does not match the string
1520 "\r\nA" because change 7.0/46 below moves the current point on by two
1521 characters after failing to match at the start. However, the pattern \nA
1522 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1523 the same is true. There doesn't seem any very clean way out of this, but
1524 what I have chosen to do makes the common cases work: PCRE now takes note
1525 of whether there can be an explicit match for \r or \n anywhere in the
1526 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1527 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1528 pattern has explicit CR or LF references.
1530 29. Added (*CR) etc for changing newline setting at start of pattern.
1533 Version 7.2 19-Jun-07
1534 ---------------------
1536 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1537 which is apparently normally available under Windows.
1539 2. Re-jig the pcregrep tests with different newline settings in an attempt
1540 to make them independent of the local environment's newline setting.
1542 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1544 4. Some of the "internals" tests were previously cut out when the link size
1545 was not 2, because the output contained actual offsets. The recent new
1546 "Z" feature of pcretest means that these can be cut out, making the tests
1547 usable with all link sizes.
1549 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1550 stack recursion. This gives a massive performance boost under BSD, but just
1551 a small improvement under Linux. However, it saves one field in the frame
1552 in all cases.
1554 6. Added more features from the forthcoming Perl 5.10:
1556 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1557 recursion call. It refers to the nth most recently opened parentheses.
1559 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1560 to be opened parentheses.
1562 (c) Conditions that refer to capturing parentheses can be specified
1563 relatively, for example, (?(-2)... or (?(+3)...
1565 (d) \K resets the start of the current match so that everything before
1566 is not part of it.
1568 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1570 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1571 reference syntax.
1573 (g) (?| introduces a group in which the numbering of parentheses in each
1574 alternative starts with the same number.
1576 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1578 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1581 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1582 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1583 for detecting groups that can match an empty string.
1585 9. A pattern with a very large number of alternatives (more than several
1586 hundred) was running out of internal workspace during the pre-compile
1587 phase, where pcre_compile() figures out how much memory will be needed. A
1588 bit of new cunning has reduced the workspace needed for groups with
1589 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1590 workspace instead of running out of the 4096 that are available.
1592 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1594 11. Applied patch from Google to remove an optimization that didn't quite work.
1595 The report of the bug said:
1597 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1598 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1599 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1601 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1602 it matched the wrong number of bytes.
1605 Version 7.1 24-Apr-07
1606 ---------------------
1608 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1609 that is more "standard", making use of automake and other Autotools. There
1610 is some re-arrangement of the files and adjustment of comments consequent
1611 on this.
1613 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1614 for recursive directory scanning broke on some systems because the files
1615 are not scanned in any specific order and on different systems the order
1616 was different. A call to "sort" has been inserted into RunGrepTest for the
1617 approprate test as a short-term fix. In the longer term there may be an
1618 alternative.
1620 3. I had an email from Eric Raymond about problems translating some of PCRE's
1621 man pages to HTML (despite the fact that I distribute HTML pages, some
1622 people do their own conversions for various reasons). The problems
1623 concerned the use of low-level troff macros .br and .in. I have therefore
1624 removed all such uses from the man pages (some were redundant, some could
1625 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1626 HTML has been updated to handle .nf/.fi and to complain if it encounters
1627 .br or .in.
1629 4. Updated comments in configure.ac that get placed in config.h.in and also
1630 arranged for config.h to be included in the distribution, with the name
1631 config.h.generic, for the benefit of those who have to compile without
1632 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1634 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1635 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1636 makevp.bat for the current PCRE, using the additional files
1637 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1639 6. A Windows user reported a minor discrepancy with test 2, which turned out
1640 to be caused by a trailing space on an input line that had got lost in his
1641 copy. The trailing space was an accident, so I've just removed it.
1643 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1644 that is needed.
1646 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1647 as "const" (a) because they are and (b) because it helps the PHP
1648 maintainers who have recently made a script to detect big data structures
1649 in the php code that should be moved to the .rodata section. I remembered
1650 to update Builducptable as well, so it won't revert if ucptable.h is ever
1651 re-created.
1653 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1654 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1655 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1656 support is not required. This saves 1.5-2K of code, which is important in
1657 some applications.
1659 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1660 so as not to refer to the tables, even though these functions will never be
1661 called when UTF-8 support is disabled. Otherwise there are problems with a
1662 shared library.
1664 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1666 (a) It was defining its arguments as char * instead of void *.
1668 (b) It was assuming that all moves were upwards in memory; this was true
1669 a long time ago when I wrote it, but is no longer the case.
1671 The emulated memove() is provided for those environments that have neither
1672 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1673 is clearly not the case, as these two bugs were recently reported.
1675 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1676 and Detrail to create the HTML documentation, the .txt form of the man
1677 pages, and it removes trailing spaces from listed files. It also creates
1678 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1679 case, it wraps all the #defines with #ifndefs. This script should be run
1680 before "make dist".
1682 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1683 with Unicode property support.
1685 (a) For a maximizing quantifier, if the two different cases of the
1686 character were of different lengths in their UTF-8 codings (there are
1687 some cases like this - I found 11), and the matching function had to
1688 back up over a mixture of the two cases, it incorrectly assumed they
1689 were both the same length.
1691 (b) When PCRE was configured to use the heap rather than the stack for
1692 recursion during matching, it was not correctly preserving the data for
1693 the other case of a UTF-8 character when checking ahead for a match
1694 while processing a minimizing repeat. If the check also involved
1695 matching a wide character, but failed, corruption could cause an
1696 erroneous result when trying to check for a repeat of the original
1697 character.
1699 13. Some tidying changes to the testing mechanism:
1701 (a) The RunTest script now detects the internal link size and whether there
1702 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1703 values substituted by "configure". (The RunGrepTest script already did
1704 this for UTF-8.) The configure.ac script no longer substitutes the
1705 relevant variables.
1707 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1708 with length and offset values. This means that the output is different
1709 for different internal link sizes. Test 2 is skipped for link sizes
1710 other than 2 because of this, bypassing the problem. Unfortunately,
1711 there was also a test in test 3 (the locale tests) that used /B and
1712 failed for link sizes other than 2. Rather than cut the whole test out,
1713 I have added a new /Z option to pcretest that replaces the length and
1714 offset values with spaces. This is now used to make test 3 independent
1715 of link size. (Test 2 will be tidied up later.)
1717 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1718 segmentation fault instead of returning the appropriate error message.
1720 15. In multiline mode when the newline sequence was set to "any", the pattern
1721 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1722 This doesn't seem right; it now treats the CRLF combination as the line
1723 ending, and so does not match in that case. It's only a pattern such as ^$
1724 that would hit this one: something like ^ABC$ would have failed after \r
1725 and then tried again after \r\n.
1727 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1728 in an attempt to make files that differ only in their line terminators
1729 compare equal. This works on Linux.
1731 17. Under certain error circumstances pcregrep might try to free random memory
1732 as it exited. This is now fixed, thanks to valgrind.
1734 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1735 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1736 was because its rules for how to advance for /g after matching an empty
1737 string at the end of a line did not allow for this case. They now check for
1738 it specially.
1740 20. pcretest is supposed to handle patterns and data of any length, by
1741 extending its buffers when necessary. It was getting this wrong when the
1742 buffer for a data line had to be extended.
1744 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1745 CRLF as a newline sequence.
1747 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1748 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1749 I have nevertheless tidied it up.
1751 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1753 24. Added a man page for pcre-config.
1756 Version 7.0 19-Dec-06
1757 ---------------------
1759 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1760 moving to gcc 4.1.1.
1762 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1763 sys/time.h, which is documented as needed for this function. It doesn't
1764 seem to matter on Linux, but it showed up on some releases of OS X.
1766 3. It seems that there are systems where bytes whose values are greater than
1767 127 match isprint() in the "C" locale. The "C" locale should be the
1768 default when a C program starts up. In most systems, only ASCII printing
1769 characters match isprint(). This difference caused the output from pcretest
1770 to vary, making some of the tests fail. I have changed pcretest so that:
1772 (a) When it is outputting text in the compiled version of a pattern, bytes
1773 other than 32-126 are always shown as hex escapes.
1775 (b) When it is outputting text that is a matched part of a subject string,
1776 it does the same, unless a different locale has been set for the match
1777 (using the /L modifier). In this case, it uses isprint() to decide.
1779 4. Fixed a major bug that caused incorrect computation of the amount of memory
1780 required for a compiled pattern when options that changed within the
1781 pattern affected the logic of the preliminary scan that determines the
1782 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1783 that the computed length was too small. The symptoms of this bug were
1784 either the PCRE error "internal error: code overflow" from pcre_compile(),
1785 or a glibc crash with a message such as "pcretest: free(): invalid next
1786 size (fast)". Examples of patterns that provoked this bug (shown in
1787 pcretest format) are:
1789 /(?-x: )/x
1790 /(?x)(?-x: \s*#\s*)/
1791 /((?i)[\x{c0}])/8
1792 /(?i:[\x{c0}])/8
1794 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1795 is now done differently.
1797 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1798 wrapper classes; (b) implement a new function in the C++ scanner that is
1799 more efficient than the old way of doing things because it avoids levels of
1800 recursion in the regex matching; (c) add a paragraph to the documentation
1801 for the FullMatch() function.
1803 6. The escape sequence \n was being treated as whatever was defined as
1804 "newline". Not only was this contrary to the documentation, which states
1805 that \n is character 10 (hex 0A), but it also went horribly wrong when
1806 "newline" was defined as CRLF. This has been fixed.
1808 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1809 was being set to -1 for the "end of line" case (supposedly a value that no
1810 character can have). Though this value is never used (the check for end of
1811 line is "zero bytes in current character"), it caused compiler complaints.
1812 I've changed it to 0xffffffff.
1814 8. In pcre_version.c, the version string was being built by a sequence of
1815 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1816 string (as it is for production releases) called a macro with an empty
1817 argument. The C standard says the result of this is undefined. The gcc
1818 compiler treats it as an empty string (which was what was wanted) but it is
1819 reported that Visual C gives an error. The source has been hacked around to
1820 avoid this problem.
1822 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1823 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1824 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1825 of them did).
1827 10. Originally, pcretest opened its input and output without "b"; then I was
1828 told that "b" was needed in some environments, so it was added for release
1829 5.0 to both the input and output. (It makes no difference on Unix-like
1830 systems.) Later I was told that it is wrong for the input on Windows. I've
1831 now abstracted the modes into two macros, to make it easier to fiddle with
1832 them, and removed "b" from the input mode under Windows.
1834 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1836 12. Added -help and --help to pcretest as an official way of being reminded
1837 of the options.
1839 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1840 and pcrecpp.cc because they annoy compilers at high warning levels.
1842 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1844 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1845 scripts are not C programs :-) and which was not noticed because it works
1846 on Linux.
1848 16. pcretest is supposed to handle any length of pattern and data line (as one
1849 line or as a continued sequence of lines) by extending its input buffer if
1850 necessary. This feature was broken for very long pattern lines, leading to
1851 a string of junk being passed to pcre_compile() if the pattern was longer
1852 than about 50K.
1854 17. I have done a major re-factoring of the way pcre_compile() computes the
1855 amount of memory needed for a compiled pattern. Previously, there was code
1856 that made a preliminary scan of the pattern in order to do this. That was
1857 OK when PCRE was new, but as the facilities have expanded, it has become
1858 harder and harder to keep it in step with the real compile phase, and there
1859 have been a number of bugs (see for example, 4 above). I have now found a
1860 cunning way of running the real compile function in a "fake" mode that
1861 enables it to compute how much memory it would need, while actually only
1862 ever using a few hundred bytes of working memory and without too many
1863 tests of the mode. This should make future maintenance and development
1864 easier. A side effect of this work is that the limit of 200 on the nesting
1865 depth of parentheses has been removed (though this was never a serious
1866 limitation, I suspect). However, there is a downside: pcre_compile() now
1867 runs more slowly than before (30% or more, depending on the pattern). I
1868 hope this isn't a big issue. There is no effect on runtime performance.
1870 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1871 newline (only possible for the last line of a file) and it was a
1872 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1874 19. Added additional timing features to pcretest. (1) The -tm option now times
1875 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1876 separate command line item, by a number that specifies the number of
1877 repeats to use when timing. The default is 50000; this gives better
1878 precision, but takes uncomfortably long for very large patterns.
1880 20. Extended pcre_study() to be more clever in cases where a branch of a
1881 subpattern has no definite first character. For example, (a*|b*)[cd] would
1882 previously give no result from pcre_study(). Now it recognizes that the
1883 first character must be a, b, c, or d.
1885 21. There was an incorrect error "recursive call could loop indefinitely" if
1886 a subpattern (or the entire pattern) that was being tested for matching an
1887 empty string contained only one non-empty item after a nested subpattern.
1888 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1889 incorrectly, because the \d was being skipped in the check.
1891 22. The pcretest program now has a new pattern option /B and a command line
1892 option -b, which is equivalent to adding /B to every pattern. This causes
1893 it to show the compiled bytecode, without the additional information that
1894 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1895 is the same as /B/I).
1897 23. A new optimization is now able automatically to treat some sequences such
1898 as a*b as a*+b. More specifically, if something simple (such as a character
1899 or a simple class like \d) has an unlimited quantifier, and is followed by
1900 something that cannot possibly match the quantified thing, the quantifier
1901 is automatically "possessified".
1903 24. A recursive reference to a subpattern whose number was greater than 39
1904 went wrong under certain circumstances in UTF-8 mode. This bug could also
1905 have affected the operation of pcre_study().
1907 25. Realized that a little bit of performance could be had by replacing
1908 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1910 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1912 27. Possessive quantifiers such as a++ were previously implemented by turning
1913 them into atomic groups such as ($>a+). Now they have their own opcodes,
1914 which improves performance. This includes the automatically created ones
1915 from 23 above.
1917 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1918 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1919 the first matched character to be a colon. This applied both to named and
1920 numbered groups.
1922 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1924 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1925 building PCRE on Windows easier, so I have included it in the distribution.
1927 31. There is now a check in pcretest against a ridiculously large number being
1928 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1929 loop, the loop is abandoned.
1931 32. Forward references to subpatterns in conditions such as (?(2)...) where
1932 subpattern 2 is defined later cause pcre_compile() to search forwards in
1933 the pattern for the relevant set of parentheses. This search went wrong
1934 when there were unescaped parentheses in a character class, parentheses
1935 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1937 33. "Subroutine" calls and backreferences were previously restricted to
1938 referencing subpatterns earlier in the regex. This restriction has now
1939 been removed.
1941 34. Added a number of extra features that are going to be in Perl 5.10. On the
1942 whole, these are just syntactic alternatives for features that PCRE had
1943 previously implemented using the Python syntax or my own invention. The
1944 other formats are all retained for compatibility.
1946 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1947 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1948 also .NET compatible.
1950 (b) A recursion or subroutine call to a named group can now be defined as
1951 (?&name) as well as (?P>name).
1953 (c) A backreference to a named group can now be defined as \k<name> or
1954 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1955 5.10, are also .NET compatible.
1957 (d) A conditional reference to a named group can now use the syntax
1958 (?(<name>) or (?('name') as well as (?(name).
1960 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1961 groups (named and numbered) that are never evaluated inline, but can be
1962 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1963 is always false. There may be only one alternative in such a group.
1965 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1966 as the simple (?(R). The condition is true only if the most recent
1967 recursion is that of the given number or name. It does not search out
1968 through the entire recursion stack.
1970 (g) The escape \gN or \g{N} has been added, where N is a positive or
1971 negative number, specifying an absolute or relative reference.
1973 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1974 some "unreachable code" warnings.
1976 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1977 things, this adds five new scripts.
1979 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1980 There were also incompatibilities regarding the handling of \Q..\E inside
1981 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1982 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1984 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1985 matches an empty string, and forcibly breaks the loop. There were bugs in
1986 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1987 against aaaa the result was just "a" rather than "aaaa", for example. Two
1988 separate and independent bugs (that affected different cases) have been
1989 fixed.
1991 39. Refactored the code to abolish the use of different opcodes for small
1992 capturing bracket numbers. This is a tidy that I avoided doing when I
1993 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1994 The new approach is not only tidier, it makes it possible to reduce the
1995 memory needed to fix the previous bug (38).
1997 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1998 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1999 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2000 mode.
2002 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2003 report.
2005 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2006 copy construction and assignment in the C++ wrapper.
2008 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2009 couple of bugs that could have given wrong results in the "--newline=crlf"
2010 case.
2012 44. Added a number of casts and did some reorganization of signed/unsigned int
2013 variables following suggestions from Dair Grant. Also renamed the variable
2014 "this" as "item" because it is a C++ keyword.
2016 45. Arranged for dftables to add
2018 #include "pcre_internal.h"
2020 to pcre_chartables.c because without it, gcc 4.x may remove the array
2021 definition from the final binary if PCRE is built into a static library and
2022 dead code stripping is activated.
2024 46. For an unanchored pattern, if a match attempt fails at the start of a
2025 newline sequence, and the newline setting is CRLF or ANY, and the next two
2026 characters are CRLF, advance by two characters instead of one.
2029 Version 6.7 04-Jul-06
2030 ---------------------
2032 1. In order to handle tests when input lines are enormously long, pcretest has
2033 been re-factored so that it automatically extends its buffers when
2034 necessary. The code is crude, but this _is_ just a test program. The
2035 default size has been increased from 32K to 50K.
2037 2. The code in pcre_study() was using the value of the re argument before
2038 testing it for NULL. (Of course, in any sensible call of the function, it
2039 won't be NULL.)
2041 3. The memmove() emulation function in pcre_internal.h, which is used on
2042 systems that lack both memmove() and bcopy() - that is, hardly ever -
2043 was missing a "static" storage class specifier.
2045 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2046 containing an extended class (one that cannot be represented by a bitmap
2047 because it contains high-valued characters or Unicode property items, e.g.
2048 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2049 pattern, but PCRE should not loop if you do not (it no longer does).
2050 [Detail: two cases were found: (a) a repeated subpattern containing an
2051 extended class; (b) a recursive reference to a subpattern that followed a
2052 previous extended class. It wasn't skipping over the extended class
2053 correctly when UTF-8 mode was not set.]
2055 5. A negated single-character class was not being recognized as fixed-length
2056 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2057 compile error "lookbehind assertion is not fixed length".
2059 6. The RunPerlTest auxiliary script was showing an unexpected difference
2060 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2061 write a Perl script that can interpret lines of an input file either as
2062 byte characters or as UTF-8, which is what "perltest" was being required to
2063 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2064 can't do is switch easily at run time between having the "use utf8;" pragma
2065 or not. In the end, I fudged it by using the RunPerlTest script to insert
2066 "use utf8;" explicitly for the UTF-8 tests.
2068 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2069 the end of the subject string, contrary to the documentation and to what
2070 Perl does. This was true of both matching functions. Now it matches only at
2071 the start of the subject and immediately after *internal* newlines.
2073 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2074 a pointer to an int instead of a pointer to an unsigned long int. This
2075 caused problems on 64-bit systems.
2077 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2078 instance of the 'standard' template library not being so standard".
2080 10. There was no check on the number of named subpatterns nor the maximum
2081 length of a subpattern name. The product of these values is used to compute
2082 the size of the memory block for a compiled pattern. By supplying a very
2083 long subpattern name and a large number of named subpatterns, the size
2084 computation could be caused to overflow. This is now prevented by limiting
2085 the length of names to 32 characters, and the number of named subpatterns
2086 to 10,000.
2088 11. Subpatterns that are repeated with specific counts have to be replicated in
2089 the compiled pattern. The size of memory for this was computed from the
2090 length of the subpattern and the repeat count. The latter is limited to
2091 65535, but there was no limit on the former, meaning that integer overflow
2092 could in principle occur. The compiled length of a repeated subpattern is
2093 now limited to 30,000 bytes in order to prevent this.
2095 12. Added the optional facility to have named substrings with the same name.
2097 13. Added the ability to use a named substring as a condition, using the
2098 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2099 are numbers (not recommended). Forward references are permitted.
2101 14. Added forward references in named backreferences (if you see what I mean).
2103 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2104 pattern could run off the end of the subject. For example, the pattern
2105 "(?s)(.{1,5})"8 did this with the subject "ab".
2107 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2108 PCRE_CASELESS was set when matching characters that were quantified with ?
2109 or *.
2111 17. A character class other than a single negated character that had a minimum
2112 but no maximum quantifier - for example [ab]{6,} - was not handled
2113 correctly by pce_dfa_exec(). It would match only one character.
2115 18. A valid (though odd) pattern that looked like a POSIX character
2116 class but used an invalid character after [ (for example [[,abc,]]) caused
2117 pcre_compile() to give the error "Failed: internal error: code overflow" or
2118 in some cases to crash with a glibc free() error. This could even happen if
2119 the pattern terminated after [[ but there just happened to be a sequence of
2120 letters, a binary zero, and a closing ] in the memory that followed.
2122 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2123 over the years. Originally (before any Unicode support), just the bottom 8
2124 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2125 output from "man perlunicode" includes this:
2127 The regular expression compiler produces polymorphic opcodes. That
2128 is, the pattern adapts to the data and automatically switches to
2129 the Unicode character scheme when presented with Unicode data--or
2130 instead uses a traditional byte scheme when presented with byte
2131 data.
2133 Sadly, a wide octal escape does not cause a switch, and in a string with
2134 no other multibyte characters, these octal escapes are treated as before.
2135 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2136 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2137 Unicode string.
2139 I have not perpetrated such confusion in PCRE. Up till now, it took just
2140 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2141 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2142 translate to the appropriate multibyte character.
2144 29. Applied some refactoring to reduce the number of warnings from Microsoft
2145 and Borland compilers. This has included removing the fudge introduced
2146 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2147 a warning about an unused variable.
2149 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2150 characters since release 4.0, because Perl (from release 5.004) does not.
2151 [Or at least, is documented not to: some releases seem to be in conflict
2152 with the documentation.] However, when a pattern was studied with
2153 pcre_study() and all its branches started with \s, PCRE still included VT
2154 as a possible starting character. Of course, this did no harm; it just
2155 caused an unnecessary match attempt.
2157 22. Removed a now-redundant internal flag bit that recorded the fact that case
2158 dependency changed within the pattern. This was once needed for "required
2159 byte" processing, but is no longer used. This recovers a now-scarce options
2160 bit. Also moved the least significant internal flag bit to the most-
2161 significant bit of the word, which was not previously used (hangover from
2162 the days when it was an int rather than a uint) to free up another bit for
2163 the future.
2165 23. Added support for CRLF line endings as well as CR and LF. As well as the
2166 default being selectable at build time, it can now be changed at runtime
2167 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2168 specify that it is scanning data with non-default line endings.
2170 24. Changed the definition of CXXLINK to make it agree with the definition of
2171 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2173 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2174 recursions. This makes a big different to stack usage for some patterns.
2176 26. If a subpattern containing a named recursion or subroutine reference such
2177 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2178 the space required for the compiled pattern went wrong and gave too small a
2179 value. Depending on the environment, this could lead to "Failed: internal
2180 error: code overflow at offset 49" or "glibc detected double free or
2181 corruption" errors.
2183 27. Applied patches from Google (a) to support the new newline modes and (b) to
2184 advance over multibyte UTF-8 characters in GlobalReplace.
2186 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2187 difference for some implementation of PCRE in some Windows version.
2189 29. Added some extra testing facilities to pcretest:
2191 \q<number> in a data line sets the "match limit" value
2192 \Q<number> in a data line sets the "match recursion limt" value
2193 -S <number> sets the stack size, where <number> is in megabytes
2195 The -S option isn't available for Windows.
2198 Version 6.6 06-Feb-06
2199 ---------------------
2201 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2202 in pcreposix.h. I have copied the definition from pcre.h.
2204 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2205 because pcre.h is no longer a built file.
2207 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2208 not normally included in the compiled code.
2211 Version 6.5 01-Feb-06
2212 ---------------------
2214 1. When using the partial match feature with pcre_dfa_exec(), it was not
2215 anchoring the second and subsequent partial matches at the new starting
2216 point. This could lead to incorrect results. For example, with the pattern
2217 /1234/, partially matching against "123" and then "a4" gave a match.
2219 2. Changes to pcregrep:
2221 (a) All non-match returns from pcre_exec() were being treated as failures
2222 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2223 error message is output. Some extra information is given for the
2225 probably the only errors that are likely to be caused by users (by
2226 specifying a regex that has nested indefinite repeats, for instance).
2227 If there are more than 20 of these errors, pcregrep is abandoned.
2229 (b) A binary zero was treated as data while matching, but terminated the
2230 output line if it was written out. This has been fixed: binary zeroes
2231 are now no different to any other data bytes.
2233 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2234 used to set a locale for matching. The --locale=xxxx long option has
2235 been added (no short equivalent) to specify a locale explicitly on the
2236 pcregrep command, overriding the environment variables.
2238 (d) When -B was used with -n, some line numbers in the output were one less
2239 than they should have been.
2241 (e) Added the -o (--only-matching) option.
2243 (f) If -A or -C was used with -c (count only), some lines of context were
2244 accidentally printed for the final match.
2246 (g) Added the -H (--with-filename) option.
2248 (h) The combination of options -rh failed to suppress file names for files
2249 that were found from directory arguments.
2251 (i) Added the -D (--devices) and -d (--directories) options.
2253 (j) Added the -F (--fixed-strings) option.
2255 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2257 (l) Added the --colo(u)r option.
2259 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2260 is not present by default.
2262 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2263 items such as (?R) or (?1), when the recursion could match a number of
2264 alternatives. If it matched one of the alternatives, but subsequently,
2265 outside the recursion, there was a failure, the code tried to back up into
2266 the recursion. However, because of the way PCRE is implemented, this is not
2267 possible, and the result was an incorrect result from the match.
2269 In order to prevent this happening, the specification of recursion has
2270 been changed so that all such subpatterns are automatically treated as
2271 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2273 4. I had overlooked the fact that, in some locales, there are characters for
2274 which isalpha() is true but neither isupper() nor islower() are true. In
2275 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2276 and ordfeminine) are like this. This affected the treatment of \w and \W
2277 when they appeared in character classes, but not when they appeared outside
2278 a character class. The bit map for "word" characters is now created
2279 separately from the results of isalnum() instead of just taking it from the
2280 upper, lower, and digit maps. (Plus the underscore character, of course.)
2282 5. The above bug also affected the handling of POSIX character classes such as
2283 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2284 permanent tables. Instead, the bit maps for such a class were previously
2285 created as the appropriate unions of the upper, lower, and digit bitmaps.
2286 Now they are created by subtraction from the [[:word:]] class, which has
2287 its own bitmap.
2289 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2290 It is created by subtracting the vertical space characters (\x09, \x0a,
2291 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2292 subtraction was done in the overall bitmap for a character class, meaning
2293 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2294 be recognized. This bug has been fixed.
2296 7. Patches from the folks at Google:
2298 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2299 real life, but is still worth protecting against".
2301 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2302 regular expressions".
2304 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2305 have it.
2307 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2308 "configure" and the latter not, in order to fix a problem somebody had
2309 with compiling the Arg class on HP-UX.
2311 (e) Improve the error-handling of the C++ wrapper a little bit.
2313 (f) New tests for checking recursion limiting.
2315 8. The pcre_memmove() function, which is used only if the environment does not
2316 have a standard memmove() function (and is therefore rarely compiled),
2317 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2318 returning a result (though PCRE never actually uses the result).
2320 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2321 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2322 returned instead of calling malloc() with an overflowing number that would
2323 most likely cause subsequent chaos.
2325 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2327 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2328 with this option is matched, the nmatch and pmatch options of regexec() are
2329 ignored.
2331 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2332 provided in case anyone wants to the the POSIX interface with UTF-8
2333 strings.
2335 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2336 C++ linking (needed for some HP-UX environments).
2338 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2339 (unused parameter) and in the pcre_printint() function (omitted "default"
2340 switch label when the default is to do nothing).
2342 15. Added some code to make it possible, when PCRE is compiled as a C++
2343 library, to replace subject pointers for pcre_exec() with a smart pointer
2344 class, thus making it possible to process discontinuous strings.
2346 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2347 much the same function. They were added by different people who were trying
2348 to make PCRE easy to compile on non-Unix systems. It has been suggested
2349 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2350 for compiling on Windows systems. I have therefore replaced it with
2351 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2352 defaults to "extern" for C or "extern C" for C++, which works fine on
2353 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2354 SCOPE with something explicit in config.h. In addition:
2356 (a) pcreposix.h still had just "extern" instead of either of these macros;
2357 I have replaced it with PCRE_DATA_SCOPE.
2359 (b) Functions such as _pcre_xclass(), which are internal to the library,
2360 but external in the C sense, all had PCRE_EXPORT in their definitions.
2361 This is apparently wrong for the Windows case, so I have removed it.
2362 (It makes no difference on Unix-like systems.)
2364 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2365 of recursive calls to match(). This is different to MATCH_LIMIT because
2366 that limits the total number of calls to match(), not all of which increase
2367 the depth of recursion. Limiting the recursion depth limits the amount of
2368 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2369 when PCRE is compiled, and changed at run time. A patch from Google adds
2370 this functionality to the C++ interface.
2372 18. Changes to the handling of Unicode character properties:
2374 (a) Updated the table to Unicode 4.1.0.
2376 (b) Recognize characters that are not in the table as "Cn" (undefined).
2378 (c) I revised the way the table is implemented to a much improved format
2379 which includes recognition of ranges. It now supports the ranges that
2380 are defined in UnicodeData.txt, and it also amalgamates other
2381 characters into ranges. This has reduced the number of entries in the
2382 table from around 16,000 to around 3,000, thus reducing its size
2383 considerably. I realized I did not need to use a tree structure after
2384 all - a binary chop search is just as efficient. Having reduced the
2385 number of entries, I extended their size from 6 bytes to 8 bytes to
2386 allow for more data.
2388 (d) Added support for Unicode script names via properties such as \p{Han}.
2390 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2391 matching that character.
2393 20. When matching a repeated Unicode property with a minimum greater than zero,
2394 (for example \pL{2,}), PCRE could look past the end of the subject if it
2395 reached it while seeking the minimum number of characters. This could
2396 happen only if some of the characters were more than one byte long, because
2397 there is a check for at least the minimum number of bytes.
2399 21. Refactored the implementation of \p and \P so as to be more general, to
2400 allow for more different types of property in future. This has changed the
2401 compiled form incompatibly. Anybody with saved compiled patterns that use
2402 \p or \P will have to recompile them.
2404 22. Added "Any" and "L&" to the supported property types.
2406 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2407 but give a compile time error if the value is greater than 0xff.
2409 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2410 accidentally not being installed or uninstalled.
2412 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2413 made were to insert the current release number. This seemed silly, because
2414 it made things harder for people building PCRE on systems that don't run
2415 "configure". I have turned pcre.h into a distributed file, no longer built
2416 by "configure", with the version identification directly included. There is
2417 no longer a pcre.h.in file.
2419 However, this change necessitated a change to the pcre-config script as
2420 well. It is built from pcre-config.in, and one of the substitutions was the
2421 release number. I have updated configure.ac so that ./configure now finds
2422 the release number by grepping pcre.h.
2424 26. Added the ability to run the tests under valgrind.
2427 Version 6.4 05-Sep-05
2428 ---------------------
2430 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2431 "--" to be printed when multiple files were scanned, even when none of the
2432 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2433 consider it to be a bug, and have restored the previous behaviour.
2435 2. A couple of code tidies to get rid of compiler warnings.
2437 3. The pcretest program used to cheat by referring to symbols in the library
2438 whose names begin with _pcre_. These are internal symbols that are not
2439 really supposed to be visible externally, and in some environments it is
2440 possible to suppress them. The cheating is now confined to including
2441 certain files from the library's source, which is a bit cleaner.
2443 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2444 file's purpose clearer.
2446 5. Reorganized pcre_ucp_findchar().
2449 Version 6.3 15-Aug-05
2450 ---------------------
2452 1. The file libpcre.pc.in did not have general read permission in the tarball.
2454 2. There were some problems when building without C++ support:
2456 (a) If C++ support was not built, "make install" and "make test" still
2457 tried to test it.
2459 (b) There were problems when the value of CXX was explicitly set. Some
2460 changes have been made to try to fix these, and ...
2462 (c) --disable-cpp can now be used to explicitly disable C++ support.
2464 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2465 backslash in a target when C++ was disabled. This confuses some
2466 versions of "make", apparently. Using an intermediate variable solves
2467 this. (Same for CPP_LOBJ.)
2469 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2470 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2471 necessary on certain architectures.
2473 4. Added a setting of -export-symbols-regex to the link command to remove
2474 those symbols that are exported in the C sense, but actually are local
2475 within the library, and not documented. Their names all begin with
2476 "_pcre_". This is not a perfect job, because (a) we have to except some
2477 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2478 available (and never for static libraries). I have made a note to try to
2479 find a way round (a) in the future.
2482 Version 6.2 01-Aug-05
2483 ---------------------
2485 1. There was no test for integer overflow of quantifier values. A construction
2486 such as {1111111111111111} would give undefined results. What is worse, if
2487 a minimum quantifier for a parenthesized subpattern overflowed and became
2488 negative, the calculation of the memory size went wrong. This could have
2489 led to memory overwriting.
2491 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2493 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2494 operating environments where this matters.
2496 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2497 PCRE options from within the C++ wrapper.
2499 5. Named capturing subpatterns were not being correctly counted when a pattern
2500 was compiled. This caused two problems: (a) If there were more than 100
2501 such subpatterns, the calculation of the memory needed for the whole
2502 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2503 back references of the form \12, where the number was greater than 9, were
2504 not recognized as back references, even though there were sufficient
2505 previous subpatterns.
2507 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2508 versions of gcc, e.g. 2.95.4.
2511 Version 6.1 21-Jun-05
2512 ---------------------
2514 1. There was one reference to the variable "posix" in pcretest.c that was not
2515 surrounded by "#if !defined NOPOSIX".
2517 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2518 the cross-check on the old pcre_info() function, for the benefit of the
2519 cut-down version of PCRE that is currently imported into Exim.
2521 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2522 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2523 patterns that start (?i) but I don't think that matters. The use of (?i) is
2524 just an example; this all applies to the other options as well.
2526 4. Since libtool seems to echo the compile commands it is issuing, the output
2527 from "make" can be reduced a bit by putting "@" in front of each libtool
2528 compile command.
2530 5. Patch from the folks at Google for configure.in to be a bit more thorough
2531 in checking for a suitable C++ installation before trying to compile the
2532 C++ stuff. This should fix a reported problem when a compiler was present,
2533 but no suitable headers.
2535 6. The man pages all had just "PCRE" as their title. I have changed them to
2536 be the relevant file name. I have also arranged that these names are
2537 retained in the file doc/pcre.txt, which is a concatenation in text format
2538 of all the man pages except the little individual ones for each function.
2540 7. The NON-UNIX-USE file had not been updated for the different set of source
2541 files that come with release 6. I also added a few comments about the C++
2542 wrapper.
2545 Version 6.0 07-Jun-05
2546 ---------------------
2548 1. Some minor internal re-organization to help with my DFA experiments.
2550 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2551 didn't matter for the library itself when fully configured, but did matter
2552 when compiling without UCP support, or within Exim, where the ucp files are
2553 not imported.
2555 3. Refactoring of the library code to split up the various functions into
2556 different source modules. The addition of the new DFA matching code (see
2557 below) to a single monolithic source would have made it really too
2558 unwieldy, quite apart from causing all the code to be include in a
2559 statically linked application, when only some functions are used. This is
2560 relevant even without the DFA addition now that patterns can be compiled in
2561 one application and matched in another.
2563 The downside of splitting up is that there have to be some external
2564 functions and data tables that are used internally in different modules of
2565 the library but which are not part of the API. These have all had their
2566 names changed to start with "_pcre_" so that they are unlikely to clash
2567 with other external names.
2569 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2570 a different (DFA) algorithm. Although it is slower than the original
2571 function, it does have some advantages for certain types of matching
2572 problem.
2574 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2575 including restarting after a partial match.
2577 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2578 defined when compiling for Windows was sent to me. I have put it into the
2579 code, though I have no means of testing or verifying it.
2581 7. Added the pcre_refcount() auxiliary function.
2583 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2584 match before or at the first newline in the subject string. In pcretest,
2585 the /f option on a pattern can be used to set this.
2587 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2588 would behave wrongly. This has been present in PCRE since release 4.0.
2590 10. A number of changes to the pcregrep command:
2592 (a) Refactored how -x works; insert ^(...)$ instead of setting
2593 PCRE_ANCHORED and checking the length, in preparation for adding
2594 something similar for -w.
2596 (b) Added the -w (match as a word) option.
2598 (c) Refactored the way lines are read and buffered so as to have more
2599 than one at a time available.
2601 (d) Implemented a pcregrep test script.
2603 (e) Added the -M (multiline match) option. This allows patterns to match
2604 over several lines of the subject. The buffering ensures that at least
2605 8K, or the rest of the document (whichever is the shorter) is available
2606 for matching (and similarly the previous 8K for lookbehind assertions).
2608 (f) Changed the --help output so that it now says
2610 -w, --word-regex(p)
2612 instead of two lines, one with "regex" and the other with "regexp"
2613 because that confused at least one person since the short forms are the
2614 same. (This required a bit of code, as the output is generated
2615 automatically from a table. It wasn't just a text change.)
2617 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2618 option but starts with a hyphen. Could be a pattern or a path name
2619 starting with a hyphen, for instance.
2621 (h) "-" can be given as a file name to represent stdin.
2623 (i) When file names are being printed, "(standard input)" is used for
2624 the standard input, for compatibility with GNU grep. Previously
2625 "<stdin>" was used.
2627 (j) The option --label=xxx can be used to supply a name to be used for
2628 stdin when file names are being printed. There is no short form.
2630 (k) Re-factored the options decoding logic because we are going to add
2631 two more options that take data. Such options can now be given in four
2632 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2634 (l) Added the -A, -B, and -C options for requesting that lines of context
2635 around matches be printed.
2637 (m) Added the -L option to print the names of files that do not contain
2638 any matching lines, that is, the complement of -l.
2640 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2641 continue to scan other files.
2643 (o) The -s option was incorrectly implemented. For compatibility with other
2644 greps, it now suppresses the error message for a non-existent or non-
2645 accessible file (but not the return code). There is a new option called
2646 -q that suppresses the output of matching lines, which was what -s was
2647 previously doing.
2649 (p) Added --include and --exclude options to specify files for inclusion
2650 and exclusion when recursing.
2652 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2653 Hopefully, it now does.
2655 12. Missing cast in pcre_study().
2657 13. Added an "uninstall" target to the makefile.
2659 14. Replaced "extern" in the function prototypes in Makefile.in with
2660 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2661 world, but is set differently for Windows.
2663 15. Added a second compiling function called pcre_compile2(). The only
2664 difference is that it has an extra argument, which is a pointer to an
2665 integer error code. When there is a compile-time failure, this is set
2666 non-zero, in addition to the error test pointer being set to point to an
2667 error message. The new argument may be NULL if no error number is required
2668 (but then you may as well call pcre_compile(), which is now just a
2669 wrapper). This facility is provided because some applications need a
2670 numeric error indication, but it has also enabled me to tidy up the way
2671 compile-time errors are handled in the POSIX wrapper.
2673 16. Added VPATH=.libs to the makefile; this should help when building with one
2674 prefix path and installing with another. (Or so I'm told by someone who
2675 knows more about this stuff than I do.)
2677 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2678 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2679 match everything, including newlines. This is not POSIX-compatible, but
2680 somebody wanted the feature. From pcretest it can be activated by using
2681 both the P and the s flags.
2683 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2685 19. libpcre.pc was being incorrectly installed as executable.
2687 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2688 it now also looks for '\r' so that it will work unmodified on Windows.
2690 21. Added Google's contributed C++ wrapper to the distribution.
2692 22. Added some untidy missing memory free() calls in pcretest, to keep
2693 Electric Fence happy when testing.
2697 Version 5.0 13-Sep-04
2698 ---------------------
2700 1. Internal change: literal characters are no longer packed up into items
2701 containing multiple characters in a single byte-string. Each character
2702 is now matched using a separate opcode. However, there may be more than one
2703 byte in the character in UTF-8 mode.
2705 2. The pcre_callout_block structure has two new fields: pattern_position and
2706 next_item_length. These contain the offset in the pattern to the next match
2707 item, and its length, respectively.
2709 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2710 insertion of callouts before each pattern item. Added the /C option to
2711 pcretest to make use of this.
2713 4. On the advice of a Windows user, the lines
2715 #if defined(_WIN32) || defined(WIN32)
2716 _setmode( _fileno( stdout ), 0x8000 );
2717 #endif /* defined(_WIN32) || defined(WIN32) */
2719 have been added to the source of pcretest. This apparently does useful
2720 magic in relation to line terminators.
2722 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2723 for the benefit of those environments where the "b" makes a difference.
2725 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2726 to know about it. I have put a hack into configure.in that adds in code
2727 to set GCC=yes if CC=icc. This seems to end up at a point in the
2728 generated configure script that is early enough to affect the setting of
2729 compiler options, which is what is needed, but I have no means of testing
2730 whether it really works. (The user who reported this had patched the
2731 generated configure script, which of course I cannot do.)
2733 LATER: After change 22 below (new libtool files), the configure script
2734 seems to know about icc (and also ecc). Therefore, I have commented out
2735 this hack in configure.in.
2737 7. Added support for pkg-config (2 patches were sent in).
2739 8. Negated POSIX character classes that used a combination of internal tables
2740 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2741 [[:^ascii]]. Typically, they would match almost any characters. The other
2742 POSIX classes were not broken in this way.
2744 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2745 to find the match, as PCRE was deluded into thinking that the match had to
2746 start at the start point or following a newline. The same bug applied to
2747 patterns with negative forward assertions or any backward assertions
2748 preceding ".*" at the start, unless the pattern required a fixed first
2749 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2751 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2752 starting at the last subject character, bytes beyond the end of the subject
2753 string were read.
2755 11. Renamed the variable "class" as "classbits" to make life easier for C++
2756 users. (Previously there was a macro definition, but it apparently wasn't
2757 enough.)
2759 12. Added the new field "tables" to the extra data so that tables can be passed
2760 in at exec time, or the internal tables can be re-selected. This allows
2761 a compiled regex to be saved and re-used at a later time by a different
2762 program that might have everything at different addresses.
2764 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2765 -R library as well as a -L library.
2767 14. The debugging options of pcretest (-d on the command line or D on a
2768 pattern) showed incorrect output for anything following an extended class
2769 that contained multibyte characters and which was followed by a quantifier.
2771 15. Added optional support for general category Unicode character properties
2772 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2773 support. It adds about 90K to the size of the library. The meanings of the
2774 inbuilt class escapes such as \d and \s have NOT been changed.
2776 16. Updated pcredemo.c to include calls to free() to release the memory for the
2777 compiled pattern.
2779 17. The generated file chartables.c was being created in the source directory
2780 instead of in the building directory. This caused the build to fail if the
2781 source directory was different from the building directory, and was
2782 read-only.
2784 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2785 file. No doubt somebody will tell me if they don't make sense... Also added
2786 Dan Mooney's comments about building on OpenVMS.
2788 19. Added support for partial matching via the PCRE_PARTIAL option for
2789 pcre_exec() and the \P data escape in pcretest.
2791 20. Extended pcretest with 3 new pattern features:
2793 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2794 write the compiled pattern to the file whose name is "rest-of-line".
2795 This is a straight binary dump of the data, with the saved pointer to
2796 the character tables forced to be NULL. The study data, if any, is
2797 written too. After writing, pcretest reads a new pattern.
2799 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2800 compiled pattern from the given file. There must not be any
2801 occurrences of "<" in the file name (pretty unlikely); if there are,
2802 pcretest will instead treat the initial "<" as a pattern delimiter.
2803 After reading in the pattern, pcretest goes on to read data lines as
2804 usual.
2806 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2807 and 16-bit fields in a compiled pattern, to simulate a pattern that
2808 was compiled on a host of opposite endianness.
2810 21. The pcre-exec() function can now cope with patterns that were compiled on
2811 hosts of opposite endianness, with this restriction:
2813 As for any compiled expression that is saved and used later, the tables
2814 pointer field cannot be preserved; the extra_data field in the arguments
2815 to pcre_exec() should be used to pass in a tables address if a value
2816 other than the default internal tables were used at compile time.
2818 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2819 now diagnosed as an error. Previously, most of the time, a negative number
2820 would have been treated as zero, but if in addition "ovector" was passed as
2821 NULL, a crash could occur.
2823 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2824 new versions from the libtool 1.5 distribution (the last one is a copy of
2825 a file called libtool.m4). This seems to have fixed the need to patch
2826 "configure" to support Darwin 1.3 (which I used to do). However, I still
2827 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2828 workstation).
2830 24. Changed the PCRE licence to be the more standard "BSD" licence.
2833 Version 4.5 01-Dec-03
2834 ---------------------
2836 1. There has been some re-arrangement of the code for the match() function so
2837 that it can be compiled in a version that does not call itself recursively.
2838 Instead, it keeps those local variables that need separate instances for
2839 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2840 needs to "recurse". Keeping track of where control must go is done by means
2841 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2842 hide most of the details from the main code, and operates only if
2843 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2844 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2845 operating.
2847 To make it easier for callers to provide specially tailored get/free
2848 functions for this usage, two new functions, pcre_stack_malloc, and
2849 pcre_stack_free, are used. They are always called in strict stacking order,
2850 and the size of block requested is always the same.
2852 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2853 PCRE has been compiled to use the stack or the heap for recursion. The
2854 -C option of pcretest uses this to show which version is compiled.
2856 A new data escape \S, is added to pcretest; it causes the amounts of store
2857 obtained and freed by both kinds of malloc/free at match time to be added
2858 to the output.
2860 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2861 what's available on my current Linux desktop machine.
2863 3. When matching a UTF-8 string, the test for a valid string at the start has
2864 been extended. If start_offset is not zero, PCRE now checks that it points
2865 to a byte that is the start of a UTF-8 character. If not, it returns
2866 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2867 this is necessary because there may be backward assertions in the pattern.
2868 When matching the same subject several times, it may save resources to use
2869 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2871 4. The code for checking the validity of UTF-8 strings has been tightened so
2872 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2873 containing "overlong sequences".
2875 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2876 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2877 should have been "&", but it just so happened that all the cases this let
2878 through by mistake were picked up later in the function.
2880 6. I had used a variable called "isblank" - this is a C99 function, causing
2881 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2883 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2884 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2885 so the version is shown; (c) stop "make test" echoing "./RunTest".
2887 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2889 9. The prototype for memmove() for systems that don't have it was using
2890 size_t, but the inclusion of the header that defines size_t was later. I've
2891 moved the #includes for the C headers earlier to avoid this.
2893 10. Added some adjustments to the code to make it easier to compiler on certain
2894 special systems:
2896 (a) Some "const" qualifiers were missing.
2897 (b) Added the macro EXPORT before all exported functions; by default this
2898 is defined to be empty.
2899 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2900 that it reads its output file name as an argument instead of writing
2901 to the standard output and assuming this can be redirected.
2903 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2904 class containing characters with values greater than 255, PCRE compilation
2905 went into a loop.
2907 12. A recursive reference to a subpattern that was within another subpattern
2908 that had a minimum quantifier of zero caused PCRE to crash. For example,
2909 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2910 recursion. If the recursively-called subpattern itself had a zero repeat,
2911 that was OK.
2913 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2914 buffer into which it was copied (for escape processing) was still set at
2915 1024, so long lines caused crashes.
2917 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2918 "internal error: code overflow...". This applied to any character class
2919 that was followed by a possessive quantifier.
2921 15. Modified the Makefile to add libpcre.la as a prerequisite for
2922 libpcreposix.la because I was told this is needed for a parallel build to
2923 work.
2925 16. If a pattern that contained .* following optional items at the start was
2926 studied, the wrong optimizing data was generated, leading to matching
2927 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2928 matching string must start with a or b or c. The correct conclusion for
2929 this pattern is that a match can start with any character.
2932 Version 4.4 13-Aug-03
2933 ---------------------
2935 1. In UTF-8 mode, a character class containing characters with values between
2936 127 and 255 was not handled correctly if the compiled pattern was studied.
2937 In fixing this, I have also improved the studying algorithm for such
2938 classes (slightly).
2940 2. Three internal functions had redundant arguments passed to them. Removal
2941 might give a very teeny performance improvement.
2943 3. Documentation bug: the value of the capture_top field in a callout is *one
2944 more than* the number of the hightest numbered captured substring.
2946 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2947 in incorrectly linking with a previously installed version. They now link
2948 explicitly with libpcre.la.
2950 5. configure.in no longer needs to recognize Cygwin specially.
2952 6. A problem in pcre.in for Windows platforms is fixed.
2954 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2955 pcretest, it used to include the size of the study block as part of its
2956 output. Unfortunately, the structure contains a field that has a different
2957 size on different hardware architectures. This meant that the tests that
2958 showed this size failed. As the block is currently always of a fixed size,
2959 this information isn't actually particularly useful in pcretest output, so
2960 I have just removed it.
2962 8. Three pre-processor statements accidentally did not start in column 1.
2963 Sadly, there are *still* compilers around that complain, even though
2964 standard C has not required this for well over a decade. Sigh.
2966 9. In pcretest, the code for checking callouts passed small integers in the
2967 callout_data field, which is a void * field. However, some picky compilers
2968 complained about the casts involved for this on 64-bit systems. Now
2969 pcretest passes the address of the small integer instead, which should get
2970 rid of the warnings.
2972 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2973 both compile and run time, and gives an error if an invalid UTF-8 sequence
2974 is found. There is a option for disabling this check in cases where the
2975 string is known to be correct and/or the maximum performance is wanted.
2977 11. In response to a bug report, I changed one line in Makefile.in from
2979 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2980 to
2981 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2983 to look similar to other lines, but I have no way of telling whether this
2984 is the right thing to do, as I do not use Windows. No doubt I'll get told
2985 if it's wrong...
2988 Version 4.3 21-May-03
2989 ---------------------
2991 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2992 Makefile.
2994 2. Some refactoring to improve the quality of the code:
2996 (i) The utf8_table... variables are now declared "const".
2998 (ii) The code for \cx, which used the "case flipping" table to upper case
2999 lower case letters, now just substracts 32. This is ASCII-specific,
3000 but the whole concept of \cx is ASCII-specific, so it seems
3001 reasonable.
3003 (iii) PCRE was using its character types table to recognize decimal and
3004 hexadecimal digits in the pattern. This is silly, because it handles
3005 only 0-9, a-f, and A-F, but the character types table is locale-
3006 specific, which means strange things might happen. A private
3007 table is now used for this - though it costs 256 bytes, a table is
3008 much faster than multiple explicit tests. Of course, the standard
3009 character types table is still used for matching digits in subject
3010 strings against \d.
3012 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3013 ending in _t are). So I've renamed it as ESC_tee.
3015 3. The first argument for regexec() in the POSIX wrapper should have been
3016 defined as "const".
3018 4. Changed pcretest to use malloc() for its buffers so that they can be
3019 Electric Fenced for debugging.
3021 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3022 to read one or more bytes before the start of the subject string. Often this
3023 had no effect on PCRE's behaviour, but in some circumstances it could
3024 provoke a segmentation fault.
3026 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3027 to try to read one or more bytes before the start of the subject string.
3029 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3030 UTF-8 support could misbehave in various ways if the subject string
3031 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3032 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3033 back over UTF-8 characters.)
3036 Version 4.2 14-Apr-03
3037 ---------------------
3039 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3041 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3042 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3043 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3044 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3045 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3047 Note: automatic setting of the BUILD variables is not yet working
3048 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3049 compile-time but not at link-time
3050 [LINK]: use for linking executables only
3051 make different versions for Windows and non-Windows
3052 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3053 libraries
3054 [LINK_FOR_BUILD]: new variable
3055 [OBJEXT]: use throughout
3056 [EXEEXT]: use throughout
3057 <winshared>: new target
3058 <wininstall>: new target
3059 <dftables.o>: use native compiler
3060 <dftables>: use native linker
3061 <install>: handle Windows platform correctly
3062 <clean>: ditto
3063 <check>: ditto
3064 copy DLL to top builddir before testing
3066 As part of these changes, -no-undefined was removed again. This was reported
3067 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3068 in any case.
3070 3. Some tidies to get rid of compiler warnings:
3072 . In the match_data structure, match_limit was an unsigned long int, whereas
3073 match_call_count was an int. I've made them both unsigned long ints.
3075 . In pcretest the fact that a const uschar * doesn't automatically cast to
3076 a void * provoked a warning.
3078 . Turning on some more compiler warnings threw up some "shadow" variables
3079 and a few more missing casts.
3081 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3082 option, a class that contained a single character with a value between 128
3083 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3085 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3086 option, a class that contained several characters, but with at least one
3087 whose value was between 128 and 255 caused PCRE to crash.
3090 Version 4.1 12-Mar-03
3091 ---------------------
3093 1. Compiling with gcc -pedantic found a couple of places where casts were
3094 needed, and a string in dftables.c that was longer than standard compilers are
3095 required to support.
3097 2. Compiling with Sun's compiler found a few more places where the code could
3098 be tidied up in order to avoid warnings.
3100 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3101 first of these names is deprecated in the latest Autoconf in favour of the name
3102 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3103 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3104 analogy I have changed it to CFLAGS_FOR_BUILD.
3106 4. Added -no-undefined to the linking command in the Makefile, because this is
3107 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3108 linking step for the pcreposix library.
3110 5. PCRE was failing to diagnose the case of two named groups with the same
3111 name.
3113 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3114 literal character that is needed in the subject for a match, and scans along to
3115 ensure that it is present before embarking on the full matching process. This
3116 saves time in cases of nested unlimited repeats that are never going to match.
3117 Problem: the scan can take a lot of time if the subject is very long (e.g.
3118 megabytes), thus penalizing straightforward matches. It is now done only if the
3119 amount of subject to be scanned is less than 1000 bytes.
3121 7. A lesser problem with the same optimization is that it was recording the
3122 first character of an anchored pattern as "needed", thus provoking a search
3123 right along the subject, even when the first match of the pattern was going to
3124 fail. The "needed" character is now not set for anchored patterns, unless it
3125 follows something in the pattern that is of non-fixed length. Thus, it still
3126 fulfils its original purpose of finding quick non-matches in cases of nested
3127 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3130 Version 4.0 17-Feb-03
3131 ---------------------
3133 1. If a comment in an extended regex that started immediately after a meta-item
3134 extended to the end of string, PCRE compiled incorrect data. This could lead to
3135 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3137 2. Moved to autoconf 2.53 and libtool 1.4.2.
3139 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3140 the special perltest8 script is no longer needed - all the tests can be run
3141 from a single perltest script.
3143 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3144 by \s. It has now been removed in PCRE. This means it isn't recognized as
3145 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3146 class [:space:] *does* include VT, thereby creating a mess.
3148 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3149 space and tab.
3151 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3152 its new features into the main test script, reducing the number of scripts.
3154 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3155 were backward compatible, and made the (?i) apply to the whole pattern, as if
3156 /i were given. Now it behaves more logically, and applies the option setting
3157 only to what follows. PCRE has been changed to follow suit. However, if it
3158 finds options settings right at the start of the pattern, it extracts them into
3159 the global options, as before. Thus, they show up in the info data.
3161 8. Added support for the \Q...\E escape sequence. Characters in between are
3162 treated as literals. This is slightly different from Perl in that $ and @ are
3163 also handled as literals inside the quotes. In Perl, they will cause variable
3164 interpolation. Note the following examples:
3166 Pattern PCRE matches Perl matches
3168 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3169 \Qabc\$xyz\E abc\$xyz abc\$xyz
3170 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3172 For compatibility with Perl, \Q...\E sequences are recognized inside character
3173 classes as well as outside them.
3175 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3176 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3177 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3178 signed/unsigned warnings.
3180 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3181 option for pcretest, so I've replaced it by a simple function that does just
3182 that job.
3184 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3185 "pcregrep -".
3187 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3188 Java package. This provides some syntactic sugar for simple cases of what my
3189 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3190 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3191 item, you can use this simplified notation. Note that only makes sense with
3192 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3193 greediness, whatever the setting of the PCRE_UNGREEDY option.
3195 13. A change of greediness default within a pattern was not taking effect at
3196 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3197 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3198 was abstracted outside.
3200 14. PCRE now supports the \G assertion. It is true when the current matching
3201 position is at the start point of the match. This differs from \A when the
3202 starting offset is non-zero. Used with the /g option of pcretest (or similar
3203 code), it works in the same way as it does for Perl's /g option. If all
3204 alternatives of a regex begin with \G, the expression is anchored to the start
3205 match position, and the "anchored" flag is set in the compiled expression.
3207 15. Some bugs concerning the handling of certain option changes within patterns
3208 have been fixed. These applied to options other than (?ims). For example,
3209 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3210 been the other way round. Some of this was related to change 7 above.
3212 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3213 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3214 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3215 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3217 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3218 mode. Unlike ".", it always matches newline, whatever the setting of
3219 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3220 assertions. Perl allows it, but it doesn't (in general) work because it can't
3221 calculate the length of the lookbehind. At least, that's the case for Perl
3222 5.8.0 - I've been told they are going to document that it doesn't work in
3223 future.
3225 18. Added an error diagnosis for escapes that PCRE does not support: these are
3226 \L, \l, \N, \P, \p, \U, \u, and \X.
3228 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3229 reading past the end of the pattern in cases such as /[abcd/.
3231 20. PCRE was getting more memory than necessary for patterns with classes that
3232 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3234 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3235 compiling PCRE for use with Virtual Pascal.
3237 22. Small fix to the Makefile to make it work properly if the build is done
3238 outside the source tree.
3240 23. Added a new extension: a condition to go with recursion. If a conditional
3241 subpattern starts with (?(R) the "true" branch is used if recursion has
3242 happened, whereas the "false" branch is used only at the top level.
3244 24. When there was a very long string of literal characters (over 255 bytes
3245 without UTF support, over 250 bytes with UTF support), the computation of how
3246 much memory was required could be incorrect, leading to segfaults or other
3247 strange effects.
3249 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3250 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3251 there was a subsequent back reference to those brackets. This meant that, for
3252 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3253 possible to check for precisely this case. All we can do is abandon the
3254 optimization if .* occurs inside capturing brackets when there are any back
3255 references whatsoever. (See below for a better fix that came later.)
3257 26. The handling of the optimization for finding the first character of a
3258 non-anchored pattern, and for finding a character that is required later in the
3259 match were failing in some cases. This didn't break the matching; it just
3260 failed to optimize when it could. The way this is done has been re-implemented.
3262 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3264 28. Added a new feature that provides some of the functionality that Perl
3265 provides with (?{...}). The facility is termed a "callout". The way it is done
3266 in PCRE is for the caller to provide an optional function, by setting
3267 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3268 global variable. By default it is unset, which disables all calling out. To get
3269 the function called, the regex must include (?C) at appropriate points. This
3270 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3271 This provides a means of identifying different callout points. When PCRE
3272 reaches such a point in the regex, if pcre_callout has been set, the external
3273 function is called. It is provided with data in a structure called
3274 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3275 matching continues; if it returns a non-zero value, the match at the current
3276 point fails. However, backtracking will occur if possible. [This was changed
3277 later and other features added - see item 49 below.]
3279 29. pcretest is upgraded to test the callout functionality. It provides a
3280 callout function that displays information. By default, it shows the start of
3281 the match and the current position in the text. There are some new data escapes
3282 to vary what happens:
3284 \C+ in addition, show current contents of captured substrings
3285 \C- do not supply a callout function
3286 \C!n return 1 when callout number n is reached
3287 \C!n!m return 1 when callout number n is reached for the mth time
3289 30. If pcregrep was called with the -l option and just a single file name, it
3290 output "<stdin>" if a match was found, instead of the file name.
3292 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3293 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3294 pcre_exec(). This saves a malloc/free per call. The default value of
3295 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3296 when configuring.
3298 32. The default maximum size of a compiled pattern is 64K. There have been a
3299 few cases of people hitting this limit. The code now uses macros to handle the
3300 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3301 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3302 configuring. Tests 2 and 5 work only with 2-byte links because they output
3303 debugging information about compiled patterns.
3305 33. Internal code re-arrangements:
3307 (a) Moved the debugging function for printing out a compiled regex into
3308 its own source file (printint.c) and used #include to pull it into
3309 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3310 separate copies.
3312 (b) Defined the list of op-code names for debugging as a macro in
3313 internal.h so that it is next to the definition of the opcodes.
3315 (c) Defined a table of op-code lengths for simpler skipping along compiled
3316 code. This is again a macro in internal.h so that it is next to the
3317 definition of the opcodes.
3319 34. Added support for recursive calls to individual subpatterns, along the
3320 lines of Robin Houston's patch (but implemented somewhat differently).
3322 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3323 allow it to read and process whole directories in Win32. This code was
3324 contributed by Lionel Fourquaux; it has not been tested by me.
3326 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3327 used to name a group. Names consist of alphanumerics and underscores, and must
3328 be unique. Back references use the syntax (?P=name) and recursive calls use
3329 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3330 numbers. The function pcre_fullinfo() can be used after compilation to extract
3331 a name/number map. There are three relevant calls:
3333 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3334 PCRE_INFO_NAMECOUNT yields the number of entries
3335 PCRE_INFO_NAMETABLE yields a pointer to the map.
3337 The map is a vector of fixed-size entries. The size of each entry depends on
3338 the length of the longest name used. The first two bytes of each entry are the
3339 group number, most significant byte first. There follows the corresponding
3340 name, zero terminated. The names are in alphabetical order.
3342 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3343 case instead of 255. Making it the same both with and without UTF-8 support
3344 means that the same test output works with both.
3346 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3347 calling malloc() with a zero argument.
3349 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3350 optimization. I've improved things by keeping a bitmap of backreferences with
3351 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3352 fact referenced, the optimization can be applied. It is unlikely that a
3353 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3354 the match to follow \n) will appear inside brackets with a number greater than
3355 31, but if it does, any back reference > 31 suppresses the optimization.
3357 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3358 of disabling numbered capturing parentheses. Any opening parenthesis that is
3359 not followed by ? behaves as if it were followed by ?: but named parentheses
3360 can still be used for capturing (and they will acquire numbers in the usual
3361 way).
3363 41. Redesigned the return codes from the match() function into yes/no/error so
3364 that errors can be passed back from deep inside the nested calls. A malloc
3365 failure while inside a recursive subpattern call now causes the
3366 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3368 42. It is now possible to set a limit on the number of times the match()
3369 function is called in a call to pcre_exec(). This facility makes it possible to
3370 limit the amount of recursion and backtracking, though not in a directly
3371 obvious way, because the match() function is used in a number of different
3372 circumstances. The count starts from zero for each position in the subject
3373 string (for non-anchored patterns). The default limit is, for compatibility, a
3374 large number, namely 10 000 000. You can change this in two ways:
3376 (a) When configuring PCRE before making, you can use --with-match-limit=n
3377 to set a default value for the compiled library.
3379 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3380 a different value is set. See 45 below.
3382 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3384 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3385 of things that can be changed at compile time. The first argument specifies
3386 what is wanted and the second points to where the information is to be placed.
3387 The current list of available information is:
3391 The output is an integer that is set to one if UTF-8 support is available;
3392 otherwise it is set to zero.
3396 The output is an integer that it set to the value of the code that is used for
3397 newline. It is either LF (10) or CR (13).
3401 The output is an integer that contains the number of bytes used for internal
3402 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3406 The output is an integer that contains the threshold above which the POSIX
3407 interface uses malloc() for output vectors. See item 31 above.
3411 The output is an unsigned integer that contains the default limit of the number
3412 of match() calls in a pcre_exec() execution. See 42 above.
3414 44. pcretest has been upgraded by the addition of the -C option. This causes it
3415 to extract all the available output from the new pcre_config() function, and to
3416 output it. The program then exits immediately.
3418 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3419 order to support additional features. One way would have been to define
3420 pcre_exec2() (for example) with extra arguments, but this would not have been
3421 extensible, and would also have required all calls to the original function to
3422 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3423 is used for passing in "extra" data from pcre_study().
3425 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3426 contains the following fields:
3428 flags a bitmap indicating which of the following fields are set
3429 study_data opaque data from pcre_study()
3430 match_limit a way of specifying a limit on match() calls for a specific
3431 call to pcre_exec()
3432 callout_data data for callouts (see 49 below)
3434 The flag bits are also defined in pcre.h, and are
3440 The pcre_study() function now returns one of these new pcre_extra blocks, with
3441 the actual study data pointed to by the study_data field, and the
3442 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3443 before. That is, this change is entirely upwards-compatible and requires no
3444 change to existing code.
3446 If you want to pass in additional data to pcre_exec(), you can either place it
3447 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3448 block.
3450 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3451 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3452 times with different match limits, until it finds the minimum value needed for
3453 pcre_exec() to complete. The value is then output. This can be instructive; for
3454 most simple matches the number is quite small, but for pathological cases it
3455 gets very large very quickly.
3457 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3458 returns the size of the data block pointed to by the study_data field in a
3459 pcre_extra block, that is, the value that was passed as the argument to
3460 pcre_malloc() when PCRE was getting memory in which to place the information
3461 created by pcre_study(). The fourth argument should point to a size_t variable.
3462 pcretest has been extended so that this information is shown after a successful
3463 pcre_study() call when information about the compiled regex is being displayed.
3465 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3466 because what follows is always an absolute path. (Later: it turns out that this
3467 is more than cosmetic for MinGW, because it doesn't like empty path
3468 components.)
3470 49. Some changes have been made to the callout feature (see 28 above):
3472 (i) A callout function now has three choices for what it returns:
3474 0 => success, carry on matching
3475 > 0 => failure at this point, but backtrack if possible
3476 < 0 => serious error, return this value from pcre_exec()
3478 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3479 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3480 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3481 use by callout functions. It will never be used by PCRE itself.
3483 (ii) The pcre_extra structure (see 45 above) has a void * field called
3484 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3485 pcre_callout_block structure has a field of the same name. The contents of
3486 the field passed in the pcre_extra structure are passed to the callout
3487 function in the corresponding field in the callout block. This makes it
3488 easier to use the same callout-containing regex from multiple threads. For
3489 testing, the pcretest program has a new data escape
3491 \C*n pass the number n (may be negative) as callout_data
3493 If the callout function in pcretest receives a non-zero value as
3494 callout_data, it returns that value.
3496 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3497 there were some redundant $(CFLAGS) in commands that are now specified as
3498 $(LINK), which already includes $(CFLAGS).
3500 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3501 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3502 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3503 one-byte characters throughout. Note that case-insensitive matching applies
3504 only to characters whose values are less than 256. PCRE doesn't support the
3505 notion of cases for higher-valued characters.
3507 (i) A character class whose characters are all within 0-255 is handled as
3508 a bit map, and the map is inverted for negative classes. Previously, a
3509 character > 255 always failed to match such a class; however it should
3510 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3512 (ii) A negated character class with a single character < 255 is coded as
3513 "not this character" (OP_NOT). This wasn't working properly when the test
3514 character was multibyte, either singly or repeated.
3516 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3517 mode, for example: \x{100}{2,3}.
3519 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3520 singly or repeated) now correctly test multibyte characters. However,
3521 PCRE doesn't recognize any characters with values greater than 255 as
3522 digits, spaces, or word characters. Such characters always match \D, \S,
3523 and \W, and never match \d, \s, or \w.
3525 (v) Classes may now contain characters and character ranges with values
3526 greater than 255. For example: [ab\x{100}-\x{400}].
3528 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3529 PCRE in UTF-8 mode.
3531 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3532 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3533 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3534 value.)
3536 53. The single man page has become too large. I have therefore split it up into
3537 a number of separate man pages. These also give rise to individual HTML pages;
3538 these are now put in a separate directory, and there is an index.html page that
3539 lists them all. Some hyperlinking between the pages has been installed.
3541 54. Added convenience functions for handling named capturing parentheses.
3543 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3544 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3545 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3546 are faulted.
3548 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3549 calling configure. These values are used when compiling the dftables.c program
3550 which is run to generate the source of the default character tables. They
3551 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3552 you will need to set these values.
3554 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3557 Version 3.9 02-Jan-02
3558 ---------------------
3560 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3562 2. If --disable-static was given, the building process failed when trying to
3563 build pcretest and pcregrep. (For some reason it was using libtool to compile
3564 them, which is not right, as they aren't part of the library.)
3567 Version 3.8 18-Dec-01
3568 ---------------------
3570 1. The experimental UTF-8 code was completely screwed up. It was packing the
3571 bytes in the wrong order. How dumb can you get?
3574 Version 3.7 29-Oct-01
3575 ---------------------
3577 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3578 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3579 this didn't happen under Solaris 8, where I normally test things.
3581 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3582 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3583 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3584 this fix an hour or so after the initial 3.7 release.)
3587 Version 3.6 23-Oct-01
3588 ---------------------
3590 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3591 offsets passed as NULL with zero offset count.
3593 2. The config.guess and config.sub files had not been updated when I moved to
3594 the latest autoconf.
3597 Version 3.5 15-Aug-01
3598 ---------------------
3600 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3601 had been forgotten.
3603 2. By using declared but undefined structures, we can avoid using "void"
3604 definitions in pcre.h while keeping the internal definitions of the structures
3605 private.
3607 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3608 user point of view, this means that both static and shared libraries are built
3609 by default, but this can be individually controlled. More of the work of
3610 handling this static/shared cases is now inside libtool instead of PCRE's make
3611 file.
3613 4. The pcretest utility is now installed along with pcregrep because it is
3614 useful for users (to test regexs) and by doing this, it automatically gets
3615 relinked by libtool. The documentation has been turned into a man page, so
3616 there are now .1, .txt, and .html versions in /doc.
3618 5. Upgrades to pcregrep:
3619 (i) Added long-form option names like gnu grep.
3620 (ii) Added --help to list all options with an explanatory phrase.
3621 (iii) Added -r, --recursive to recurse into sub-directories.
3622 (iv) Added -f, --file to read patterns from a file.
3624 6. pcre_exec() was referring to its "code" argument before testing that
3625 argument for NULL (and giving an error if it was NULL).
3627 7. Upgraded Makefile.in to allow for compiling in a different directory from
3628 the source directory.
3630 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3631 options bits, the pointer it was passed was to an int instead of to an unsigned
3632 long int. This mattered only on 64-bit systems.
3634 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3635 generated) instead of pcre.in, which it its source. Also made the same change
3636 in several of the .c files.
3638 10. A new release of gcc defines printf() as a macro, which broke pcretest
3639 because it had an ifdef in the middle of a string argument for printf(). Fixed
3640 by using separate calls to printf().
3642 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3643 script, to force use of CR or LF instead of \n in the source. On non-Unix
3644 systems, the value can be set in config.h.
3646 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3647 absolute limit. Changed the text of the error message to make this clear, and
3648 likewise updated the man page.
3650 13. The limit of 99 on the number of capturing subpatterns has been removed.
3651 The new limit is 65535, which I hope will not be a "real" limit.
3654 Version 3.4 22-Aug-00
3655 ---------------------
3657 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3659 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3662 Version 3.3 01-Aug-00
3663 ---------------------
3665 1. If an octal character was given, but the value was greater than \377, it
3666 was not getting masked to the least significant bits, as documented. This could
3667 lead to crashes in some systems.
3669 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3670 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3672 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3673 These just pass their arguments on to (pcre_free)(), but they are provided
3674 because some uses of PCRE bind it to non-C systems that can call its functions,
3675 but cannot call free() or pcre_free() directly.
3677 4. Add "make test" as a synonym for "make check". Corrected some comments in
3678 the Makefile.
3680 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3681 Makefile.
3683 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3684 command called pgrep for grepping around the active processes.
3686 7. Added the beginnings of support for UTF-8 character strings.
3688 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3689 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3690 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3691 out for the ar command.)
3694 Version 3.2 12-May-00
3695 ---------------------
3697 This is purely a bug fixing release.
3699 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3700 of ZA. This was just one example of several cases that could provoke this bug,
3701 which was introduced by change 9 of version 2.00. The code for breaking
3702 infinite loops after an iteration that matches an empty string was't working
3703 correctly.
3705 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3706 when matched against abbab (for example). After matching an empty string, it
3707 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3708 caused it to match further down the string than it should.
3710 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3711 was there because it doesn't seem to be needed, and it causes trouble on some
3712 systems, as it is not a Standard C header. It has been removed.
3714 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3715 were reported on the Macintosh. The changes were from
3717 while ((c = *(++ptr)) != 0 && c != '\n');
3718 to
3719 while ((c = *(++ptr)) != 0 && c != '\n') ;
3721 Totally extraordinary, but if that's what it takes...
3723 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3724 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3725 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3726 assumes the way PCRE uses memmove() (always moving upwards).
3728 6. PCRE is being used in one environment where strchr() is not available. There
3729 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3730 faster code anyway.
3733 Version 3.1 09-Feb-00
3734 ---------------------
3736 The only change in this release is the fixing of some bugs in Makefile.in for
3737 the "install" target:
3739 (1) It was failing to install pcreposix.h.
3741 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3744 Version 3.0 01-Feb-00
3745 ---------------------
3747 1. Add support for the /+ modifier to perltest (to output $` like it does in
3748 pcretest).
3750 2. Add support for the /g modifier to perltest.
3752 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3753 matches null strings.
3755 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3756 pattern. Perl treats empty patterns specially - it reuses the most recent
3757 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3758 effect.
3760 5. The POSIX interface was broken in that it was just handing over the POSIX
3761 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3762 required a bigger vector, with some working space on the end. This means that
3763 the POSIX wrapper now has to get and free some memory, and copy the results.
3765 6. Added some simple autoconf support, placing the test data and the
3766 documentation in separate directories, re-organizing some of the
3767 information files, and making it build pcre-config (a GNU standard). Also added
3768 libtool support for building PCRE as a shared library, which is now the
3769 default.
3771 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3772 09 are not valid octal constants. Single digits will be used for minor values
3773 less than 10.
3775 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3776 existing programs that set these in the POSIX interface can use PCRE without
3777 modification.
3779 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3780 return all that pcre_info() returns, plus additional data. The pcre_info()
3781 function is retained for compatibility, but is considered to be obsolete.
3783 10. Added experimental recursion feature (?R) to handle one common case that
3784 Perl 5.6 will be able to do with (?p{...}).
3786 11. Added support for POSIX character classes like [:alpha:], which Perl is
3787 adopting.
3790 Version 2.08 31-Aug-99
3791 ----------------------
3793 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3794 trying to match at the startoffset position, but instead was moving forward to
3795 the next newline as if a previous match had failed.
3797 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3798 and could get into a loop if a null string was matched other than at the start
3799 of the subject.
3801 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3802 be distinguished at compile time, and for completeness also added PCRE_DATE.
3804 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3805 in GnuWin32 environments.
3808 Version 2.07 29-Jul-99
3809 ----------------------
3811 1. The documentation is now supplied in plain text form and HTML as well as in
3812 the form of man page sources.
3814 2. C++ compilers don't like assigning (void *) values to other pointer types.
3815 In particular this affects malloc(). Although there is no problem in Standard
3816 C, I've put in casts to keep C++ compilers happy.
3818 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3819 should be (const char *).
3821 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3822 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3823 However, I haven't made this a standard facility. The documentation doesn't
3824 mention it, and the Makefile doesn't support it.
3826 5. The Makefile now contains an "install" target, with editable destinations at
3827 the top of the file. The pcretest program is not installed.
3829 6. pgrep -V now gives the PCRE version number and date.
3831 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3832 causing the entire string to be ignored, instead of just the last character.
3834 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3835 non-matching string, it can take a very, very long time, even for strings of
3836 quite modest length, because of the nested recursion. PCRE now does better in
3837 some of these cases. It does this by remembering the last required literal
3838 character in the pattern, and pre-searching the subject to ensure it is present
3839 before running the real match. In other words, it applies a heuristic to detect
3840 some types of certain failure quickly, and in the above example, if presented
3841 with a string that has no trailing " it gives "no match" very quickly.
3843 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3844 other alternatives are tried instead.
3847 Version 2.06 09-Jun-99
3848 ----------------------
3850 1. Change pcretest's output for amount of store used to show just the code
3851 space, because the remainder (the data block) varies in size between 32-bit and
3852 64-bit systems.
3854 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3855 start matching at. This allows lookbehinds to work when searching for multiple
3856 occurrences in a string.
3858 3. Added additional options to pcretest for testing multiple occurrences:
3860 /+ outputs the rest of the string that follows a match
3861 /g loops for multiple occurrences, using the new startoffset argument
3862 /G loops for multiple occurrences by passing an incremented pointer
3864 4. PCRE wasn't doing the "first character" optimization for patterns starting
3865 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3866 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3867 the letter 'x'. On long subject strings, this gives a significant speed-up.
3870 Version 2.05 21-Apr-99
3871 ----------------------
3873 1. Changed the type of magic_number from int to long int so that it works
3874 properly on 16-bit systems.
3876 2. Fixed a bug which caused patterns starting with .* not to work correctly
3877 when the subject string contained newline characters. PCRE was assuming
3878 anchoring for such patterns in all cases, which is not correct because .* will
3879 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3880 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3881 must be retried after every newline in the subject.
3884 Version 2.04 18-Feb-99
3885 ----------------------
3887 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3888 computation of the store needed to hold the pattern was incorrect (too large).
3889 If such patterns were nested a few deep, this could multiply and become a real
3890 problem.
3892 2. Added /M option to pcretest to show the memory requirement of a specific
3893 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3895 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3896 compiled in such a way that the backtracking after subsequent failure was
3897 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3898 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3901 Version 2.03 02-Feb-99
3902 ----------------------
3904 1. Fixed typo and small mistake in man page.
3906 2. Added 4th condition (GPL supersedes if conflict) and created separate
3907 LICENCE file containing the conditions.
3909 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3910 Perl, that is the internal \ allows the delimiter to be included in the
3911 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3912 the final delimiter, add \ to the end of the pattern (to test the error).
3914 4. Added the convenience functions for extracting substrings after a successful
3915 match. Updated pcretest to make it able to test these functions.
3918 Version 2.02 14-Jan-99
3919 ----------------------
3921 1. Initialized the working variables associated with each extraction so that
3922 their saving and restoring doesn't refer to uninitialized store.
3924 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3925 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3926 fix the problem.
3928 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3929 calls, and wasn't printing the correct value for compiling calls. Increased the
3930 default value of LOOPREPEAT, and the number of significant figures in the
3931 times.
3933 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3935 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3936 a building problem on Windows NT with a FAT file system.
3939 Version 2.01 21-Oct-98
3940 ----------------------
3942 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3943 to character tables built by pcre_maketables() in the current locale. If NULL
3944 is passed, the default tables are used.
3947 Version 2.00 24-Sep-98
3948 ----------------------
3950 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3951 it any more.
3953 2. Allow quantification of (?>) groups, and make it work correctly.
3955 3. The first character computation wasn't working for (?>) groups.
3957 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3958 end of the subject) and add 5.005's \z, which really does match only at the
3959 very end of the subject.
3961 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3963 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3964 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3965 localized options. All options to pcre_study() were also removed.
3967 7. Add other new features from 5.005:
3969 $(?<= positive lookbehind
3970 $(?<! negative lookbehind
3971 (?imsx-imsx) added the unsetting capability
3972 such a setting is global if at outer level; local otherwise
3973 (?imsx-imsx:) non-capturing groups with option setting
3974 (?(cond)re|re) conditional pattern matching
3976 A backreference to itself in a repeated group matches the previous
3977 captured string.
3979 8. General tidying up of studying (both automatic and via "study")
3980 consequential on the addition of new assertions.
3982 9. As in 5.005, unlimited repeated groups that could match an empty substring
3983 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3984 runtime if any iteration does actually match an empty substring.
3986 10. Include the RunTest script in the distribution.
3988 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3989 discrepancies, some of which were old and were also with respect to 5.004. They
3990 have now been fixed.
3993 Version 1.09 28-Apr-98
3994 ----------------------
3996 1. A negated single character class followed by a quantifier with a minimum
3997 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3998 program crashes, or just wrong answers. This did not apply to negated classes
3999 containing more than one character, or to minima other than one.
4002 Version 1.08 27-Mar-98
4003 ----------------------
4005 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4007 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4008 latter must appear before anything that relies on it in the pattern.
4011 Version 1.07 16-Feb-98
4012 ----------------------
4014 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4015 repeat of a potentially empty string).
4018 Version 1.06 23-Jan-98
4019 ----------------------
4021 1. Added Markus Oberhumer's little patches for C++.
4023 2. Literal strings longer than 255 characters were broken.
4026 Version 1.05 23-Dec-97
4027 ----------------------
4029 1. Negated character classes containing more than one character were failing if
4030 PCRE_CASELESS was set at run time.
4033 Version 1.04 19-Dec-97
4034 ----------------------
4036 1. Corrected the man page, where some "const" qualifiers had been omitted.
4038 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4039 input syntax.
4041 3. Fixed memory leak which occurred when a regex with back references was
4042 matched with an offsets vector that wasn't big enough. The temporary memory
4043 that is used in this case wasn't being freed if the match failed.
4045 4. Tidied pcretest to ensure it frees memory that it gets.
4047 5. Temporary memory was being obtained in the case where the passed offsets
4048 vector was exactly big enough.
4050 6. Corrected definition of offsetof() from change 5 below.
4052 7. I had screwed up change 6 below and broken the rules for the use of
4053 setjmp(). Now fixed.
4056 Version 1.03 18-Dec-97
4057 ----------------------
4059 1. A erroneous regex with a missing opening parenthesis was correctly
4060 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4061 on some systems.
4063 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4064 it was reported that one broken compiler failed on the former because "code" is
4065 also an independent variable.
4067 3. The erroneous regex a[]b caused an array overrun reference.
4069 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4070 fail on data ending with that character. (It was going on too far, and checking
4071 the next character, typically a binary zero.) This was specific to the
4072 optimized code for single-character negative classes.
4074 5. Added a contributed patch from the TIN world which does the following:
4076 + Add an undef for memmove, in case the the system defines a macro for it.
4078 + Add a definition of offsetof(), in case there isn't one. (I don't know
4079 the reason behind this - offsetof() is part of the ANSI standard - but
4080 it does no harm).
4082 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4083 most of the places where whitespace preceded '#'. I have given up and
4084 allowed the remaining 2 cases to be at the margin.
4086 + Rename some variables in pcre to eliminate shadowing. This seems very
4087 pedantic, but does no harm, of course.
4089 6. Moved the call to setjmp() into its own function, to get rid of warnings
4090 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4092 7. Constructs such as \d{8,} were compiling into the equivalent of
4093 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4094 outcome, but in this particular case used more store than had been allocated,
4095 which caused the bug to be discovered because it threw up an internal error.
4097 8. The debugging code in both pcre and pcretest for outputting the compiled
4098 form of a regex was going wrong in the case of back references followed by
4099 curly-bracketed repeats.
4102 Version 1.02 12-Dec-97
4103 ----------------------
4105 1. Typos in pcre.3 and comments in the source fixed.
4107 2. Applied a contributed patch to get rid of places where it used to remove
4108 'const' from variables, and fixed some signed/unsigned and uninitialized
4109 variable warnings.
4111 3. Added the "runtest" target to Makefile.
4113 4. Set default compiler flag to -O2 rather than just -O.
4116 Version 1.01 19-Nov-97
4117 ----------------------
4119 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4120 like /([ab]*)*/, that is, for classes with more than one character in them.
4122 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4123 as /((?>a*))*/ (a PCRE_EXTRA facility).
4126 Version 1.00 18-Nov-97
4127 ----------------------
4129 1. Added compile-time macros to support systems such as SunOS4 which don't have
4130 memmove() or strerror() but have other things that can be used instead.
4132 2. Arranged that "make clean" removes the executables.
4135 Version 0.99 27-Oct-97
4136 ----------------------
4138 1. Fixed bug in code for optimizing classes with only one character. It was
4139 initializing a 32-byte map regardless, which could cause it to run off the end
4140 of the memory it had got.
4142 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4145 Version 0.98 22-Oct-97
4146 ----------------------
4148 1. Fixed bug in code for handling temporary memory usage when there are more
4149 back references than supplied space in the ovector. This could cause segfaults.
4152 Version 0.97 21-Oct-97
4153 ----------------------
4155 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4157 2. Optimized negated single characters not to use a bit map.
4159 3. Brought error texts together as macro definitions; clarified some of them;
4160 fixed one that was wrong - it said "range out of order" when it meant "invalid
4161 escape sequence".
4163 4. Changed some char * arguments to const char *.
4165 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4167 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4168 pcretest.
4171 Version 0.96 16-Oct-97
4172 ----------------------
4174 1. Added a simple "pgrep" utility to the distribution.
4176 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4177 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4178 where "ddd" means "one or more decimal digits".
4180 3. Fixed serious bug. If a pattern had a back reference, but the call to
4181 pcre_exec() didn't supply a large enough ovector to record the related
4182 identifying subpattern, the match always failed. PCRE now remembers the number
4183 of the largest back reference, and gets some temporary memory in which to save
4184 the offsets during matching if necessary, in order to ensure that
4185 backreferences always work.
4187 4. Increased the compatibility with Perl in a number of ways:
4189 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4190 to request this handling. The option can be set at compile or exec time.
4192 (b) $ matches before a terminating newline by default; an option
4193 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4194 mode). The option can be set at compile or exec time.
4196 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4197 the same as Perl's. If the decimal number it represents is less than 10
4198 or there aren't that many previous left capturing parentheses, an octal
4199 escape is read. Inside a character class, it's always an octal escape,
4200 even if it is a single digit.
4202 (d) An escaped but undefined alphabetic character is taken as a literal,
4203 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4204 escapes.
4206 (e) {0} is now permitted. (The previous item is removed from the compiled
4207 pattern).
4209 5. Changed all the names of code files so that the basic parts are no longer
4210 than 10 characters, and abolished the teeny "globals.c" file.
4212 6. Changed the handling of character classes; they are now done with a 32-byte
4213 bit map always.
4215 7. Added the -d and /D options to pcretest to make it possible to look at the
4216 internals of compilation without having to recompile pcre.
4219 Version 0.95 23-Sep-97
4220 ----------------------
4222 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4223 \x20 at the start of a run of normal characters. These were being treated as
4224 real characters, instead of the source characters being re-checked.
4227 Version 0.94 18-Sep-97
4228 ----------------------
4230 1. The functions are now thread-safe, with the caveat that the global variables
4231 containing pointers to malloc() and free() or alternative functions are the
4232 same for all threads.
4234 2. Get pcre_study() to generate a bitmap of initial characters for non-
4235 anchored patterns when this is possible, and use it if passed to pcre_exec().
4238 Version 0.93 15-Sep-97
4239 ----------------------
4241 1. /(b)|(:+)/ was computing an incorrect first character.
4243 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4244 but not actually doing anything yet.
4246 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4247 as Perl does (e.g. [-az] or [az-]).
4249 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4250 all possible positions.
4252 5. Split up into different modules to avoid including unneeded functions in a
4253 compiled binary. However, compile and exec are still in one module. The "study"
4254 function is split off.
4256 6. The character tables are now in a separate module whose source is generated
4257 by an auxiliary program - but can then be edited by hand if required. There are
4258 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4259 toupper() in the code.
4261 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4262 make them global. Abolish the function for setting them, as the caller can now
4263 set them directly.
4266 Version 0.92 11-Sep-97
4267 ----------------------
4269 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4270 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4272 2. Caseless matching was not working in character classes if the characters in
4273 the pattern were in upper case.
4275 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4277 4. Make PCRE_ANCHORED public and accept as a compile option.
4279 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4280 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4281 pass them.
4283 6. Give an error if bad option bits passed at compile or run time.
4285 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4286 pcretest to cause it to pass that flag.
4288 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4289 options, and the first character, if set.
4291 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4294 Version 0.91 10-Sep-97
4295 ----------------------
4297 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4298 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4300 2. PCRE was looping on encountering an indefinitely repeated back reference to
4301 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4302 Perl does - treats the match as successful.
4304 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5