ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 961 - (show annotations)
Fri Apr 20 11:49:13 2012 UTC (9 years, 2 months ago) by ph10
File size: 229560 byte(s)
Fix auto-possessify bugs for \s*\R and \S*R.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.31
5 -----------------------------
7 1. Fixing a wrong JIT test case and some compiler warnings.
9 2. Removed a bashism from the RunTest script.
11 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12 to unsigned type, result still unsigned" that was given by an MS compiler
13 on encountering the code "-sizeof(xxx)".
15 4. Partial matching support is added to the JIT compiler.
17 5. Fixed several bugs concerned with partial matching of items that consist
18 of more than one character:
20 (a) /^(..)\1/ did not partially match "aba" because checking references was
21 done on an "all or nothing" basis. This also applied to repeated
22 references.
24 (b) \R did not give a hard partial match if \r was found at the end of the
25 subject.
27 (c) \X did not give a hard partial match after matching one or more
28 characters at the end of the subject.
30 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31 a partial match for the string "\r".
33 (e) When newline was set to CRLF, the metacharacter "." did not recognize
34 a partial match for a CR character at the end of the subject string.
36 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
37 running pcretest, the text "(JIT)" added to the output whenever JIT is
38 actually used to run the match.
40 7. Individual JIT compile options can be set in pcretest by following -s+[+]
41 or /S+[+] with a digit between 1 and 7.
43 8. OP_NOT now supports any UTF character not just single-byte ones.
45 9. (*MARK) control verb is now supported by the JIT compiler.
47 10. The command "./RunTest list" lists the available tests without actually
48 running any of them. (Because I keep forgetting what they all are.)
52 12. Applied a (slightly modified) user-supplied patch that improves performance
53 when the heap is used for recursion (compiled with --disable-stack-for-
54 recursion). Instead of malloc and free for each heap frame each time a
55 logical recursion happens, frames are retained on a chain and re-used where
56 possible. This sometimes gives as much as 30% improvement.
58 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59 call.
61 14. As documented, (*COMMIT) is now confined to within a positive assertion.
63 15. It is now possible to link pcretest with libedit as an alternative to
64 libreadline.
66 16. (*COMMIT) control verb is now supported by the JIT compiler.
68 17. The Unicode data tables have been updated to Unicode 6.1.0.
70 18. Added --file-list option to pcregrep.
72 19. Added binary file support to pcregrep, including the -a, --binary-files,
73 -I, and --text options.
75 20. The madvise function is renamed for posix_madvise for QNX compatibility
76 reasons. Fixed by Giuseppe D'Angelo.
78 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
80 22. Changed the option for creating symbolic links for 16-bit man pages from
81 -s to -sf so that re-installing does not cause issues.
83 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
85 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
86 matches in certain environments (the workspace was not being correctly
87 retained). Also added to pcre_dfa_exec() a simple plausibility check on
88 some of the workspace data at the beginning of a restart.
90 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
91 was not doing so when it should - probably a typo introduced by SVN 528
92 (change 8.10/14).
95 Version 8.30 04-February-2012
96 -----------------------------
98 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
99 name is defined in ctype.h.
101 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
102 only in quite long subpatterns.
104 3. Removed the function pcre_info(), which has been obsolete and deprecated
105 since it was replaced by pcre_fullinfo() in February 2000.
107 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
108 match a (*MARK), and the match failed at the start of the subject, a
109 reference to memory before the start of the subject could occur. This bug
110 was introduced by fix 17 of release 8.21.
112 5. A reference to an unset group with zero minimum repetition was giving
113 totally wrong answers (in non-JavaScript-compatibility mode). For example,
114 /(another)?(\1?)test/ matched against "hello world test". This bug was
115 introduced in release 8.13.
117 6. Add support for 16-bit character strings (a large amount of work involving
118 many changes and refactorings).
120 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
121 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
122 from a file.
124 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
125 rounding is not applied in this particular case).
127 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
128 if they appear, or are escaped, in patterns.
130 10. Get rid of a number of -Wunused-but-set-variable warnings.
132 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
133 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
134 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
135 also returns the mark "x". This bug applied to capturing parentheses,
136 non-capturing parentheses, and atomic parentheses. It also applied to some
137 assertions.
139 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
140 information out of configure.ac instead of relying on pcre.h.generic, which
141 is not stored in the repository.
143 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
144 -lreadline.
146 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
148 15. Applied Graycode's patch to put the top-level frame on the stack rather
149 than the heap when not using the stack for recursion. This gives a
150 performance improvement in many cases when recursion is not deep.
152 16. Experimental code added to "pcretest -C" to output the stack frame size.
155 Version 8.21 12-Dec-2011
156 ------------------------
158 1. Updating the JIT compiler.
160 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
161 are added as well.
163 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
164 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
165 calling _pcre_jit_exec. Some extra comments are added.
167 4. (*MARK) settings inside atomic groups that do not contain any capturing
168 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
169 was introduced by change 18 for 8.20.
171 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
172 ECMA-262 standard.
174 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
175 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
176 This bug was probably introduced by change 9 of 8.13.
178 7. While fixing 6 above, I noticed that a number of other items were being
179 incorrectly rejected as "not fixed length". This arose partly because newer
180 opcodes had not been added to the fixed-length checking code. I have (a)
181 corrected the bug and added tests for these items, and (b) arranged for an
182 error to occur if an unknown opcode is encountered while checking for fixed
183 length instead of just assuming "not fixed length". The items that were
184 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
185 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
186 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
188 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
189 being incorrectly compiled and would have given unpredicatble results.
191 9. A possessively repeated subpattern with minimum repeat count greater than
192 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
193 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
194 the first (A) could occur when it should not.
196 10. Add a cast and remove a redundant test from the code.
198 11. JIT should use pcre_malloc/pcre_free for allocation.
200 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
201 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
202 is anything other than /usr, -L is still shown).
204 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
206 14. Perl does not support \N without a following name in a [] class; PCRE now
207 also gives an error.
209 15. If a forward reference was repeated with an upper limit of around 2000,
210 it caused the error "internal error: overran compiling workspace". The
211 maximum number of forward references (including repeats) was limited by the
212 internal workspace, and dependent on the LINK_SIZE. The code has been
213 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
214 the default depends on LINK_SIZE. There is a new upper limit (for safety)
215 of around 200,000 forward references. While doing this, I also speeded up
216 the filling in of repeated forward references.
218 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
219 incorrectly expecting the subject to contain another "a" after the start.
221 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
222 in the match, the SKIP should be ignored. This was not happening; instead
223 the SKIP was being treated as NOMATCH. For patterns such as
224 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
225 tested.
227 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
228 now much more compatible with Perl, in particular in cases where the result
229 is a non-match for a non-anchored pattern. For example, if
230 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
231 "m", where previously it did not return a name. A side effect of this
232 change is that for partial matches, the last encountered mark name is
233 returned, as for non matches. A number of tests that were previously not
234 Perl-compatible have been moved into the Perl-compatible test files. The
235 refactoring has had the pleasing side effect of removing one argument from
236 the match() function, thus reducing its stack requirements.
238 19. If the /S+ option was used in pcretest to study a pattern using JIT,
239 subsequent uses of /S (without +) incorrectly behaved like /S+.
241 21. Retrieve executable code size support for the JIT compiler and fixing
242 some warnings.
244 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
245 not work when the shorter character appeared right at the end of the
246 subject string.
248 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
249 systems.
251 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
252 output it when the /M option is used in pcretest.
254 25. The CheckMan script was not being included in the distribution. Also, added
255 an explicit "perl" to run Perl scripts from the PrepareRelease script
256 because this is reportedly needed in Windows.
258 26. If study data was being save in a file and studying had not found a set of
259 "starts with" bytes for the pattern, the data written to the file (though
260 never used) was taken from uninitialized memory and so caused valgrind to
261 complain.
263 27. Updated RunTest.bat as provided by Sheri Pierce.
265 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
267 29. Computation of memory usage for the table of capturing group names was
268 giving an unnecessarily large value.
271 Version 8.20 21-Oct-2011
272 ------------------------
274 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
275 a POSIX class. After further experiments with Perl, which convinced me that
276 Perl has bugs and confusions, a closing square bracket is no longer allowed
277 in a POSIX name. This bug also affected patterns with classes that started
278 with full stops.
280 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
281 captured substring, but while checking the failing first alternative,
282 substring 1 is temporarily captured. If the output vector supplied to
283 pcre_exec() was not big enough for this capture, the yield of the function
284 was still zero ("insufficient space for captured substrings"). This cannot
285 be totally fixed without adding another stack variable, which seems a lot
286 of expense for a edge case. However, I have improved the situation in cases
287 such as /(a)(b)x|abc/ matched against "abc", where the return code
288 indicates that fewer than the maximum number of slots in the ovector have
289 been set.
291 3. Related to (2) above: when there are more back references in a pattern than
292 slots in the output vector, pcre_exec() uses temporary memory during
293 matching, and copies in the captures as far as possible afterwards. It was
294 using the entire output vector, but this conflicts with the specification
295 that only 2/3 is used for passing back captured substrings. Now it uses
296 only the first 2/3, for compatibility. This is, of course, another edge
297 case.
299 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
300 main code base, and can be used by building with --enable-jit. When this is
301 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
302 runtime --no-jit option is given.
304 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
305 ovector, the return from the function was zero, implying that there were
306 other matches that did not fit. The correct "exactly full" value is now
307 returned.
309 6. If a subpattern that was called recursively or as a subroutine contained
310 (*PRUNE) or any other control that caused it to give a non-standard return,
311 invalid errors such as "Error -26 (nested recursion at the same subject
312 position)" or even infinite loops could occur.
314 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
315 computing the minimum length on reaching *ACCEPT, and so ended up with the
316 wrong value of 1 rather than 0. Further investigation indicates that
317 computing a minimum subject length in the presence of *ACCEPT is difficult
318 (think back references, subroutine calls), and so I have changed the code
319 so that no minimum is registered for a pattern that contains *ACCEPT.
321 8. If (*THEN) was present in the first (true) branch of a conditional group,
322 it was not handled as intended. [But see 16 below.]
324 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
325 Sheri Pierce.
327 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
328 the first byte in a match must be "a".
330 11. Change 17 for 8.13 increased the recursion depth for patterns like
331 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
332 pattern contains any instances of (*THEN). If it does not, the old
333 optimizations are restored. It would be nice to do this on a per-group
334 basis, but at the moment that is not feasible.
336 12. In some environments, the output of pcretest -C is CRLF terminated. This
337 broke RunTest's code that checks for the link size. A single white space
338 character after the value is now allowed for.
340 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
341 For "fr", it uses the Windows-specific input and output files.
343 14. If (*THEN) appeared in a group that was called recursively or as a
344 subroutine, it did not work as intended. [But see next item.]
346 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
347 pattern fragments (but not containing any | characters). If A and B are
348 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
349 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
350 D. In other words, Perl considers parentheses that do not contain any |
351 characters to be part of a surrounding alternative, whereas PCRE was
352 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
353 differently. PCRE now behaves in the same way as Perl, except in the case
354 of subroutine/recursion calls such as (?1) which have in any case always
355 been different (but PCRE had them first :-).
357 16. Related to 15 above: Perl does not treat the | in a conditional group as
358 creating alternatives. Such a group is treated in the same way as an
359 ordinary group without any | characters when processing (*THEN). PCRE has
360 been changed to match Perl's behaviour.
362 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
363 RunGrepTest script failed.
365 18. Change 22 for version 13 caused atomic groups to use more stack. This is
366 inevitable for groups that contain captures, but it can lead to a lot of
367 stack use in large patterns. The old behaviour has been restored for atomic
368 groups that do not contain any capturing parentheses.
370 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
371 suppress the check for a minimum subject length at run time. (If it was
372 given to pcre_exec() or pcre_dfa_exec() it did work.)
374 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
375 fail to work when decoding hex characters in data strings in EBCDIC
376 environments.
378 21. It appears that in at least one Mac OS environment, the isxdigit() function
379 is implemented as a macro that evaluates to its argument more than once,
380 contravening the C 90 Standard (I haven't checked a later standard). There
381 was an instance in pcretest which caused it to go wrong when processing
382 \x{...} escapes in subject strings. The has been rewritten to avoid using
383 things like p++ in the argument of isxdigit().
386 Version 8.13 16-Aug-2011
387 ------------------------
389 1. The Unicode data tables have been updated to Unicode 6.0.0.
391 2. Two minor typos in pcre_internal.h have been fixed.
393 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
394 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
395 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
397 4. There were a number of related bugs in the code for matching backrefences
398 caselessly in UTF-8 mode when codes for the characters concerned were
399 different numbers of bytes. For example, U+023A and U+2C65 are an upper
400 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
401 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
402 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
403 2-byte code at the end of the subject (it thought there wasn't enough data
404 left).
406 5. Comprehensive information about what went wrong is now returned by
407 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
408 as the output vector has at least 2 elements. The offset of the start of
409 the failing character and a reason code are placed in the vector.
411 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
412 now returned is for the first byte of the failing character, instead of the
413 last byte inspected. This is an incompatible change, but I hope it is small
414 enough not to be a problem. It makes the returned offset consistent with
415 pcre_exec() and pcre_dfa_exec().
417 7. pcretest now gives a text phrase as well as the error number when
418 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
419 failure, the offset and reason code are output.
421 8. When \R was used with a maximizing quantifier it failed to skip backwards
422 over a \r\n pair if the subsequent match failed. Instead, it just skipped
423 back over a single character (\n). This seems wrong (because it treated the
424 two characters as a single entity when going forwards), conflicts with the
425 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
426 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
427 has been changed.
429 9. Some internal refactoring has changed the processing so that the handling
430 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
431 time (the PCRE_DOTALL option was changed this way some time ago: version
432 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
433 which was always a bit of a fudge. It also means that there is one less
434 argument for the match() function, which reduces its stack requirements
435 slightly. This change also fixes an incompatibility with Perl: the pattern
436 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
438 10. More internal refactoring has drastically reduced the number of recursive
439 calls to match() for possessively repeated groups such as (abc)++ when
440 using pcre_exec().
442 11. While implementing 10, a number of bugs in the handling of groups were
443 discovered and fixed:
445 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
446 (a|)*(?1) gave a compile-time internal error.
447 ((a|)+)+ did not notice that the outer group could match an empty string.
448 (^a|^)+ was not marked as anchored.
449 (.*a|.*)+ was not marked as matching at start or after a newline.
451 12. Yet more internal refactoring has removed another argument from the match()
452 function. Special calls to this function are now indicated by setting a
453 value in a variable in the "match data" data block.
455 13. Be more explicit in pcre_study() instead of relying on "default" for
456 opcodes that mean there is no starting character; this means that when new
457 ones are added and accidentally left out of pcre_study(), testing should
458 pick them up.
460 14. The -s option of pcretest has been documented for ages as being an old
461 synonym of -m (show memory usage). I have changed it to mean "force study
462 for every regex", that is, assume /S for every regex. This is similar to -i
463 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
464 using it. It makes it easier to run collections of tests with and without
465 study enabled, and thereby test pcre_study() more easily. All the standard
466 tests are now run with and without -s (but some patterns can be marked as
467 "never study" - see 20 below).
469 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
470 restoration of the capturing data to the outer values was not happening
471 correctly.
473 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
474 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
475 pattern had matched an empty string, and so incorrectly returned a no
476 match.
478 17. There was optimizing code for the last branch of non-capturing parentheses,
479 and also for the obeyed branch of a conditional subexpression, which used
480 tail recursion to cut down on stack usage. Unfortunately, now that there is
481 the possibility of (*THEN) occurring in these branches, tail recursion is
482 no longer possible because the return has to be checked for (*THEN). These
483 two optimizations have therefore been removed. [But see 8.20/11 above.]
485 18. If a pattern containing \R was studied, it was assumed that \R always
486 matched two bytes, thus causing the minimum subject length to be
487 incorrectly computed because \R can also match just one byte.
489 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
490 was incorrectly computed.
492 20. If /S is present twice on a test pattern in pcretest input, it now
493 *disables* studying, thereby overriding the use of -s on the command line
494 (see 14 above). This is necessary for one or two tests to keep the output
495 identical in both cases.
497 21. When (*ACCEPT) was used in an assertion that matched an empty string and
498 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
500 22. When an atomic group that contained a capturing parenthesis was
501 successfully matched, but the branch in which it appeared failed, the
502 capturing was not being forgotten if a higher numbered group was later
503 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
504 group 1 to "a", when in fact it should be unset. This applied to multi-
505 branched capturing and non-capturing groups, repeated or not, and also to
506 positive assertions (capturing in negative assertions does not happen
507 in PCRE) and also to nested atomic groups.
509 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
510 subject after a captured substring, to make it easier to tell which of a
511 number of identical substrings has been captured.
513 24. The way atomic groups are processed by pcre_exec() has been changed so that
514 if they are repeated, backtracking one repetition now resets captured
515 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
516 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
517 "aaa". Previously, it would have been "a". As part of this code
518 refactoring, the way recursive calls are handled has also been changed.
520 25. If an assertion condition captured any substrings, they were not passed
521 back unless some other capturing happened later. For example, if
522 (?(?=(a))a) was matched against "a", no capturing was returned.
524 26. When studying a pattern that contained subroutine calls or assertions,
525 the code for finding the minimum length of a possible match was handling
526 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
527 group 1 called group 2 while simultaneously a separate group 2 called group
528 1). A stack overflow occurred in this case. I have fixed this by limiting
529 the recursion depth to 10.
531 27. Updated RunTest.bat in the distribution to the version supplied by Tom
532 Fortmann. This supports explicit test numbers on the command line, and has
533 argument validation and error reporting.
535 28. An instance of \X with an unlimited repeat could fail if at any point the
536 first character it looked at was a mark character.
538 29. Some minor code refactoring concerning Unicode properties and scripts
539 should reduce the stack requirement of match() slightly.
541 30. Added the '=' option to pcretest to check the setting of unused capturing
542 slots at the end of the pattern, which are documented as being -1, but are
543 not included in the return count.
545 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
546 compiled something random. Now it gives a compile-time error (as does
547 Perl).
549 32. A *MARK encountered during the processing of a positive assertion is now
550 recorded and passed back (compatible with Perl).
552 33. If --only-matching or --colour was set on a pcregrep call whose pattern
553 had alternative anchored branches, the search for a second match in a line
554 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
555 matched the line "0102" twice. The same bug affected patterns that started
556 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
557 twice.
559 34. Previously, PCRE did not allow quantification of assertions. However, Perl
560 does, and because of capturing effects, quantifying parenthesized
561 assertions may at times be useful. Quantifiers are now allowed for
562 parenthesized assertions.
564 35. A minor code tidy in pcre_compile() when checking options for \R usage.
566 36. \g was being checked for fancy things in a character class, when it should
567 just be a literal "g".
569 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
570 appearance of a nested POSIX class supersedes an apparent external class.
571 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
572 unescaped square brackets may also appear as part of class names. For
573 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
574 more like Perl. (But see 8.20/1 above.)
576 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
577 was because it thought it was \N{name}, which is not supported).
579 39. Add minix to OS list not supporting the -S option in pcretest.
581 40. PCRE tries to detect cases of infinite recursion at compile time, but it
582 cannot analyze patterns in sufficient detail to catch mutual recursions
583 such as ((?1))((?2)). There is now a runtime test that gives an error if a
584 subgroup is called recursively as a subpattern for a second time at the
585 same position in the subject string. In previous releases this might have
586 been caught by the recursion limit, or it might have run out of stack.
588 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
589 happen only once. PCRE was, however incorrectly giving a compile time error
590 "recursive call could loop indefinitely" because it cannot analyze the
591 pattern in sufficient detail. The compile time test no longer happens when
592 PCRE is compiling a conditional subpattern, but actual runaway loops are
593 now caught at runtime (see 40 above).
595 42. It seems that Perl allows any characters other than a closing parenthesis
596 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
597 has been changed to be the same.
599 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
600 as not to get warnings when autogen.sh is called. Also changed
601 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
603 44. To help people who use pcregrep to scan files containing exceedingly long
604 lines, the following changes have been made:
606 (a) The default value of the buffer size parameter has been increased from
607 8K to 20K. (The actual buffer used is three times this size.)
609 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
610 PCRE is built.
612 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
613 to be set at run time.
615 (d) Numerical values in pcregrep options can be followed by K or M, for
616 example --buffer-size=50K.
618 (e) If a line being scanned overflows pcregrep's buffer, an error is now
619 given and the return code is set to 2.
621 45. Add a pointer to the latest mark to the callout data block.
623 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
624 partial match of an empty string instead of no match. This was specific to
625 the use of ".".
627 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
628 complete match instead of a partial match. This bug was dependent on both
629 the PCRE_UTF8 and PCRE_DOTALL options being set.
631 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
632 starting byte set, because \b was not being ignored.
635 Version 8.12 15-Jan-2011
636 ------------------------
638 1. Fixed some typos in the markup of the man pages, and wrote a script that
639 checks for such things as part of the documentation building process.
641 2. On a big-endian 64-bit system, pcregrep did not correctly process the
642 --match-limit and --recursion-limit options (added for 8.11). In
643 particular, this made one of the standard tests fail. (The integer value
644 went into the wrong half of a long int.)
646 3. If the --colour option was given to pcregrep with -v (invert match), it
647 did strange things, either producing crazy output, or crashing. It should,
648 of course, ignore a request for colour when reporting lines that do not
649 match.
651 4. Another pcregrep bug caused similar problems if --colour was specified with
652 -M (multiline) and the pattern match finished with a line ending.
654 5. In pcregrep, when a pattern that ended with a literal newline sequence was
655 matched in multiline mode, the following line was shown as part of the
656 match. This seems wrong, so I have changed it.
658 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
659 the check for further matches in the same line (so they could be coloured)
660 to overrun the end of the current line. If another match was found, it was
661 incorrectly shown (and then shown again when found in the next line).
663 7. If pcregrep was compiled under Windows, there was a reference to the
664 function pcregrep_exit() before it was defined. I am assuming this was
665 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
666 reported by a user. I've moved the definition above the reference.
669 Version 8.11 10-Dec-2010
670 ------------------------
672 1. (*THEN) was not working properly if there were untried alternatives prior
673 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
674 backtracked to try for "b" instead of moving to the next alternative branch
675 at the same level (in this case, to look for "c"). The Perl documentation
676 is clear that when (*THEN) is backtracked onto, it goes to the "next
677 alternative in the innermost enclosing group".
679 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
680 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
681 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
682 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
683 (*THEN).
685 3. If \s appeared in a character class, it removed the VT character from
686 the class, even if it had been included by some previous item, for example
687 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
688 of \s, but is part of the POSIX "space" class.)
690 4. A partial match never returns an empty string (because you can always
691 match an empty string at the end of the subject); however the checking for
692 an empty string was starting at the "start of match" point. This has been
693 changed to the "earliest inspected character" point, because the returned
694 data for a partial match starts at this character. This means that, for
695 example, /(?<=abc)def/ gives a partial match for the subject "abc"
696 (previously it gave "no match").
698 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
699 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
700 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
701 has an implication that the given string is incomplete (because a partial
702 match is preferred over a full match). For this reason, these items now
703 give a partial match in this situation. [Aside: previously, the one case
704 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
705 match rather than a full match, which was wrong by the old rules, but is
706 now correct.]
708 6. There was a bug in the handling of #-introduced comments, recognized when
709 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
710 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
711 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
712 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
713 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
714 places in pcre_compile().
716 7. Related to (6) above, when pcre_compile() was skipping #-introduced
717 comments when looking ahead for named forward references to subpatterns,
718 the only newline sequence it recognized was NL. It now handles newlines
719 according to the set newline convention.
721 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
722 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
723 cater for a lack of strerror(). These oversights have been fixed.
725 9. Added --match-limit and --recursion-limit to pcregrep.
727 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
729 11. When the -o option was used, pcregrep was setting a return code of 1, even
730 when matches were found, and --line-buffered was not being honoured.
732 12. Added an optional parentheses number to the -o and --only-matching options
733 of pcregrep.
735 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
736 can match an empty string. The code to do it in pcretest and pcredemo
737 needed fixing:
739 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
740 only one byte after an empty string match just before CRLF (this case
741 just got forgotten; "any" and "anycrlf" were OK).
743 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
744 mode when an empty string match preceded an ASCII character followed by
745 a non-ASCII character. (The code for advancing by one character rather
746 than one byte was nonsense.)
748 (c) The pcredemo.c sample program did not have any code at all to handle
749 the cases when CRLF is a valid newline sequence.
751 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
752 as a starting offset was within the subject string. There is now a new
753 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
754 negative or greater than the length of the string. In order to test this,
755 pcretest is extended to allow the setting of negative starting offsets.
757 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
758 starting offset points to the beginning of a UTF-8 character was
759 unnecessarily clumsy. I tidied it up.
761 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
762 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
764 17. Nobody had reported that the --include_dir option, which was added in
765 release 7.7 should have been called --include-dir (hyphen, not underscore)
766 for compatibility with GNU grep. I have changed it to --include-dir, but
767 left --include_dir as an undocumented synonym, and the same for
768 --exclude-dir, though that is not available in GNU grep, at least as of
769 release 2.5.4.
771 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
772 characters from a string of bytes) have been redefined so as not to use
773 loops, in order to improve performance in some environments. At the same
774 time, I abstracted some of the common code into auxiliary macros to save
775 repetition (this should not affect the compiled code).
777 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
778 compile-time error is now given if \c is not followed by an ASCII
779 character, that is, a byte less than 128. (In EBCDIC mode, the code is
780 different, and any byte value is allowed.)
782 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
783 START_OPTIMIZE option, which is now allowed at compile time - but just
784 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
785 to pcregrep and other applications that have no direct access to PCRE
786 options. The new /Y option in pcretest sets this option when calling
787 pcre_compile().
789 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
790 back references. Groups containing recursive back references were forced to
791 be atomic by that change, but in the case of named groups, the amount of
792 memory required was incorrectly computed, leading to "Failed: internal
793 error: code overflow". This has been fixed.
795 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
796 pcretest.c, to avoid build problems in some Borland environments.
799 Version 8.10 25-Jun-2010
800 ------------------------
802 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
803 THEN.
805 2. (*ACCEPT) was not working when inside an atomic group.
807 3. Inside a character class, \B is treated as a literal by default, but
808 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
809 causes the error). The code is unchanged, but I tidied the documentation.
811 4. Inside a character class, PCRE always treated \R and \X as literals,
812 whereas Perl faults them if its -w option is set. I have changed PCRE so
813 that it faults them when PCRE_EXTRA is set.
815 5. Added support for \N, which always matches any character other than
816 newline. (It is the same as "." when PCRE_DOTALL is not set.)
818 6. When compiling pcregrep with newer versions of gcc which may have
819 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
820 declared with attribute warn_unused_result" were given. Just casting the
821 result to (void) does not stop the warnings; a more elaborate fudge is
822 needed. I've used a macro to implement this.
824 7. Minor change to pcretest.c to avoid a compiler warning.
826 8. Added four artifical Unicode properties to help with an option to make
827 \s etc use properties (see next item). The new properties are: Xan
828 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
830 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
831 use Unicode properties. (*UCP) at the start of a pattern can be used to set
832 this option. Modified pcretest to add /W to test this facility. Added
833 REG_UCP to make it available via the POSIX interface.
835 10. Added --line-buffered to pcregrep.
837 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
838 studied, and the match started with a letter with a code point greater than
839 127 whose first byte was different to the first byte of the other case of
840 the letter, the other case of this starting letter was not recognized
841 (#976).
843 12. If a pattern that was studied started with a repeated Unicode property
844 test, for example, \p{Nd}+, there was the theoretical possibility of
845 setting up an incorrect bitmap of starting bytes, but fortunately it could
846 not have actually happened in practice until change 8 above was made (it
847 added property types that matched character-matching opcodes).
849 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
850 possible starting bytes for non-anchored patterns.
852 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
853 \R, and also a number of cases that involve Unicode properties, both
854 explicit and implicit when PCRE_UCP is set.
856 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
857 input, it could crash or give wrong results if characters with values
858 greater than 0xc0 were present in the subject string. (Detail: it assumed
859 UTF-8 input when processing these items.)
861 16. Added a lot of (int) casts to avoid compiler warnings in systems where
862 size_t is 64-bit (#991).
864 17. Added a check for running out of memory when PCRE is compiled with
865 --disable-stack-for-recursion (#990).
867 18. If the last data line in a file for pcretest does not have a newline on
868 the end, a newline was missing in the output.
870 19. The default pcre_chartables.c file recognizes only ASCII characters (values
871 less than 128) in its various bitmaps. However, there is a facility for
872 generating tables according to the current locale when PCRE is compiled. It
873 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
874 characters, are recognized by isspace() and therefore were getting set in
875 these tables, and indeed these tables seem to approximate to ISO 8859. This
876 caused a problem in UTF-8 mode when pcre_study() was used to create a list
877 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
878 which of course cannot start UTF-8 characters. I have changed the code so
879 that only real ASCII characters (less than 128) and the correct starting
880 bytes for UTF-8 encodings are set for characters greater than 127 when in
881 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
882 altogether.)
884 20. Added the /T option to pcretest so as to be able to run tests with non-
885 standard character tables, thus making it possible to include the tests
886 used for 19 above in the standard set of tests.
888 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
889 reference to a subpattern the other side of a comment that contains an
890 opening parenthesis caused either an internal compiling error, or a
891 reference to the wrong subpattern.
894 Version 8.02 19-Mar-2010
895 ------------------------
897 1. The Unicode data tables have been updated to Unicode 5.2.0.
899 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
900 configured.
902 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
903 original author of that file, following a query about its status.
905 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
906 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
908 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
909 quantifier applied to a forward-referencing subroutine call, could compile
910 incorrect code or give the error "internal error: previously-checked
911 referenced subpattern not found".
913 6. Both MS Visual Studio and Symbian OS have problems with initializing
914 variables to point to external functions. For these systems, therefore,
915 pcre_malloc etc. are now initialized to local functions that call the
916 relevant global functions.
918 7. There were two entries missing in the vectors called coptable and poptable
919 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
920 I've fixed the data, and added a kludgy way of testing at compile time that
921 the lengths are correct (equal to the number of opcodes).
923 8. Following on from 7, I added a similar kludge to check the length of the
924 eint vector in pcreposix.c.
926 9. Error texts for pcre_compile() are held as one long string to avoid too
927 much relocation at load time. To find a text, the string is searched,
928 counting zeros. There was no check for running off the end of the string,
929 which could happen if a new error number was added without updating the
930 string.
932 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
934 11. \K was not working if it appeared in an atomic group or in a group that
935 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
936 \K is "not well defined" if used in an assertion. PCRE now accepts it if
937 the assertion is positive, but not if it is negative.
939 12. Change 11 fortuitously reduced the size of the stack frame used in the
940 "match()" function of pcre_exec.c by one pointer. Forthcoming
941 implementation of support for (*MARK) will need an extra pointer on the
942 stack; I have reserved it now, so that the stack frame size does not
943 decrease.
945 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
946 item in branch that calls a recursion is a subroutine call - as in the
947 second branch in the above example - was incorrectly given the compile-
948 time error "recursive call could loop indefinitely" because pcre_compile()
949 was not correctly checking the subroutine for matching a non-empty string.
951 14. The checks for overrunning compiling workspace could trigger after an
952 overrun had occurred. This is a "should never occur" error, but it can be
953 triggered by pathological patterns such as hundreds of nested parentheses.
954 The checks now trigger 100 bytes before the end of the workspace.
956 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
959 Version 8.01 19-Jan-2010
960 ------------------------
962 1. If a pattern contained a conditional subpattern with only one branch (in
963 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
964 computed the wrong minimum data length (which is of course zero for such
965 subpatterns). This could cause incorrect "no match" results.
967 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
968 the pattern is reset in the first branch, pcre_compile() failed with
969 "internal error: code overflow at offset...". This happened only when
970 the reset was to the original external option setting. (An optimization
971 abstracts leading options settings into an external setting, which was the
972 cause of this.)
974 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
975 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
976 assertion pattern did not match (meaning that the assertion was true), it
977 was incorrectly treated as false if the SKIP had been reached during the
978 matching. This also applied to assertions used as conditions.
980 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
981 assertion subpattern, including such a pattern used as a condition,
982 unpredictable results occurred, instead of the error return
985 5. The C++ GlobalReplace function was not working like Perl for the special
986 situation when an empty string is matched. It now does the fancy magic
987 stuff that is necessary.
989 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
990 removed. (These were left over from very, very early versions of PCRE.)
992 7. Some cosmetic changes to the code to make life easier when compiling it
993 as part of something else:
995 (a) Change DEBUG to PCRE_DEBUG.
997 (b) In pcre_compile(), rename the member of the "branch_chain" structure
998 called "current" as "current_branch", to prevent a collision with the
999 Linux macro when compiled as a kernel module.
1001 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1002 prevent a collision with the Linux macro when compiled as a kernel
1003 module.
1005 8. In pcre_compile() there are some checks for integer overflows that used to
1006 cast potentially large values to (double). This has been changed to that
1007 when building, a check for int64_t is made, and if it is found, it is used
1008 instead, thus avoiding the use of floating point arithmetic. (There is no
1009 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1010 double.
1012 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1013 2005 (difference between two addresses compared to an unsigned value).
1015 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1016 custom one, because of the following reported problem in Windows:
1018 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1019 under Win32.
1020 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1021 therefore missing the function definition.
1022 - The compiler thus generates a "C" signature for the test function.
1023 - The linker fails to find the "C" function.
1024 - PCRE fails to configure if asked to do so against libbz2.
1026 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1027 messages were output:
1029 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1030 rerunning libtoolize, to keep the correct libtool macros in-tree.
1031 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1033 I have done both of these things.
1035 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1036 most of the time, it *can* run out if it is given a pattern that contains a
1037 runaway infinite recursion. I updated the discussion in the pcrestack man
1038 page.
1040 13. Now that we have gone to the x.xx style of version numbers, the minor
1041 version may start with zero. Using 08 or 09 is a bad idea because users
1042 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1043 interpreted as invalid octal numbers. I've updated the previous comment in
1044 configure.ac, and also added a check that gives an error if 08 or 09 are
1045 used.
1047 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1048 causing partial matching to fail when the end of the subject matched \W
1049 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1051 15. There were some discrepancies between the declarations in pcre_internal.h
1052 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1053 their definitions. The declarations used "const uschar *" and the
1054 definitions used USPTR. Even though USPTR is normally defined as "const
1055 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1056 reported that: "This difference in casting confuses some C++ compilers, for
1057 example, SunCC recognizes above declarations as different functions and
1058 generates broken code for hbpcre." I have changed the declarations to use
1059 USPTR.
1061 16. GNU libtool is named differently on some systems. The autogen.sh script now
1062 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1063 (FreeBSD).
1065 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1066 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1067 comment: "Figure out how to create a longlong from a string: strtoll and
1068 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1069 instance, but it only takes 2 args instead of 3!"
1071 18. A subtle bug concerned with back references has been fixed by a change of
1072 specification, with a corresponding code fix. A pattern such as
1073 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1074 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1075 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1076 same bug. Such groups have to be quantified to be useful, or contained
1077 inside another quantified group. (If there's no repetition, the reference
1078 can never match.) The problem arises because, having left the group and
1079 moved on to the rest of the pattern, a later failure that backtracks into
1080 the group uses the captured value from the final iteration of the group
1081 rather than the correct earlier one. I have fixed this in PCRE by forcing
1082 any group that contains a reference to itself to be an atomic group; that
1083 is, there cannot be any backtracking into it once it has completed. This is
1084 similar to recursive and subroutine calls.
1087 Version 8.00 19-Oct-09
1088 ----------------------
1090 1. The table for translating pcre_compile() error codes into POSIX error codes
1091 was out-of-date, and there was no check on the pcre_compile() error code
1092 being within the table. This could lead to an OK return being given in
1093 error.
1095 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1096 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1097 in a Windows environment.
1099 3. The pcregrep --count option prints the count for each file even when it is
1100 zero, as does GNU grep. However, pcregrep was also printing all files when
1101 --files-with-matches was added. Now, when both options are given, it prints
1102 counts only for those files that have at least one match. (GNU grep just
1103 prints the file name in this circumstance, but including the count seems
1104 more useful - otherwise, why use --count?) Also ensured that the
1105 combination -clh just lists non-zero counts, with no names.
1107 4. The long form of the pcregrep -F option was incorrectly implemented as
1108 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1109 but it seems right to fix it, and I didn't think it was worth preserving
1110 the old behaviour.
1112 5. The command line items --regex=pattern and --regexp=pattern were not
1113 recognized by pcregrep, which required --regex pattern or --regexp pattern
1114 (with a space rather than an '='). The man page documented the '=' forms,
1115 which are compatible with GNU grep; these now work.
1117 6. No libpcreposix.pc file was created for pkg-config; there was just
1118 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1120 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1121 when UCP support is not needed, by modifying the Python script that
1122 generates it from Unicode data files. This should not matter if the module
1123 is correctly used as a library, but I received one complaint about 50K of
1124 unwanted data. My guess is that the person linked everything into his
1125 program rather than using a library. Anyway, it does no harm.
1127 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1128 was a minimum greater than 1 for a wide character in a possessive
1129 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1130 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1131 character. Chaos in the form of incorrect output or a compiling loop could
1132 result.
1134 9. The restrictions on what a pattern can contain when partial matching is
1135 requested for pcre_exec() have been removed. All patterns can now be
1136 partially matched by this function. In addition, if there are at least two
1137 slots in the offset vector, the offset of the earliest inspected character
1138 for the match and the offset of the end of the subject are set in them when
1139 PCRE_ERROR_PARTIAL is returned.
1141 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1142 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1143 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1144 and may be more useful for multi-segment matching.
1146 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1147 used to be given if ever the end of the subject was reached; now it is
1148 given only if matching could not proceed because another character was
1149 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1150 string "Z", which now yields "no match" instead of "partial match". In the
1151 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1152 final character ended with (*FAIL).
1154 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1155 if the pattern had a "must contain" character that was already found in the
1156 earlier partial match, unless partial matching was again requested. For
1157 example, with the pattern /dog.(body)?/, the "must contain" character is
1158 "g". If the first part-match was for the string "dog", restarting with
1159 "sbody" failed. This bug has been fixed.
1161 13. The string returned by pcre_dfa_exec() after a partial match has been
1162 changed so that it starts at the first inspected character rather than the
1163 first character of the match. This makes a difference only if the pattern
1164 starts with a lookbehind assertion or \b or \B (\K is not supported by
1165 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1166 matching functions compatible, and I think it's the right thing to do.
1168 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1169 so that the demonstration program is easily available in environments where
1170 PCRE has not been installed from source.
1172 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1173 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1174 library.
1176 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1177 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1178 is not the first non-POSIX option to be added. Clearly some people find
1179 these options useful.
1181 17. If a caller to the POSIX matching function regexec() passes a non-zero
1182 value for nmatch with a NULL value for pmatch, the value of
1183 nmatch is forced to zero.
1185 18. RunGrepTest did not have a test for the availability of the -u option of
1186 the diff command, as RunTest does. It now checks in the same way as
1187 RunTest, and also checks for the -b option.
1189 19. If an odd number of negated classes containing just a single character
1190 interposed, within parentheses, between a forward reference to a named
1191 subpattern and the definition of the subpattern, compilation crashed with
1192 an internal error, complaining that it could not find the referenced
1193 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1194 [The bug was that it was starting one character too far in when skipping
1195 over the character class, thus treating the ] as data rather than
1196 terminating the class. This meant it could skip too much.]
1198 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1199 /g option in pcretest when the pattern contains \K, which makes it possible
1200 to have an empty string match not at the start, even when the pattern is
1201 anchored. Updated pcretest and pcredemo to use this option.
1203 21. If the maximum number of capturing subpatterns in a recursion was greater
1204 than the maximum at the outer level, the higher number was returned, but
1205 with unset values at the outer level. The correct (outer level) value is
1206 now given.
1208 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1209 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1210 make it do so. The string so far is captured, making this feature
1211 compatible with Perl.
1213 23. The tests have been re-organized, adding tests 11 and 12, to make it
1214 possible to check the Perl 5.10 features against Perl 5.10.
1216 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1217 pattern matches a fixed length string. PCRE did not allow this; now it
1218 does. Neither allows recursion.
1220 25. I finally figured out how to implement a request to provide the minimum
1221 length of subject string that was needed in order to match a given pattern.
1222 (It was back references and recursion that I had previously got hung up
1223 on.) This code has now been added to pcre_study(); it finds a lower bound
1224 to the length of subject needed. It is not necessarily the greatest lower
1225 bound, but using it to avoid searching strings that are too short does give
1226 some useful speed-ups. The value is available to calling programs via
1227 pcre_fullinfo().
1229 26. While implementing 25, I discovered to my embarrassment that pcretest had
1230 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1231 study optimizations had never been tested with that matching function.
1232 Oops. What is worse, even when it was passed study data, there was a bug in
1233 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1234 were also very few tests of studied patterns with pcre_dfa_exec().
1236 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1237 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1238 on the other side of the coin, they are no longer allowed to have different
1239 names, because these cannot be distinguished in PCRE, and this has caused
1240 confusion. (This is a difference from Perl.)
1242 28. When duplicate subpattern names are present (necessarily with different
1243 numbers, as required by 27 above), and a test is made by name in a
1244 conditional pattern, either for a subpattern having been matched, or for
1245 recursion in such a pattern, all the associated numbered subpatterns are
1246 tested, and the overall condition is true if the condition is true for any
1247 one of them. This is the way Perl works, and is also more like the way
1248 testing by number works.
1251 Version 7.9 11-Apr-09
1252 ---------------------
1254 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1255 (pcretest), all targets were linked against these libraries. This included
1256 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1257 libraries. This caused unwanted dependencies to be created. This problem
1258 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1259 pcretest is linked with readline.
1261 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1262 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1263 moved outside it again, because FALSE and TRUE are already defined in AIX,
1264 but BOOL is not.
1266 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1267 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1269 4. The pcregrep documentation said spaces were inserted as well as colons (or
1270 hyphens) following file names and line numbers when outputting matching
1271 lines. This is not true; no spaces are inserted. I have also clarified the
1272 wording for the --colour (or --color) option.
1274 5. In pcregrep, when --colour was used with -o, the list of matching strings
1275 was not coloured; this is different to GNU grep, so I have changed it to be
1276 the same.
1278 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1279 each matching line was coloured. Now it goes on to look for further matches
1280 of any of the test patterns, which is the same behaviour as GNU grep.
1282 7. A pattern that could match an empty string could cause pcregrep to loop; it
1283 doesn't make sense to accept an empty string match in pcregrep, so I have
1284 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1285 seems to be how GNU grep behaves.
1287 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1288 start or after a newline", because the conditional assertion was not being
1289 correctly handled. The rule now is that both the assertion and what follows
1290 in the first alternative must satisfy the test.
1292 9. If auto-callout was enabled in a pattern with a conditional group whose
1293 condition was an assertion, PCRE could crash during matching, both with
1294 pcre_exec() and pcre_dfa_exec().
1296 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1297 used for matching.
1299 11. Unicode property support in character classes was not working for
1300 characters (bytes) greater than 127 when not in UTF-8 mode.
1302 12. Added the -M command line option to pcretest.
1304 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1306 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1308 16. Added comments and documentation about mis-use of no_arg in the C++
1309 wrapper.
1311 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1312 from Martin Jerabek that uses macro names for all relevant character and
1313 string constants.
1315 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1316 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1317 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1318 these, but not everybody uses configure.
1320 19. A conditional group that had only one branch was not being correctly
1321 recognized as an item that could match an empty string. This meant that an
1322 enclosing group might also not be so recognized, causing infinite looping
1323 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1324 with the subject "ab", where knowledge that the repeated group can match
1325 nothing is needed in order to break the loop.
1327 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1328 exec(), but without supplying a callout function, matching went wrong.
1330 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1331 leak if the size of the offset vector was greater than 30. When the vector
1332 is smaller, the saved offsets during recursion go onto a local stack
1333 vector, but for larger vectors malloc() is used. It was failing to free
1334 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1335 error, in fact).
1337 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1338 heapframe that is used only when UTF-8 support is enabled. This caused no
1339 problem, but was untidy.
1341 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1342 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1343 included within another project.
1345 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1346 slightly modified by me:
1348 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1349 not building pcregrep.
1351 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1352 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1354 25. Forward references, both numeric and by name, in patterns that made use of
1355 duplicate group numbers, could behave incorrectly or give incorrect errors,
1356 because when scanning forward to find the reference group, PCRE was not
1357 taking into account the duplicate group numbers. A pattern such as
1358 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1360 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1361 the feature of a custom pointer more persuasive (as requested by a user).
1363 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1364 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1365 pre-defined.
1367 28. Added support for (*UTF8) at the start of a pattern.
1369 29. Arrange for flags added by the "release type" setting in CMake to be shown
1370 in the configuration summary.
1373 Version 7.8 05-Sep-08
1374 ---------------------
1376 1. Replaced UCP searching code with optimized version as implemented for Ad
1377 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1378 stage table and inline lookup instead of a function, giving speed ups of 2
1379 to 5 times on some simple patterns that I tested. Permission was given to
1380 distribute the MultiStage2.py script that generates the tables (it's not in
1381 the tarball, but is in the Subversion repository).
1383 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1384 scripts.
1386 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1387 a group with a zero qualifier. The result of the study could be incorrect,
1388 or the function might crash, depending on the pattern.
1390 4. Caseless matching was not working for non-ASCII characters in back
1391 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1392 It now works when Unicode Property Support is available.
1394 5. In pcretest, an escape such as \x{de} in the data was always generating
1395 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1396 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1397 truncation.
1399 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1401 7. Added two (int) casts to pcregrep when printing the difference of two
1402 pointers, in case they are 64-bit values.
1404 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1405 test 2 if it fails.
1407 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1408 and a #define of that name to empty if it is not externally set. This is to
1409 allow users of MSVC to set it if necessary.
1411 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1412 the convenience functions in the pcre_get.c source file.
1414 11. An option change at the start of a pattern that had top-level alternatives
1415 could cause overwriting and/or a crash. This command provoked a crash in
1416 some environments:
1418 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1420 This potential security problem was recorded as CVE-2008-2371.
1422 12. For a pattern where the match had to start at the beginning or immediately
1423 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1424 pcre_dfa_exec() could read past the end of the passed subject if there was
1425 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1426 pcretest so that it places the subject at the end of its malloc-ed buffer.
1428 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1429 exec() might read past the end of the data buffer in UTF-8 mode.
1431 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1432 the data contained the byte 0x85 as part of a UTF-8 character within its
1433 first line. This applied both to normal and DFA matching.
1435 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1436 /^[^d]*?$/8 failed to match "abc".
1438 16. Added a missing copyright notice to pcrecpp_internal.h.
1440 17. Make it more clear in the documentation that values returned from
1441 pcre_exec() in ovector are byte offsets, not character counts.
1443 18. Tidied a few places to stop certain compilers from issuing warnings.
1445 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1446 supplied by Stefan Weber. I made a further small update for 7.8 because
1447 there is a change of source arrangements: the pcre_searchfuncs.c module is
1448 replaced by pcre_ucd.c.
1451 Version 7.7 07-May-08
1452 ---------------------
1454 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1455 a string to a long long, pretend we don't even have a long long." This is
1456 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1458 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1459 pre-7.6 versions, which defined a global no_arg variable instead of putting
1460 it in the RE class. (See also #8 below.)
1462 3. Remove a line of dead code, identified by coverity and reported by Nuno
1463 Lopes.
1465 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1467 (1) The include/exclude patterns were being applied to the whole pathnames
1468 of files, instead of just to the final components.
1470 (2) If there was more than one level of directory, the subdirectories were
1471 skipped unless they satisfied the include/exclude conditions. This is
1472 inconsistent with GNU grep (and could even be seen as contrary to the
1473 pcregrep specification - which I improved to make it absolutely clear).
1474 The action now is always to scan all levels of directory, and just
1475 apply the include/exclude patterns to regular files.
1477 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1478 --exclude_dir in the tests to avoid scanning .svn directories.
1480 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1481 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1482 doesn't support NULs in patterns.
1484 7. Added some missing "const"s to declarations of static tables in
1485 pcre_compile.c and pcre_dfa_exec.c.
1487 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1488 caused by fix #2 above. (Subsequently also a second patch to fix the
1489 first patch. And a third patch - this was a messy problem.)
1491 9. Applied Craig's patch to remove the use of push_back().
1493 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1494 matching function regexec().
1496 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1497 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1498 references. PCRE supports relative numbers with this syntax (I don't think
1499 Oniguruma does).
1501 12. Previously, a group with a zero repeat such as (...){0} was completely
1502 omitted from the compiled regex. However, this means that if the group
1503 was called as a subroutine from elsewhere in the pattern, things went wrong
1504 (an internal error was given). Such groups are now left in the compiled
1505 pattern, with a new opcode that causes them to be skipped at execution
1506 time.
1508 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1509 to the way PCRE behaves:
1511 (a) A lone ] character is dis-allowed (Perl treats it as data).
1513 (b) A back reference to an unmatched subpattern matches an empty string
1514 (Perl fails the current match path).
1516 (c) A data ] in a character class must be notated as \] because if the
1517 first data character in a class is ], it defines an empty class. (In
1518 Perl it is not possible to have an empty class.) The empty class []
1519 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1520 The negative empty class [^] matches any one character, independently
1521 of the DOTALL setting.
1523 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1524 non-existent subpattern following a character class starting with ']' and
1525 containing () gave an internal compiling error instead of "reference to
1526 non-existent subpattern". Fortunately, when the pattern did exist, the
1527 compiled code was correct. (When scanning forwards to check for the
1528 existencd of the subpattern, it was treating the data ']' as terminating
1529 the class, so got the count wrong. When actually compiling, the reference
1530 was subsequently set up correctly.)
1532 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1533 it was being rejected as not supported by pcre_dfa_exec(), even though
1534 other assertions are supported. I have made pcre_dfa_exec() support
1535 (*FAIL).
1537 16. The implementation of 13c above involved the invention of a new opcode,
1538 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1539 cannot be changed at match time, I realized I could make a small
1540 improvement to matching performance by compiling OP_ALLANY instead of
1541 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1542 on the OP_ANY path.
1544 17. Compiling pcretest on Windows with readline support failed without the
1545 following two fixes: (1) Make the unistd.h include conditional on
1546 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1548 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1549 ncurses library to be included for pcretest when ReadLine support is
1550 requested, but also to allow for it to be overridden. This patch came from
1551 Daniel Bergström.
1553 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1554 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1555 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1556 spotting this.
1559 Version 7.6 28-Jan-08
1560 ---------------------
1562 1. A character class containing a very large number of characters with
1563 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1564 overflow.
1566 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1567 HAVE_LONG_LONG is not defined.
1569 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1570 bring them up to date and include new features. This patch includes:
1572 - Fixed PH's badly added libz and libbz2 support.
1573 - Fixed a problem with static linking.
1574 - Added pcredemo. [But later removed - see 7 below.]
1575 - Fixed dftables problem and added an option.
1576 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1578 - Added readline support for pcretest.
1579 - Added an listing of the option settings after cmake has run.
1581 4. A user submitted a patch to Makefile that makes it easy to create
1582 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1583 Makefile.am that cause it to include this special target, without
1584 affecting anything else. Note that the same mingw target plus all
1585 the other distribution libraries and programs are now supported
1586 when configuring with CMake (see 6 below) instead of with
1587 Configure/Make.
1589 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1590 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1591 exported in the Windows port". It has not yet been confirmed that the patch
1592 solves the problem, but it does no harm.
1594 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1595 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1596 with CMake, and also correct the comment about stack recursion.
1598 7. Remove the automatic building of pcredemo from the ./configure system and
1599 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1600 of a program that users should build themselves after PCRE is installed, so
1601 building it automatically is not really right. What is more, it gave
1602 trouble in some build environments.
1604 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1607 Version 7.5 10-Jan-08
1608 ---------------------
1610 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1611 values in parens when parsing an RE using the C++ wrapper."
1613 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1614 Characters greater than 255 were excluded from the class instead of being
1615 included.
1617 3. The same bug as (2) above applied to negated POSIX classes such as
1618 [:^space:].
1620 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1621 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1622 I have changed it.
1624 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1625 first named subpattern) and a construct such as (?&a) would reference the
1626 first named subpattern whose name started with "a" (in other words, the
1627 length check was missing). Both these problems are fixed. "Subpattern name
1628 expected" is now given for (?&) (a zero-length name), and this patch also
1629 makes it give the same error for \k'' (previously it complained that that
1630 was a reference to a non-existent subpattern).
1632 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1633 this is right because (?- can be followed by option settings as well as by
1634 digits. I have, however, made the messages clearer.
1636 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1637 than the number used in the conditional) now cause a compile-time error.
1638 This is actually not compatible with Perl, which accepts such patterns, but
1639 treats the conditional as always being FALSE (as PCRE used to), but it
1640 seems to me that giving a diagnostic is better.
1642 8. Change "alphameric" to the more common word "alphanumeric" in comments
1643 and messages.
1645 9. Fix two occurrences of "backslash" in comments that should have been
1646 "backspace".
1648 10. Remove two redundant lines of code that can never be obeyed (their function
1649 was moved elsewhere).
1651 11. The program that makes PCRE's Unicode character property table had a bug
1652 which caused it to generate incorrect table entries for sequences of
1653 characters that have the same character type, but are in different scripts.
1654 It amalgamated them into a single range, with the script of the first of
1655 them. In other words, some characters were in the wrong script. There were
1656 thirteen such cases, affecting characters in the following ranges:
1658 U+002b0 - U+002c1
1659 U+0060c - U+0060d
1660 U+0061e - U+00612
1661 U+0064b - U+0065e
1662 U+0074d - U+0076d
1663 U+01800 - U+01805
1664 U+01d00 - U+01d77
1665 U+01d9b - U+01dbf
1666 U+0200b - U+0200f
1667 U+030fc - U+030fe
1668 U+03260 - U+0327f
1669 U+0fb46 - U+0fbb1
1670 U+10450 - U+1049d
1672 12. The -o option (show only the matching part of a line) for pcregrep was not
1673 compatible with GNU grep in that, if there was more than one match in a
1674 line, it showed only the first of them. It now behaves in the same way as
1675 GNU grep.
1677 13. If the -o and -v options were combined for pcregrep, it printed a blank
1678 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1679 does the same. The return code can be used to tell if there were any
1680 non-matching lines.
1682 14. Added --file-offsets and --line-offsets to pcregrep.
1684 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1685 infinitely looping recursion. The bug was that positive lookaheads were not
1686 being skipped when checking for a possible empty match (negative lookaheads
1687 and both kinds of lookbehind were skipped).
1689 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1690 inclusion of <windows.h> to before rather than after the definition of
1691 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1693 17. Specifying a possessive quantifier with a specific limit for a Unicode
1694 character property caused pcre_compile() to compile bad code, which led at
1695 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1696 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1697 caused the error; without that there was no problem.
1699 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1701 19. Added --enable-pcretest-libreadline.
1703 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1704 RE::GlobalReplace(). As a result, the number of replacements returned was
1705 double what it should be. I removed one of the increments, but Craig sent a
1706 later patch that removed the other one (the right fix) and added unit tests
1707 that check the return values (which was not done before).
1709 21. Several CMake things:
1711 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1712 the names libpcre and libpcreposix, not just pcre and pcreposix.
1714 (2) The above change means that pcretest and pcregrep are now correctly
1715 linked with the newly-built libraries, not previously installed ones.
1719 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1720 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1721 UTF-8 newline character). The key issue is that the pattern starts .*;
1722 this means that the match must be either at the beginning, or after a
1723 newline. The bug was in the code for advancing after a failed match and
1724 checking that the new position followed a newline. It was not taking
1725 account of UTF-8 characters correctly.
1727 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1728 character classes. PCRE was not treating the sequence [:...:] as a
1729 character class unless the ... were all letters. Perl, however, seems to
1730 allow any characters between [: and :], though of course it rejects as
1731 unknown any "names" that contain non-letters, because all the known class
1732 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1733 for example, whereas PCRE did not - it did not recognize a POSIX character
1734 class. This seemed a bit dangerous, so the code has been changed to be
1735 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1736 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1737 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1738 Perl does, and where it didn't before.
1740 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1741 Windows environments %n is disabled by default.
1744 Version 7.4 21-Sep-07
1745 ---------------------
1747 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1748 means that a class such as [\s] counted as "explicit reference to CR or
1749 LF". That isn't really right - the whole point of the change was to try to
1750 help when there was an actual mention of one of the two characters. So now
1751 the change happens only if \r or \n (or a literal CR or LF) character is
1752 encountered.
1754 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1755 of both had grown to the point where there were only 3 bits left.
1756 Fortunately, there was spare space in the data structure, and so I have
1757 moved the internal flags into a new 16-bit field to free up more option
1758 bits.
1760 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1761 but did not set the internal JCHANGED flag - either of these is enough to
1762 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1763 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1764 start sets both bits.
1766 4. Added options (at build time, compile time, exec time) to change \R from
1767 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1769 5. doc/pcresyntax.html was missing from the distribution.
1771 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1772 compatibility, even though it is no longer used.
1774 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1775 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1776 windows.h file is present (where different names are used). [This was
1777 reversed later after testing - see 16 below.]
1779 8. Changed all #include <config.h> to #include "config.h". There were also
1780 some further <pcre.h> cases that I changed to "pcre.h".
1782 9. When pcregrep was used with the --colour option, it missed the line ending
1783 sequence off the lines that it output.
1785 10. It was pointed out to me that arrays of string pointers cause lots of
1786 relocations when a shared library is dynamically loaded. A technique of
1787 using a single long string with a table of offsets can drastically reduce
1788 these. I have refactored PCRE in four places to do this. The result is
1789 dramatic:
1791 Originally: 290
1792 After changing UCP table: 187
1793 After changing error message table: 43
1794 After changing table of "verbs" 36
1795 After changing table of Posix names 22
1797 Thanks to the folks working on Gregex for glib for this insight.
1799 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1800 unicode-properties was also set.
1802 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1804 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1805 checked only for CRLF.
1807 14. Added casts to pcretest.c to avoid compiler warnings.
1809 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1811 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1812 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1813 entirely. This removes changes made in 7 above.
1815 17. The CMake files have been updated, and there is now more information about
1816 building with CMake in the NON-UNIX-USE document.
1819 Version 7.3 28-Aug-07
1820 ---------------------
1822 1. In the rejigging of the build system that eventually resulted in 7.1, the
1823 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1824 brackets there is not right, since it causes compilers to look for an
1825 installed pcre.h, not the version that is in the source that is being
1826 compiled (which of course may be different). I have changed it back to:
1828 #include "pcre.h"
1830 I have a vague recollection that the change was concerned with compiling in
1831 different directories, but in the new build system, that is taken care of
1832 by the VPATH setting the Makefile.
1834 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1835 when the subject happened to end in the byte 0x85 (e.g. if the last
1836 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1837 characters but of course it shouldn't be taken as a newline when it is part
1838 of another character. The bug was that, for an unlimited repeat of . in
1839 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1840 characters when looking for a newline.
1842 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1844 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1845 in debug output.
1847 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1848 long printing in the pcrecpp unittest when running under MinGW.
1850 6. ESC_K was left out of the EBCDIC table.
1852 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1853 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1854 limit also applies to "virtual nesting" when a pattern is recursive, and in
1855 this case 1000 isn't so big. I have been able to remove this limit at the
1856 expense of backing off one optimization in certain circumstances. Normally,
1857 when pcre_exec() would call its internal match() function recursively and
1858 immediately return the result unconditionally, it uses a "tail recursion"
1859 feature to save stack. However, when a subpattern that can match an empty
1860 string has an unlimited repetition quantifier, it no longer makes this
1861 optimization. That gives it a stack frame in which to save the data for
1862 checking that an empty string has been matched. Previously this was taken
1863 from the 1000-entry workspace that had been reserved. So now there is no
1864 explicit limit, but more stack is used.
1866 8. Applied Daniel's patches to solve problems with the import/export magic
1867 syntax that is required for Windows, and which was going wrong for the
1868 pcreposix and pcrecpp parts of the library. These were overlooked when this
1869 problem was solved for the main library.
1871 9. There were some crude static tests to avoid integer overflow when computing
1872 the size of patterns that contain repeated groups with explicit upper
1873 limits. As the maximum quantifier is 65535, the maximum group length was
1874 set at 30,000 so that the product of these two numbers did not overflow a
1875 32-bit integer. However, it turns out that people want to use groups that
1876 are longer than 30,000 bytes (though not repeat them that many times).
1877 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1878 made it possible to implement the integer overflow checks in a much more
1879 dynamic way, which I have now done. The artificial limitation on group
1880 length has been removed - we now have only the limit on the total length of
1881 the compiled pattern, which depends on the LINK_SIZE setting.
1883 10. Fixed a bug in the documentation for get/copy named substring when
1884 duplicate names are permitted. If none of the named substrings are set, the
1885 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1886 empty string.
1888 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1889 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1890 because the ] is interpreted as the first data character and the
1891 terminating ] is not found. PCRE has been made compatible with Perl in this
1892 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1893 cause memory overwriting.
1895 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1896 string has been matched (to stop an infinite loop). It was not recognizing
1897 a conditional subpattern that could match an empty string if that
1898 subpattern was within another subpattern. For example, it looped when
1899 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1900 condition was not nested. This bug has been fixed.
1902 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1903 past the start of the subject in the presence of bytes with the top bit
1904 set, for example "\x8aBCD".
1906 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1907 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1909 14. Optimized (?!) to (*FAIL).
1911 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1912 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1913 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1914 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1915 does: it's just the validity check that is more restrictive.
1917 16. Inserted checks for integer overflows during escape sequence (backslash)
1918 processing, and also fixed erroneous offset values for syntax errors during
1919 backslash processing.
1921 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1922 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1924 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1925 caused an overrun.
1927 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1928 something other than just ASCII characters) inside a group that had an
1929 unlimited repeat caused a loop at compile time (while checking to see
1930 whether the group could match an empty string).
1932 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1933 [\P{Any}] did so. (Error in the code for printing property names.)
1935 21. An orphan \E inside a character class could cause a crash.
1937 22. A repeated capturing bracket such as (A)? could cause a wild memory
1938 reference during compilation.
1940 23. There are several functions in pcre_compile() that scan along a compiled
1941 expression for various reasons (e.g. to see if it's fixed length for look
1942 behind). There were bugs in these functions when a repeated \p or \P was
1943 present in the pattern. These operators have additional parameters compared
1944 with \d, etc, and these were not being taken into account when moving along
1945 the compiled data. Specifically:
1947 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1948 length.
1950 (b) An item such as \pL+ within a repeated group could cause crashes or
1951 loops.
1953 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1954 "reference to non-existent subpattern" error.
1956 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1958 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1959 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1961 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1963 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1964 character were causing crashes (broken optimization).
1966 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1967 \p or \P) caused a compile-time loop.
1969 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1970 break. For example, the unstudied pattern [\r\n]A does not match the string
1971 "\r\nA" because change 7.0/46 below moves the current point on by two
1972 characters after failing to match at the start. However, the pattern \nA
1973 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1974 the same is true. There doesn't seem any very clean way out of this, but
1975 what I have chosen to do makes the common cases work: PCRE now takes note
1976 of whether there can be an explicit match for \r or \n anywhere in the
1977 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1978 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1979 pattern has explicit CR or LF references.
1981 29. Added (*CR) etc for changing newline setting at start of pattern.
1984 Version 7.2 19-Jun-07
1985 ---------------------
1987 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1988 which is apparently normally available under Windows.
1990 2. Re-jig the pcregrep tests with different newline settings in an attempt
1991 to make them independent of the local environment's newline setting.
1993 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1995 4. Some of the "internals" tests were previously cut out when the link size
1996 was not 2, because the output contained actual offsets. The recent new
1997 "Z" feature of pcretest means that these can be cut out, making the tests
1998 usable with all link sizes.
2000 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2001 stack recursion. This gives a massive performance boost under BSD, but just
2002 a small improvement under Linux. However, it saves one field in the frame
2003 in all cases.
2005 6. Added more features from the forthcoming Perl 5.10:
2007 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2008 recursion call. It refers to the nth most recently opened parentheses.
2010 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2011 to be opened parentheses.
2013 (c) Conditions that refer to capturing parentheses can be specified
2014 relatively, for example, (?(-2)... or (?(+3)...
2016 (d) \K resets the start of the current match so that everything before
2017 is not part of it.
2019 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2021 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2022 reference syntax.
2024 (g) (?| introduces a group in which the numbering of parentheses in each
2025 alternative starts with the same number.
2027 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2029 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2032 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2033 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2034 for detecting groups that can match an empty string.
2036 9. A pattern with a very large number of alternatives (more than several
2037 hundred) was running out of internal workspace during the pre-compile
2038 phase, where pcre_compile() figures out how much memory will be needed. A
2039 bit of new cunning has reduced the workspace needed for groups with
2040 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2041 workspace instead of running out of the 4096 that are available.
2043 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2045 11. Applied patch from Google to remove an optimization that didn't quite work.
2046 The report of the bug said:
2048 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2049 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2050 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2052 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2053 it matched the wrong number of bytes.
2056 Version 7.1 24-Apr-07
2057 ---------------------
2059 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2060 that is more "standard", making use of automake and other Autotools. There
2061 is some re-arrangement of the files and adjustment of comments consequent
2062 on this.
2064 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2065 for recursive directory scanning broke on some systems because the files
2066 are not scanned in any specific order and on different systems the order
2067 was different. A call to "sort" has been inserted into RunGrepTest for the
2068 approprate test as a short-term fix. In the longer term there may be an
2069 alternative.
2071 3. I had an email from Eric Raymond about problems translating some of PCRE's
2072 man pages to HTML (despite the fact that I distribute HTML pages, some
2073 people do their own conversions for various reasons). The problems
2074 concerned the use of low-level troff macros .br and .in. I have therefore
2075 removed all such uses from the man pages (some were redundant, some could
2076 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2077 HTML has been updated to handle .nf/.fi and to complain if it encounters
2078 .br or .in.
2080 4. Updated comments in configure.ac that get placed in config.h.in and also
2081 arranged for config.h to be included in the distribution, with the name
2082 config.h.generic, for the benefit of those who have to compile without
2083 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2085 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2086 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2087 makevp.bat for the current PCRE, using the additional files
2088 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2090 6. A Windows user reported a minor discrepancy with test 2, which turned out
2091 to be caused by a trailing space on an input line that had got lost in his
2092 copy. The trailing space was an accident, so I've just removed it.
2094 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2095 that is needed.
2097 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2098 as "const" (a) because they are and (b) because it helps the PHP
2099 maintainers who have recently made a script to detect big data structures
2100 in the php code that should be moved to the .rodata section. I remembered
2101 to update Builducptable as well, so it won't revert if ucptable.h is ever
2102 re-created.
2104 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2105 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2106 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2107 support is not required. This saves 1.5-2K of code, which is important in
2108 some applications.
2110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2111 so as not to refer to the tables, even though these functions will never be
2112 called when UTF-8 support is disabled. Otherwise there are problems with a
2113 shared library.
2115 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2117 (a) It was defining its arguments as char * instead of void *.
2119 (b) It was assuming that all moves were upwards in memory; this was true
2120 a long time ago when I wrote it, but is no longer the case.
2122 The emulated memove() is provided for those environments that have neither
2123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2124 is clearly not the case, as these two bugs were recently reported.
2126 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2127 and Detrail to create the HTML documentation, the .txt form of the man
2128 pages, and it removes trailing spaces from listed files. It also creates
2129 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2130 case, it wraps all the #defines with #ifndefs. This script should be run
2131 before "make dist".
2133 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2134 with Unicode property support.
2136 (a) For a maximizing quantifier, if the two different cases of the
2137 character were of different lengths in their UTF-8 codings (there are
2138 some cases like this - I found 11), and the matching function had to
2139 back up over a mixture of the two cases, it incorrectly assumed they
2140 were both the same length.
2142 (b) When PCRE was configured to use the heap rather than the stack for
2143 recursion during matching, it was not correctly preserving the data for
2144 the other case of a UTF-8 character when checking ahead for a match
2145 while processing a minimizing repeat. If the check also involved
2146 matching a wide character, but failed, corruption could cause an
2147 erroneous result when trying to check for a repeat of the original
2148 character.
2150 13. Some tidying changes to the testing mechanism:
2152 (a) The RunTest script now detects the internal link size and whether there
2153 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2154 values substituted by "configure". (The RunGrepTest script already did
2155 this for UTF-8.) The configure.ac script no longer substitutes the
2156 relevant variables.
2158 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2159 with length and offset values. This means that the output is different
2160 for different internal link sizes. Test 2 is skipped for link sizes
2161 other than 2 because of this, bypassing the problem. Unfortunately,
2162 there was also a test in test 3 (the locale tests) that used /B and
2163 failed for link sizes other than 2. Rather than cut the whole test out,
2164 I have added a new /Z option to pcretest that replaces the length and
2165 offset values with spaces. This is now used to make test 3 independent
2166 of link size. (Test 2 will be tidied up later.)
2168 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2169 segmentation fault instead of returning the appropriate error message.
2171 15. In multiline mode when the newline sequence was set to "any", the pattern
2172 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2173 This doesn't seem right; it now treats the CRLF combination as the line
2174 ending, and so does not match in that case. It's only a pattern such as ^$
2175 that would hit this one: something like ^ABC$ would have failed after \r
2176 and then tried again after \r\n.
2178 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2179 in an attempt to make files that differ only in their line terminators
2180 compare equal. This works on Linux.
2182 17. Under certain error circumstances pcregrep might try to free random memory
2183 as it exited. This is now fixed, thanks to valgrind.
2185 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2186 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2187 was because its rules for how to advance for /g after matching an empty
2188 string at the end of a line did not allow for this case. They now check for
2189 it specially.
2191 20. pcretest is supposed to handle patterns and data of any length, by
2192 extending its buffers when necessary. It was getting this wrong when the
2193 buffer for a data line had to be extended.
2195 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2196 CRLF as a newline sequence.
2198 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2199 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2200 I have nevertheless tidied it up.
2202 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2204 24. Added a man page for pcre-config.
2207 Version 7.0 19-Dec-06
2208 ---------------------
2210 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2211 moving to gcc 4.1.1.
2213 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2214 sys/time.h, which is documented as needed for this function. It doesn't
2215 seem to matter on Linux, but it showed up on some releases of OS X.
2217 3. It seems that there are systems where bytes whose values are greater than
2218 127 match isprint() in the "C" locale. The "C" locale should be the
2219 default when a C program starts up. In most systems, only ASCII printing
2220 characters match isprint(). This difference caused the output from pcretest
2221 to vary, making some of the tests fail. I have changed pcretest so that:
2223 (a) When it is outputting text in the compiled version of a pattern, bytes
2224 other than 32-126 are always shown as hex escapes.
2226 (b) When it is outputting text that is a matched part of a subject string,
2227 it does the same, unless a different locale has been set for the match
2228 (using the /L modifier). In this case, it uses isprint() to decide.
2230 4. Fixed a major bug that caused incorrect computation of the amount of memory
2231 required for a compiled pattern when options that changed within the
2232 pattern affected the logic of the preliminary scan that determines the
2233 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2234 that the computed length was too small. The symptoms of this bug were
2235 either the PCRE error "internal error: code overflow" from pcre_compile(),
2236 or a glibc crash with a message such as "pcretest: free(): invalid next
2237 size (fast)". Examples of patterns that provoked this bug (shown in
2238 pcretest format) are:
2240 /(?-x: )/x
2241 /(?x)(?-x: \s*#\s*)/
2242 /((?i)[\x{c0}])/8
2243 /(?i:[\x{c0}])/8
2245 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2246 is now done differently.
2248 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2249 wrapper classes; (b) implement a new function in the C++ scanner that is
2250 more efficient than the old way of doing things because it avoids levels of
2251 recursion in the regex matching; (c) add a paragraph to the documentation
2252 for the FullMatch() function.
2254 6. The escape sequence \n was being treated as whatever was defined as
2255 "newline". Not only was this contrary to the documentation, which states
2256 that \n is character 10 (hex 0A), but it also went horribly wrong when
2257 "newline" was defined as CRLF. This has been fixed.
2259 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2260 was being set to -1 for the "end of line" case (supposedly a value that no
2261 character can have). Though this value is never used (the check for end of
2262 line is "zero bytes in current character"), it caused compiler complaints.
2263 I've changed it to 0xffffffff.
2265 8. In pcre_version.c, the version string was being built by a sequence of
2266 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2267 string (as it is for production releases) called a macro with an empty
2268 argument. The C standard says the result of this is undefined. The gcc
2269 compiler treats it as an empty string (which was what was wanted) but it is
2270 reported that Visual C gives an error. The source has been hacked around to
2271 avoid this problem.
2273 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2274 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2275 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2276 of them did).
2278 10. Originally, pcretest opened its input and output without "b"; then I was
2279 told that "b" was needed in some environments, so it was added for release
2280 5.0 to both the input and output. (It makes no difference on Unix-like
2281 systems.) Later I was told that it is wrong for the input on Windows. I've
2282 now abstracted the modes into two macros, to make it easier to fiddle with
2283 them, and removed "b" from the input mode under Windows.
2285 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2287 12. Added -help and --help to pcretest as an official way of being reminded
2288 of the options.
2290 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2291 and pcrecpp.cc because they annoy compilers at high warning levels.
2293 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2295 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2296 scripts are not C programs :-) and which was not noticed because it works
2297 on Linux.
2299 16. pcretest is supposed to handle any length of pattern and data line (as one
2300 line or as a continued sequence of lines) by extending its input buffer if
2301 necessary. This feature was broken for very long pattern lines, leading to
2302 a string of junk being passed to pcre_compile() if the pattern was longer
2303 than about 50K.
2305 17. I have done a major re-factoring of the way pcre_compile() computes the
2306 amount of memory needed for a compiled pattern. Previously, there was code
2307 that made a preliminary scan of the pattern in order to do this. That was
2308 OK when PCRE was new, but as the facilities have expanded, it has become
2309 harder and harder to keep it in step with the real compile phase, and there
2310 have been a number of bugs (see for example, 4 above). I have now found a
2311 cunning way of running the real compile function in a "fake" mode that
2312 enables it to compute how much memory it would need, while actually only
2313 ever using a few hundred bytes of working memory and without too many
2314 tests of the mode. This should make future maintenance and development
2315 easier. A side effect of this work is that the limit of 200 on the nesting
2316 depth of parentheses has been removed (though this was never a serious
2317 limitation, I suspect). However, there is a downside: pcre_compile() now
2318 runs more slowly than before (30% or more, depending on the pattern). I
2319 hope this isn't a big issue. There is no effect on runtime performance.
2321 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2322 newline (only possible for the last line of a file) and it was a
2323 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2325 19. Added additional timing features to pcretest. (1) The -tm option now times
2326 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2327 separate command line item, by a number that specifies the number of
2328 repeats to use when timing. The default is 50000; this gives better
2329 precision, but takes uncomfortably long for very large patterns.
2331 20. Extended pcre_study() to be more clever in cases where a branch of a
2332 subpattern has no definite first character. For example, (a*|b*)[cd] would
2333 previously give no result from pcre_study(). Now it recognizes that the
2334 first character must be a, b, c, or d.
2336 21. There was an incorrect error "recursive call could loop indefinitely" if
2337 a subpattern (or the entire pattern) that was being tested for matching an
2338 empty string contained only one non-empty item after a nested subpattern.
2339 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2340 incorrectly, because the \d was being skipped in the check.
2342 22. The pcretest program now has a new pattern option /B and a command line
2343 option -b, which is equivalent to adding /B to every pattern. This causes
2344 it to show the compiled bytecode, without the additional information that
2345 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2346 is the same as /B/I).
2348 23. A new optimization is now able automatically to treat some sequences such
2349 as a*b as a*+b. More specifically, if something simple (such as a character
2350 or a simple class like \d) has an unlimited quantifier, and is followed by
2351 something that cannot possibly match the quantified thing, the quantifier
2352 is automatically "possessified".
2354 24. A recursive reference to a subpattern whose number was greater than 39
2355 went wrong under certain circumstances in UTF-8 mode. This bug could also
2356 have affected the operation of pcre_study().
2358 25. Realized that a little bit of performance could be had by replacing
2359 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2361 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2363 27. Possessive quantifiers such as a++ were previously implemented by turning
2364 them into atomic groups such as ($>a+). Now they have their own opcodes,
2365 which improves performance. This includes the automatically created ones
2366 from 23 above.
2368 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2369 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2370 the first matched character to be a colon. This applied both to named and
2371 numbered groups.
2373 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2375 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2376 building PCRE on Windows easier, so I have included it in the distribution.
2378 31. There is now a check in pcretest against a ridiculously large number being
2379 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2380 loop, the loop is abandoned.
2382 32. Forward references to subpatterns in conditions such as (?(2)...) where
2383 subpattern 2 is defined later cause pcre_compile() to search forwards in
2384 the pattern for the relevant set of parentheses. This search went wrong
2385 when there were unescaped parentheses in a character class, parentheses
2386 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2388 33. "Subroutine" calls and backreferences were previously restricted to
2389 referencing subpatterns earlier in the regex. This restriction has now
2390 been removed.
2392 34. Added a number of extra features that are going to be in Perl 5.10. On the
2393 whole, these are just syntactic alternatives for features that PCRE had
2394 previously implemented using the Python syntax or my own invention. The
2395 other formats are all retained for compatibility.
2397 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2398 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2399 also .NET compatible.
2401 (b) A recursion or subroutine call to a named group can now be defined as
2402 (?&name) as well as (?P>name).
2404 (c) A backreference to a named group can now be defined as \k<name> or
2405 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2406 5.10, are also .NET compatible.
2408 (d) A conditional reference to a named group can now use the syntax
2409 (?(<name>) or (?('name') as well as (?(name).
2411 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2412 groups (named and numbered) that are never evaluated inline, but can be
2413 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2414 is always false. There may be only one alternative in such a group.
2416 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2417 as the simple (?(R). The condition is true only if the most recent
2418 recursion is that of the given number or name. It does not search out
2419 through the entire recursion stack.
2421 (g) The escape \gN or \g{N} has been added, where N is a positive or
2422 negative number, specifying an absolute or relative reference.
2424 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2425 some "unreachable code" warnings.
2427 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2428 things, this adds five new scripts.
2430 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2431 There were also incompatibilities regarding the handling of \Q..\E inside
2432 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2433 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2435 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2436 matches an empty string, and forcibly breaks the loop. There were bugs in
2437 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2438 against aaaa the result was just "a" rather than "aaaa", for example. Two
2439 separate and independent bugs (that affected different cases) have been
2440 fixed.
2442 39. Refactored the code to abolish the use of different opcodes for small
2443 capturing bracket numbers. This is a tidy that I avoided doing when I
2444 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2445 The new approach is not only tidier, it makes it possible to reduce the
2446 memory needed to fix the previous bug (38).
2448 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2449 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2450 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2451 mode.
2453 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2454 report.
2456 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2457 copy construction and assignment in the C++ wrapper.
2459 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2460 couple of bugs that could have given wrong results in the "--newline=crlf"
2461 case.
2463 44. Added a number of casts and did some reorganization of signed/unsigned int
2464 variables following suggestions from Dair Grant. Also renamed the variable
2465 "this" as "item" because it is a C++ keyword.
2467 45. Arranged for dftables to add
2469 #include "pcre_internal.h"
2471 to pcre_chartables.c because without it, gcc 4.x may remove the array
2472 definition from the final binary if PCRE is built into a static library and
2473 dead code stripping is activated.
2475 46. For an unanchored pattern, if a match attempt fails at the start of a
2476 newline sequence, and the newline setting is CRLF or ANY, and the next two
2477 characters are CRLF, advance by two characters instead of one.
2480 Version 6.7 04-Jul-06
2481 ---------------------
2483 1. In order to handle tests when input lines are enormously long, pcretest has
2484 been re-factored so that it automatically extends its buffers when
2485 necessary. The code is crude, but this _is_ just a test program. The
2486 default size has been increased from 32K to 50K.
2488 2. The code in pcre_study() was using the value of the re argument before
2489 testing it for NULL. (Of course, in any sensible call of the function, it
2490 won't be NULL.)
2492 3. The memmove() emulation function in pcre_internal.h, which is used on
2493 systems that lack both memmove() and bcopy() - that is, hardly ever -
2494 was missing a "static" storage class specifier.
2496 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2497 containing an extended class (one that cannot be represented by a bitmap
2498 because it contains high-valued characters or Unicode property items, e.g.
2499 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2500 pattern, but PCRE should not loop if you do not (it no longer does).
2501 [Detail: two cases were found: (a) a repeated subpattern containing an
2502 extended class; (b) a recursive reference to a subpattern that followed a
2503 previous extended class. It wasn't skipping over the extended class
2504 correctly when UTF-8 mode was not set.]
2506 5. A negated single-character class was not being recognized as fixed-length
2507 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2508 compile error "lookbehind assertion is not fixed length".
2510 6. The RunPerlTest auxiliary script was showing an unexpected difference
2511 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2512 write a Perl script that can interpret lines of an input file either as
2513 byte characters or as UTF-8, which is what "perltest" was being required to
2514 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2515 can't do is switch easily at run time between having the "use utf8;" pragma
2516 or not. In the end, I fudged it by using the RunPerlTest script to insert
2517 "use utf8;" explicitly for the UTF-8 tests.
2519 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2520 the end of the subject string, contrary to the documentation and to what
2521 Perl does. This was true of both matching functions. Now it matches only at
2522 the start of the subject and immediately after *internal* newlines.
2524 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2525 a pointer to an int instead of a pointer to an unsigned long int. This
2526 caused problems on 64-bit systems.
2528 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2529 instance of the 'standard' template library not being so standard".
2531 10. There was no check on the number of named subpatterns nor the maximum
2532 length of a subpattern name. The product of these values is used to compute
2533 the size of the memory block for a compiled pattern. By supplying a very
2534 long subpattern name and a large number of named subpatterns, the size
2535 computation could be caused to overflow. This is now prevented by limiting
2536 the length of names to 32 characters, and the number of named subpatterns
2537 to 10,000.
2539 11. Subpatterns that are repeated with specific counts have to be replicated in
2540 the compiled pattern. The size of memory for this was computed from the
2541 length of the subpattern and the repeat count. The latter is limited to
2542 65535, but there was no limit on the former, meaning that integer overflow
2543 could in principle occur. The compiled length of a repeated subpattern is
2544 now limited to 30,000 bytes in order to prevent this.
2546 12. Added the optional facility to have named substrings with the same name.
2548 13. Added the ability to use a named substring as a condition, using the
2549 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2550 are numbers (not recommended). Forward references are permitted.
2552 14. Added forward references in named backreferences (if you see what I mean).
2554 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2555 pattern could run off the end of the subject. For example, the pattern
2556 "(?s)(.{1,5})"8 did this with the subject "ab".
2558 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2559 PCRE_CASELESS was set when matching characters that were quantified with ?
2560 or *.
2562 17. A character class other than a single negated character that had a minimum
2563 but no maximum quantifier - for example [ab]{6,} - was not handled
2564 correctly by pce_dfa_exec(). It would match only one character.
2566 18. A valid (though odd) pattern that looked like a POSIX character
2567 class but used an invalid character after [ (for example [[,abc,]]) caused
2568 pcre_compile() to give the error "Failed: internal error: code overflow" or
2569 in some cases to crash with a glibc free() error. This could even happen if
2570 the pattern terminated after [[ but there just happened to be a sequence of
2571 letters, a binary zero, and a closing ] in the memory that followed.
2573 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2574 over the years. Originally (before any Unicode support), just the bottom 8
2575 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2576 output from "man perlunicode" includes this:
2578 The regular expression compiler produces polymorphic opcodes. That
2579 is, the pattern adapts to the data and automatically switches to
2580 the Unicode character scheme when presented with Unicode data--or
2581 instead uses a traditional byte scheme when presented with byte
2582 data.
2584 Sadly, a wide octal escape does not cause a switch, and in a string with
2585 no other multibyte characters, these octal escapes are treated as before.
2586 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2587 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2588 Unicode string.
2590 I have not perpetrated such confusion in PCRE. Up till now, it took just
2591 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2592 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2593 translate to the appropriate multibyte character.
2595 29. Applied some refactoring to reduce the number of warnings from Microsoft
2596 and Borland compilers. This has included removing the fudge introduced
2597 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2598 a warning about an unused variable.
2600 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2601 characters since release 4.0, because Perl (from release 5.004) does not.
2602 [Or at least, is documented not to: some releases seem to be in conflict
2603 with the documentation.] However, when a pattern was studied with
2604 pcre_study() and all its branches started with \s, PCRE still included VT
2605 as a possible starting character. Of course, this did no harm; it just
2606 caused an unnecessary match attempt.
2608 22. Removed a now-redundant internal flag bit that recorded the fact that case
2609 dependency changed within the pattern. This was once needed for "required
2610 byte" processing, but is no longer used. This recovers a now-scarce options
2611 bit. Also moved the least significant internal flag bit to the most-
2612 significant bit of the word, which was not previously used (hangover from
2613 the days when it was an int rather than a uint) to free up another bit for
2614 the future.
2616 23. Added support for CRLF line endings as well as CR and LF. As well as the
2617 default being selectable at build time, it can now be changed at runtime
2618 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2619 specify that it is scanning data with non-default line endings.
2621 24. Changed the definition of CXXLINK to make it agree with the definition of
2622 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2624 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2625 recursions. This makes a big different to stack usage for some patterns.
2627 26. If a subpattern containing a named recursion or subroutine reference such
2628 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2629 the space required for the compiled pattern went wrong and gave too small a
2630 value. Depending on the environment, this could lead to "Failed: internal
2631 error: code overflow at offset 49" or "glibc detected double free or
2632 corruption" errors.
2634 27. Applied patches from Google (a) to support the new newline modes and (b) to
2635 advance over multibyte UTF-8 characters in GlobalReplace.
2637 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2638 difference for some implementation of PCRE in some Windows version.
2640 29. Added some extra testing facilities to pcretest:
2642 \q<number> in a data line sets the "match limit" value
2643 \Q<number> in a data line sets the "match recursion limt" value
2644 -S <number> sets the stack size, where <number> is in megabytes
2646 The -S option isn't available for Windows.
2649 Version 6.6 06-Feb-06
2650 ---------------------
2652 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2653 in pcreposix.h. I have copied the definition from pcre.h.
2655 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2656 because pcre.h is no longer a built file.
2658 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2659 not normally included in the compiled code.
2662 Version 6.5 01-Feb-06
2663 ---------------------
2665 1. When using the partial match feature with pcre_dfa_exec(), it was not
2666 anchoring the second and subsequent partial matches at the new starting
2667 point. This could lead to incorrect results. For example, with the pattern
2668 /1234/, partially matching against "123" and then "a4" gave a match.
2670 2. Changes to pcregrep:
2672 (a) All non-match returns from pcre_exec() were being treated as failures
2673 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2674 error message is output. Some extra information is given for the
2676 probably the only errors that are likely to be caused by users (by
2677 specifying a regex that has nested indefinite repeats, for instance).
2678 If there are more than 20 of these errors, pcregrep is abandoned.
2680 (b) A binary zero was treated as data while matching, but terminated the
2681 output line if it was written out. This has been fixed: binary zeroes
2682 are now no different to any other data bytes.
2684 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2685 used to set a locale for matching. The --locale=xxxx long option has
2686 been added (no short equivalent) to specify a locale explicitly on the
2687 pcregrep command, overriding the environment variables.
2689 (d) When -B was used with -n, some line numbers in the output were one less
2690 than they should have been.
2692 (e) Added the -o (--only-matching) option.
2694 (f) If -A or -C was used with -c (count only), some lines of context were
2695 accidentally printed for the final match.
2697 (g) Added the -H (--with-filename) option.
2699 (h) The combination of options -rh failed to suppress file names for files
2700 that were found from directory arguments.
2702 (i) Added the -D (--devices) and -d (--directories) options.
2704 (j) Added the -F (--fixed-strings) option.
2706 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2708 (l) Added the --colo(u)r option.
2710 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2711 is not present by default.
2713 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2714 items such as (?R) or (?1), when the recursion could match a number of
2715 alternatives. If it matched one of the alternatives, but subsequently,
2716 outside the recursion, there was a failure, the code tried to back up into
2717 the recursion. However, because of the way PCRE is implemented, this is not
2718 possible, and the result was an incorrect result from the match.
2720 In order to prevent this happening, the specification of recursion has
2721 been changed so that all such subpatterns are automatically treated as
2722 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2724 4. I had overlooked the fact that, in some locales, there are characters for
2725 which isalpha() is true but neither isupper() nor islower() are true. In
2726 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2727 and ordfeminine) are like this. This affected the treatment of \w and \W
2728 when they appeared in character classes, but not when they appeared outside
2729 a character class. The bit map for "word" characters is now created
2730 separately from the results of isalnum() instead of just taking it from the
2731 upper, lower, and digit maps. (Plus the underscore character, of course.)
2733 5. The above bug also affected the handling of POSIX character classes such as
2734 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2735 permanent tables. Instead, the bit maps for such a class were previously
2736 created as the appropriate unions of the upper, lower, and digit bitmaps.
2737 Now they are created by subtraction from the [[:word:]] class, which has
2738 its own bitmap.
2740 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2741 It is created by subtracting the vertical space characters (\x09, \x0a,
2742 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2743 subtraction was done in the overall bitmap for a character class, meaning
2744 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2745 be recognized. This bug has been fixed.
2747 7. Patches from the folks at Google:
2749 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2750 real life, but is still worth protecting against".
2752 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2753 regular expressions".
2755 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2756 have it.
2758 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2759 "configure" and the latter not, in order to fix a problem somebody had
2760 with compiling the Arg class on HP-UX.
2762 (e) Improve the error-handling of the C++ wrapper a little bit.
2764 (f) New tests for checking recursion limiting.
2766 8. The pcre_memmove() function, which is used only if the environment does not
2767 have a standard memmove() function (and is therefore rarely compiled),
2768 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2769 returning a result (though PCRE never actually uses the result).
2771 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2772 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2773 returned instead of calling malloc() with an overflowing number that would
2774 most likely cause subsequent chaos.
2776 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2778 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2779 with this option is matched, the nmatch and pmatch options of regexec() are
2780 ignored.
2782 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2783 provided in case anyone wants to the the POSIX interface with UTF-8
2784 strings.
2786 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2787 C++ linking (needed for some HP-UX environments).
2789 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2790 (unused parameter) and in the pcre_printint() function (omitted "default"
2791 switch label when the default is to do nothing).
2793 15. Added some code to make it possible, when PCRE is compiled as a C++
2794 library, to replace subject pointers for pcre_exec() with a smart pointer
2795 class, thus making it possible to process discontinuous strings.
2797 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2798 much the same function. They were added by different people who were trying
2799 to make PCRE easy to compile on non-Unix systems. It has been suggested
2800 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2801 for compiling on Windows systems. I have therefore replaced it with
2802 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2803 defaults to "extern" for C or "extern C" for C++, which works fine on
2804 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2805 SCOPE with something explicit in config.h. In addition:
2807 (a) pcreposix.h still had just "extern" instead of either of these macros;
2808 I have replaced it with PCRE_DATA_SCOPE.
2810 (b) Functions such as _pcre_xclass(), which are internal to the library,
2811 but external in the C sense, all had PCRE_EXPORT in their definitions.
2812 This is apparently wrong for the Windows case, so I have removed it.
2813 (It makes no difference on Unix-like systems.)
2815 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2816 of recursive calls to match(). This is different to MATCH_LIMIT because
2817 that limits the total number of calls to match(), not all of which increase
2818 the depth of recursion. Limiting the recursion depth limits the amount of
2819 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2820 when PCRE is compiled, and changed at run time. A patch from Google adds
2821 this functionality to the C++ interface.
2823 18. Changes to the handling of Unicode character properties:
2825 (a) Updated the table to Unicode 4.1.0.
2827 (b) Recognize characters that are not in the table as "Cn" (undefined).
2829 (c) I revised the way the table is implemented to a much improved format
2830 which includes recognition of ranges. It now supports the ranges that
2831 are defined in UnicodeData.txt, and it also amalgamates other
2832 characters into ranges. This has reduced the number of entries in the
2833 table from around 16,000 to around 3,000, thus reducing its size
2834 considerably. I realized I did not need to use a tree structure after
2835 all - a binary chop search is just as efficient. Having reduced the
2836 number of entries, I extended their size from 6 bytes to 8 bytes to
2837 allow for more data.
2839 (d) Added support for Unicode script names via properties such as \p{Han}.
2841 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2842 matching that character.
2844 20. When matching a repeated Unicode property with a minimum greater than zero,
2845 (for example \pL{2,}), PCRE could look past the end of the subject if it
2846 reached it while seeking the minimum number of characters. This could
2847 happen only if some of the characters were more than one byte long, because
2848 there is a check for at least the minimum number of bytes.
2850 21. Refactored the implementation of \p and \P so as to be more general, to
2851 allow for more different types of property in future. This has changed the
2852 compiled form incompatibly. Anybody with saved compiled patterns that use
2853 \p or \P will have to recompile them.
2855 22. Added "Any" and "L&" to the supported property types.
2857 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2858 but give a compile time error if the value is greater than 0xff.
2860 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2861 accidentally not being installed or uninstalled.
2863 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2864 made were to insert the current release number. This seemed silly, because
2865 it made things harder for people building PCRE on systems that don't run
2866 "configure". I have turned pcre.h into a distributed file, no longer built
2867 by "configure", with the version identification directly included. There is
2868 no longer a pcre.h.in file.
2870 However, this change necessitated a change to the pcre-config script as
2871 well. It is built from pcre-config.in, and one of the substitutions was the
2872 release number. I have updated configure.ac so that ./configure now finds
2873 the release number by grepping pcre.h.
2875 26. Added the ability to run the tests under valgrind.
2878 Version 6.4 05-Sep-05
2879 ---------------------
2881 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2882 "--" to be printed when multiple files were scanned, even when none of the
2883 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2884 consider it to be a bug, and have restored the previous behaviour.
2886 2. A couple of code tidies to get rid of compiler warnings.
2888 3. The pcretest program used to cheat by referring to symbols in the library
2889 whose names begin with _pcre_. These are internal symbols that are not
2890 really supposed to be visible externally, and in some environments it is
2891 possible to suppress them. The cheating is now confined to including
2892 certain files from the library's source, which is a bit cleaner.
2894 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2895 file's purpose clearer.
2897 5. Reorganized pcre_ucp_findchar().
2900 Version 6.3 15-Aug-05
2901 ---------------------
2903 1. The file libpcre.pc.in did not have general read permission in the tarball.
2905 2. There were some problems when building without C++ support:
2907 (a) If C++ support was not built, "make install" and "make test" still
2908 tried to test it.
2910 (b) There were problems when the value of CXX was explicitly set. Some
2911 changes have been made to try to fix these, and ...
2913 (c) --disable-cpp can now be used to explicitly disable C++ support.
2915 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2916 backslash in a target when C++ was disabled. This confuses some
2917 versions of "make", apparently. Using an intermediate variable solves
2918 this. (Same for CPP_LOBJ.)
2920 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2921 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2922 necessary on certain architectures.
2924 4. Added a setting of -export-symbols-regex to the link command to remove
2925 those symbols that are exported in the C sense, but actually are local
2926 within the library, and not documented. Their names all begin with
2927 "_pcre_". This is not a perfect job, because (a) we have to except some
2928 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2929 available (and never for static libraries). I have made a note to try to
2930 find a way round (a) in the future.
2933 Version 6.2 01-Aug-05
2934 ---------------------
2936 1. There was no test for integer overflow of quantifier values. A construction
2937 such as {1111111111111111} would give undefined results. What is worse, if
2938 a minimum quantifier for a parenthesized subpattern overflowed and became
2939 negative, the calculation of the memory size went wrong. This could have
2940 led to memory overwriting.
2942 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2944 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2945 operating environments where this matters.
2947 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2948 PCRE options from within the C++ wrapper.
2950 5. Named capturing subpatterns were not being correctly counted when a pattern
2951 was compiled. This caused two problems: (a) If there were more than 100
2952 such subpatterns, the calculation of the memory needed for the whole
2953 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2954 back references of the form \12, where the number was greater than 9, were
2955 not recognized as back references, even though there were sufficient
2956 previous subpatterns.
2958 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2959 versions of gcc, e.g. 2.95.4.
2962 Version 6.1 21-Jun-05
2963 ---------------------
2965 1. There was one reference to the variable "posix" in pcretest.c that was not
2966 surrounded by "#if !defined NOPOSIX".
2968 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2969 the cross-check on the old pcre_info() function, for the benefit of the
2970 cut-down version of PCRE that is currently imported into Exim.
2972 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2973 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2974 patterns that start (?i) but I don't think that matters. The use of (?i) is
2975 just an example; this all applies to the other options as well.
2977 4. Since libtool seems to echo the compile commands it is issuing, the output
2978 from "make" can be reduced a bit by putting "@" in front of each libtool
2979 compile command.
2981 5. Patch from the folks at Google for configure.in to be a bit more thorough
2982 in checking for a suitable C++ installation before trying to compile the
2983 C++ stuff. This should fix a reported problem when a compiler was present,
2984 but no suitable headers.
2986 6. The man pages all had just "PCRE" as their title. I have changed them to
2987 be the relevant file name. I have also arranged that these names are
2988 retained in the file doc/pcre.txt, which is a concatenation in text format
2989 of all the man pages except the little individual ones for each function.
2991 7. The NON-UNIX-USE file had not been updated for the different set of source
2992 files that come with release 6. I also added a few comments about the C++
2993 wrapper.
2996 Version 6.0 07-Jun-05
2997 ---------------------
2999 1. Some minor internal re-organization to help with my DFA experiments.
3001 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3002 didn't matter for the library itself when fully configured, but did matter
3003 when compiling without UCP support, or within Exim, where the ucp files are
3004 not imported.
3006 3. Refactoring of the library code to split up the various functions into
3007 different source modules. The addition of the new DFA matching code (see
3008 below) to a single monolithic source would have made it really too
3009 unwieldy, quite apart from causing all the code to be include in a
3010 statically linked application, when only some functions are used. This is
3011 relevant even without the DFA addition now that patterns can be compiled in
3012 one application and matched in another.
3014 The downside of splitting up is that there have to be some external
3015 functions and data tables that are used internally in different modules of
3016 the library but which are not part of the API. These have all had their
3017 names changed to start with "_pcre_" so that they are unlikely to clash
3018 with other external names.
3020 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3021 a different (DFA) algorithm. Although it is slower than the original
3022 function, it does have some advantages for certain types of matching
3023 problem.
3025 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3026 including restarting after a partial match.
3028 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3029 defined when compiling for Windows was sent to me. I have put it into the
3030 code, though I have no means of testing or verifying it.
3032 7. Added the pcre_refcount() auxiliary function.
3034 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3035 match before or at the first newline in the subject string. In pcretest,
3036 the /f option on a pattern can be used to set this.
3038 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3039 would behave wrongly. This has been present in PCRE since release 4.0.
3041 10. A number of changes to the pcregrep command:
3043 (a) Refactored how -x works; insert ^(...)$ instead of setting
3044 PCRE_ANCHORED and checking the length, in preparation for adding
3045 something similar for -w.
3047 (b) Added the -w (match as a word) option.
3049 (c) Refactored the way lines are read and buffered so as to have more
3050 than one at a time available.
3052 (d) Implemented a pcregrep test script.
3054 (e) Added the -M (multiline match) option. This allows patterns to match
3055 over several lines of the subject. The buffering ensures that at least
3056 8K, or the rest of the document (whichever is the shorter) is available
3057 for matching (and similarly the previous 8K for lookbehind assertions).
3059 (f) Changed the --help output so that it now says
3061 -w, --word-regex(p)
3063 instead of two lines, one with "regex" and the other with "regexp"
3064 because that confused at least one person since the short forms are the
3065 same. (This required a bit of code, as the output is generated
3066 automatically from a table. It wasn't just a text change.)
3068 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3069 option but starts with a hyphen. Could be a pattern or a path name
3070 starting with a hyphen, for instance.
3072 (h) "-" can be given as a file name to represent stdin.
3074 (i) When file names are being printed, "(standard input)" is used for
3075 the standard input, for compatibility with GNU grep. Previously
3076 "<stdin>" was used.
3078 (j) The option --label=xxx can be used to supply a name to be used for
3079 stdin when file names are being printed. There is no short form.
3081 (k) Re-factored the options decoding logic because we are going to add
3082 two more options that take data. Such options can now be given in four
3083 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3085 (l) Added the -A, -B, and -C options for requesting that lines of context
3086 around matches be printed.
3088 (m) Added the -L option to print the names of files that do not contain
3089 any matching lines, that is, the complement of -l.
3091 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3092 continue to scan other files.
3094 (o) The -s option was incorrectly implemented. For compatibility with other
3095 greps, it now suppresses the error message for a non-existent or non-
3096 accessible file (but not the return code). There is a new option called
3097 -q that suppresses the output of matching lines, which was what -s was
3098 previously doing.
3100 (p) Added --include and --exclude options to specify files for inclusion
3101 and exclusion when recursing.
3103 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3104 Hopefully, it now does.
3106 12. Missing cast in pcre_study().
3108 13. Added an "uninstall" target to the makefile.
3110 14. Replaced "extern" in the function prototypes in Makefile.in with
3111 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3112 world, but is set differently for Windows.
3114 15. Added a second compiling function called pcre_compile2(). The only
3115 difference is that it has an extra argument, which is a pointer to an
3116 integer error code. When there is a compile-time failure, this is set
3117 non-zero, in addition to the error test pointer being set to point to an
3118 error message. The new argument may be NULL if no error number is required
3119 (but then you may as well call pcre_compile(), which is now just a
3120 wrapper). This facility is provided because some applications need a
3121 numeric error indication, but it has also enabled me to tidy up the way
3122 compile-time errors are handled in the POSIX wrapper.
3124 16. Added VPATH=.libs to the makefile; this should help when building with one
3125 prefix path and installing with another. (Or so I'm told by someone who
3126 knows more about this stuff than I do.)
3128 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3129 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3130 match everything, including newlines. This is not POSIX-compatible, but
3131 somebody wanted the feature. From pcretest it can be activated by using
3132 both the P and the s flags.
3134 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3136 19. libpcre.pc was being incorrectly installed as executable.
3138 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3139 it now also looks for '\r' so that it will work unmodified on Windows.
3141 21. Added Google's contributed C++ wrapper to the distribution.
3143 22. Added some untidy missing memory free() calls in pcretest, to keep
3144 Electric Fence happy when testing.
3148 Version 5.0 13-Sep-04
3149 ---------------------
3151 1. Internal change: literal characters are no longer packed up into items
3152 containing multiple characters in a single byte-string. Each character
3153 is now matched using a separate opcode. However, there may be more than one
3154 byte in the character in UTF-8 mode.
3156 2. The pcre_callout_block structure has two new fields: pattern_position and
3157 next_item_length. These contain the offset in the pattern to the next match
3158 item, and its length, respectively.
3160 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3161 insertion of callouts before each pattern item. Added the /C option to
3162 pcretest to make use of this.
3164 4. On the advice of a Windows user, the lines
3166 #if defined(_WIN32) || defined(WIN32)
3167 _setmode( _fileno( stdout ), 0x8000 );
3168 #endif /* defined(_WIN32) || defined(WIN32) */
3170 have been added to the source of pcretest. This apparently does useful
3171 magic in relation to line terminators.
3173 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3174 for the benefit of those environments where the "b" makes a difference.
3176 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3177 to know about it. I have put a hack into configure.in that adds in code
3178 to set GCC=yes if CC=icc. This seems to end up at a point in the
3179 generated configure script that is early enough to affect the setting of
3180 compiler options, which is what is needed, but I have no means of testing
3181 whether it really works. (The user who reported this had patched the
3182 generated configure script, which of course I cannot do.)
3184 LATER: After change 22 below (new libtool files), the configure script
3185 seems to know about icc (and also ecc). Therefore, I have commented out
3186 this hack in configure.in.
3188 7. Added support for pkg-config (2 patches were sent in).
3190 8. Negated POSIX character classes that used a combination of internal tables
3191 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3192 [[:^ascii]]. Typically, they would match almost any characters. The other
3193 POSIX classes were not broken in this way.
3195 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3196 to find the match, as PCRE was deluded into thinking that the match had to
3197 start at the start point or following a newline. The same bug applied to
3198 patterns with negative forward assertions or any backward assertions
3199 preceding ".*" at the start, unless the pattern required a fixed first
3200 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3202 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3203 starting at the last subject character, bytes beyond the end of the subject
3204 string were read.
3206 11. Renamed the variable "class" as "classbits" to make life easier for C++
3207 users. (Previously there was a macro definition, but it apparently wasn't
3208 enough.)
3210 12. Added the new field "tables" to the extra data so that tables can be passed
3211 in at exec time, or the internal tables can be re-selected. This allows
3212 a compiled regex to be saved and re-used at a later time by a different
3213 program that might have everything at different addresses.
3215 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3216 -R library as well as a -L library.
3218 14. The debugging options of pcretest (-d on the command line or D on a
3219 pattern) showed incorrect output for anything following an extended class
3220 that contained multibyte characters and which was followed by a quantifier.
3222 15. Added optional support for general category Unicode character properties
3223 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3224 support. It adds about 90K to the size of the library. The meanings of the
3225 inbuilt class escapes such as \d and \s have NOT been changed.
3227 16. Updated pcredemo.c to include calls to free() to release the memory for the
3228 compiled pattern.
3230 17. The generated file chartables.c was being created in the source directory
3231 instead of in the building directory. This caused the build to fail if the
3232 source directory was different from the building directory, and was
3233 read-only.
3235 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3236 file. No doubt somebody will tell me if they don't make sense... Also added
3237 Dan Mooney's comments about building on OpenVMS.
3239 19. Added support for partial matching via the PCRE_PARTIAL option for
3240 pcre_exec() and the \P data escape in pcretest.
3242 20. Extended pcretest with 3 new pattern features:
3244 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3245 write the compiled pattern to the file whose name is "rest-of-line".
3246 This is a straight binary dump of the data, with the saved pointer to
3247 the character tables forced to be NULL. The study data, if any, is
3248 written too. After writing, pcretest reads a new pattern.
3250 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3251 compiled pattern from the given file. There must not be any
3252 occurrences of "<" in the file name (pretty unlikely); if there are,
3253 pcretest will instead treat the initial "<" as a pattern delimiter.
3254 After reading in the pattern, pcretest goes on to read data lines as
3255 usual.
3257 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3258 and 16-bit fields in a compiled pattern, to simulate a pattern that
3259 was compiled on a host of opposite endianness.
3261 21. The pcre-exec() function can now cope with patterns that were compiled on
3262 hosts of opposite endianness, with this restriction:
3264 As for any compiled expression that is saved and used later, the tables
3265 pointer field cannot be preserved; the extra_data field in the arguments
3266 to pcre_exec() should be used to pass in a tables address if a value
3267 other than the default internal tables were used at compile time.
3269 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3270 now diagnosed as an error. Previously, most of the time, a negative number
3271 would have been treated as zero, but if in addition "ovector" was passed as
3272 NULL, a crash could occur.
3274 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3275 new versions from the libtool 1.5 distribution (the last one is a copy of
3276 a file called libtool.m4). This seems to have fixed the need to patch
3277 "configure" to support Darwin 1.3 (which I used to do). However, I still
3278 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3279 workstation).
3281 24. Changed the PCRE licence to be the more standard "BSD" licence.
3284 Version 4.5 01-Dec-03
3285 ---------------------
3287 1. There has been some re-arrangement of the code for the match() function so
3288 that it can be compiled in a version that does not call itself recursively.
3289 Instead, it keeps those local variables that need separate instances for
3290 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3291 needs to "recurse". Keeping track of where control must go is done by means
3292 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3293 hide most of the details from the main code, and operates only if
3294 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3295 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3296 operating.
3298 To make it easier for callers to provide specially tailored get/free
3299 functions for this usage, two new functions, pcre_stack_malloc, and
3300 pcre_stack_free, are used. They are always called in strict stacking order,
3301 and the size of block requested is always the same.
3303 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3304 PCRE has been compiled to use the stack or the heap for recursion. The
3305 -C option of pcretest uses this to show which version is compiled.
3307 A new data escape \S, is added to pcretest; it causes the amounts of store
3308 obtained and freed by both kinds of malloc/free at match time to be added
3309 to the output.
3311 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3312 what's available on my current Linux desktop machine.
3314 3. When matching a UTF-8 string, the test for a valid string at the start has
3315 been extended. If start_offset is not zero, PCRE now checks that it points
3316 to a byte that is the start of a UTF-8 character. If not, it returns
3317 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3318 this is necessary because there may be backward assertions in the pattern.
3319 When matching the same subject several times, it may save resources to use
3320 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3322 4. The code for checking the validity of UTF-8 strings has been tightened so
3323 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3324 containing "overlong sequences".
3326 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3327 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3328 should have been "&", but it just so happened that all the cases this let
3329 through by mistake were picked up later in the function.
3331 6. I had used a variable called "isblank" - this is a C99 function, causing
3332 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3334 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3335 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3336 so the version is shown; (c) stop "make test" echoing "./RunTest".
3338 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3340 9. The prototype for memmove() for systems that don't have it was using
3341 size_t, but the inclusion of the header that defines size_t was later. I've
3342 moved the #includes for the C headers earlier to avoid this.
3344 10. Added some adjustments to the code to make it easier to compiler on certain
3345 special systems:
3347 (a) Some "const" qualifiers were missing.
3348 (b) Added the macro EXPORT before all exported functions; by default this
3349 is defined to be empty.
3350 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3351 that it reads its output file name as an argument instead of writing
3352 to the standard output and assuming this can be redirected.
3354 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3355 class containing characters with values greater than 255, PCRE compilation
3356 went into a loop.
3358 12. A recursive reference to a subpattern that was within another subpattern
3359 that had a minimum quantifier of zero caused PCRE to crash. For example,
3360 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3361 recursion. If the recursively-called subpattern itself had a zero repeat,
3362 that was OK.
3364 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3365 buffer into which it was copied (for escape processing) was still set at
3366 1024, so long lines caused crashes.
3368 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3369 "internal error: code overflow...". This applied to any character class
3370 that was followed by a possessive quantifier.
3372 15. Modified the Makefile to add libpcre.la as a prerequisite for
3373 libpcreposix.la because I was told this is needed for a parallel build to
3374 work.
3376 16. If a pattern that contained .* following optional items at the start was
3377 studied, the wrong optimizing data was generated, leading to matching
3378 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3379 matching string must start with a or b or c. The correct conclusion for
3380 this pattern is that a match can start with any character.
3383 Version 4.4 13-Aug-03
3384 ---------------------
3386 1. In UTF-8 mode, a character class containing characters with values between
3387 127 and 255 was not handled correctly if the compiled pattern was studied.
3388 In fixing this, I have also improved the studying algorithm for such
3389 classes (slightly).
3391 2. Three internal functions had redundant arguments passed to them. Removal
3392 might give a very teeny performance improvement.
3394 3. Documentation bug: the value of the capture_top field in a callout is *one
3395 more than* the number of the hightest numbered captured substring.
3397 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3398 in incorrectly linking with a previously installed version. They now link
3399 explicitly with libpcre.la.
3401 5. configure.in no longer needs to recognize Cygwin specially.
3403 6. A problem in pcre.in for Windows platforms is fixed.
3405 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3406 pcretest, it used to include the size of the study block as part of its
3407 output. Unfortunately, the structure contains a field that has a different
3408 size on different hardware architectures. This meant that the tests that
3409 showed this size failed. As the block is currently always of a fixed size,
3410 this information isn't actually particularly useful in pcretest output, so
3411 I have just removed it.
3413 8. Three pre-processor statements accidentally did not start in column 1.
3414 Sadly, there are *still* compilers around that complain, even though
3415 standard C has not required this for well over a decade. Sigh.
3417 9. In pcretest, the code for checking callouts passed small integers in the
3418 callout_data field, which is a void * field. However, some picky compilers
3419 complained about the casts involved for this on 64-bit systems. Now
3420 pcretest passes the address of the small integer instead, which should get
3421 rid of the warnings.
3423 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3424 both compile and run time, and gives an error if an invalid UTF-8 sequence
3425 is found. There is a option for disabling this check in cases where the
3426 string is known to be correct and/or the maximum performance is wanted.
3428 11. In response to a bug report, I changed one line in Makefile.in from
3430 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3431 to
3432 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3434 to look similar to other lines, but I have no way of telling whether this
3435 is the right thing to do, as I do not use Windows. No doubt I'll get told
3436 if it's wrong...
3439 Version 4.3 21-May-03
3440 ---------------------
3442 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3443 Makefile.
3445 2. Some refactoring to improve the quality of the code:
3447 (i) The utf8_table... variables are now declared "const".
3449 (ii) The code for \cx, which used the "case flipping" table to upper case
3450 lower case letters, now just substracts 32. This is ASCII-specific,
3451 but the whole concept of \cx is ASCII-specific, so it seems
3452 reasonable.
3454 (iii) PCRE was using its character types table to recognize decimal and
3455 hexadecimal digits in the pattern. This is silly, because it handles
3456 only 0-9, a-f, and A-F, but the character types table is locale-
3457 specific, which means strange things might happen. A private
3458 table is now used for this - though it costs 256 bytes, a table is
3459 much faster than multiple explicit tests. Of course, the standard
3460 character types table is still used for matching digits in subject
3461 strings against \d.
3463 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3464 ending in _t are). So I've renamed it as ESC_tee.
3466 3. The first argument for regexec() in the POSIX wrapper should have been
3467 defined as "const".
3469 4. Changed pcretest to use malloc() for its buffers so that they can be
3470 Electric Fenced for debugging.
3472 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3473 to read one or more bytes before the start of the subject string. Often this
3474 had no effect on PCRE's behaviour, but in some circumstances it could
3475 provoke a segmentation fault.
3477 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3478 to try to read one or more bytes before the start of the subject string.
3480 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3481 UTF-8 support could misbehave in various ways if the subject string
3482 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3483 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3484 back over UTF-8 characters.)
3487 Version 4.2 14-Apr-03
3488 ---------------------
3490 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3492 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3493 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3494 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3495 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3496 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3498 Note: automatic setting of the BUILD variables is not yet working
3499 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3500 compile-time but not at link-time
3501 [LINK]: use for linking executables only
3502 make different versions for Windows and non-Windows
3503 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3504 libraries
3505 [LINK_FOR_BUILD]: new variable
3506 [OBJEXT]: use throughout
3507 [EXEEXT]: use throughout
3508 <winshared>: new target
3509 <wininstall>: new target
3510 <dftables.o>: use native compiler
3511 <dftables>: use native linker
3512 <install>: handle Windows platform correctly
3513 <clean>: ditto
3514 <check>: ditto
3515 copy DLL to top builddir before testing
3517 As part of these changes, -no-undefined was removed again. This was reported
3518 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3519 in any case.
3521 3. Some tidies to get rid of compiler warnings:
3523 . In the match_data structure, match_limit was an unsigned long int, whereas
3524 match_call_count was an int. I've made them both unsigned long ints.
3526 . In pcretest the fact that a const uschar * doesn't automatically cast to
3527 a void * provoked a warning.
3529 . Turning on some more compiler warnings threw up some "shadow" variables
3530 and a few more missing casts.
3532 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3533 option, a class that contained a single character with a value between 128
3534 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3536 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3537 option, a class that contained several characters, but with at least one
3538 whose value was between 128 and 255 caused PCRE to crash.
3541 Version 4.1 12-Mar-03
3542 ---------------------
3544 1. Compiling with gcc -pedantic found a couple of places where casts were
3545 needed, and a string in dftables.c that was longer than standard compilers are
3546 required to support.
3548 2. Compiling with Sun's compiler found a few more places where the code could
3549 be tidied up in order to avoid warnings.
3551 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3552 first of these names is deprecated in the latest Autoconf in favour of the name
3553 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3554 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3555 analogy I have changed it to CFLAGS_FOR_BUILD.
3557 4. Added -no-undefined to the linking command in the Makefile, because this is
3558 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3559 linking step for the pcreposix library.
3561 5. PCRE was failing to diagnose the case of two named groups with the same
3562 name.
3564 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3565 literal character that is needed in the subject for a match, and scans along to
3566 ensure that it is present before embarking on the full matching process. This
3567 saves time in cases of nested unlimited repeats that are never going to match.
3568 Problem: the scan can take a lot of time if the subject is very long (e.g.
3569 megabytes), thus penalizing straightforward matches. It is now done only if the
3570 amount of subject to be scanned is less than 1000 bytes.
3572 7. A lesser problem with the same optimization is that it was recording the
3573 first character of an anchored pattern as "needed", thus provoking a search
3574 right along the subject, even when the first match of the pattern was going to
3575 fail. The "needed" character is now not set for anchored patterns, unless it
3576 follows something in the pattern that is of non-fixed length. Thus, it still
3577 fulfils its original purpose of finding quick non-matches in cases of nested
3578 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3581 Version 4.0 17-Feb-03
3582 ---------------------
3584 1. If a comment in an extended regex that started immediately after a meta-item
3585 extended to the end of string, PCRE compiled incorrect data. This could lead to
3586 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3588 2. Moved to autoconf 2.53 and libtool 1.4.2.
3590 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3591 the special perltest8 script is no longer needed - all the tests can be run
3592 from a single perltest script.
3594 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3595 by \s. It has now been removed in PCRE. This means it isn't recognized as
3596 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3597 class [:space:] *does* include VT, thereby creating a mess.
3599 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3600 space and tab.
3602 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3603 its new features into the main test script, reducing the number of scripts.
3605 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3606 were backward compatible, and made the (?i) apply to the whole pattern, as if
3607 /i were given. Now it behaves more logically, and applies the option setting
3608 only to what follows. PCRE has been changed to follow suit. However, if it
3609 finds options settings right at the start of the pattern, it extracts them into
3610 the global options, as before. Thus, they show up in the info data.
3612 8. Added support for the \Q...\E escape sequence. Characters in between are
3613 treated as literals. This is slightly different from Perl in that $ and @ are
3614 also handled as literals inside the quotes. In Perl, they will cause variable
3615 interpolation. Note the following examples:
3617 Pattern PCRE matches Perl matches
3619 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3620 \Qabc\$xyz\E abc\$xyz abc\$xyz
3621 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3623 For compatibility with Perl, \Q...\E sequences are recognized inside character
3624 classes as well as outside them.
3626 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3627 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3628 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3629 signed/unsigned warnings.
3631 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3632 option for pcretest, so I've replaced it by a simple function that does just
3633 that job.
3635 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3636 "pcregrep -".
3638 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3639 Java package. This provides some syntactic sugar for simple cases of what my
3640 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3641 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3642 item, you can use this simplified notation. Note that only makes sense with
3643 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3644 greediness, whatever the setting of the PCRE_UNGREEDY option.
3646 13. A change of greediness default within a pattern was not taking effect at
3647 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3648 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3649 was abstracted outside.
3651 14. PCRE now supports the \G assertion. It is true when the current matching
3652 position is at the start point of the match. This differs from \A when the
3653 starting offset is non-zero. Used with the /g option of pcretest (or similar
3654 code), it works in the same way as it does for Perl's /g option. If all
3655 alternatives of a regex begin with \G, the expression is anchored to the start
3656 match position, and the "anchored" flag is set in the compiled expression.
3658 15. Some bugs concerning the handling of certain option changes within patterns
3659 have been fixed. These applied to options other than (?ims). For example,
3660 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3661 been the other way round. Some of this was related to change 7 above.
3663 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3664 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3665 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3666 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3668 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3669 mode. Unlike ".", it always matches newline, whatever the setting of
3670 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3671 assertions. Perl allows it, but it doesn't (in general) work because it can't
3672 calculate the length of the lookbehind. At least, that's the case for Perl
3673 5.8.0 - I've been told they are going to document that it doesn't work in
3674 future.
3676 18. Added an error diagnosis for escapes that PCRE does not support: these are
3677 \L, \l, \N, \P, \p, \U, \u, and \X.
3679 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3680 reading past the end of the pattern in cases such as /[abcd/.
3682 20. PCRE was getting more memory than necessary for patterns with classes that
3683 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3685 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3686 compiling PCRE for use with Virtual Pascal.
3688 22. Small fix to the Makefile to make it work properly if the build is done
3689 outside the source tree.
3691 23. Added a new extension: a condition to go with recursion. If a conditional
3692 subpattern starts with (?(R) the "true" branch is used if recursion has
3693 happened, whereas the "false" branch is used only at the top level.
3695 24. When there was a very long string of literal characters (over 255 bytes
3696 without UTF support, over 250 bytes with UTF support), the computation of how
3697 much memory was required could be incorrect, leading to segfaults or other
3698 strange effects.
3700 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3701 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3702 there was a subsequent back reference to those brackets. This meant that, for
3703 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3704 possible to check for precisely this case. All we can do is abandon the
3705 optimization if .* occurs inside capturing brackets when there are any back
3706 references whatsoever. (See below for a better fix that came later.)
3708 26. The handling of the optimization for finding the first character of a
3709 non-anchored pattern, and for finding a character that is required later in the
3710 match were failing in some cases. This didn't break the matching; it just
3711 failed to optimize when it could. The way this is done has been re-implemented.
3713 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3715 28. Added a new feature that provides some of the functionality that Perl
3716 provides with (?{...}). The facility is termed a "callout". The way it is done
3717 in PCRE is for the caller to provide an optional function, by setting
3718 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3719 global variable. By default it is unset, which disables all calling out. To get
3720 the function called, the regex must include (?C) at appropriate points. This
3721 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3722 This provides a means of identifying different callout points. When PCRE
3723 reaches such a point in the regex, if pcre_callout has been set, the external
3724 function is called. It is provided with data in a structure called
3725 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3726 matching continues; if it returns a non-zero value, the match at the current
3727 point fails. However, backtracking will occur if possible. [This was changed
3728 later and other features added - see item 49 below.]
3730 29. pcretest is upgraded to test the callout functionality. It provides a
3731 callout function that displays information. By default, it shows the start of
3732 the match and the current position in the text. There are some new data escapes
3733 to vary what happens:
3735 \C+ in addition, show current contents of captured substrings
3736 \C- do not supply a callout function
3737 \C!n return 1 when callout number n is reached
3738 \C!n!m return 1 when callout number n is reached for the mth time
3740 30. If pcregrep was called with the -l option and just a single file name, it
3741 output "<stdin>" if a match was found, instead of the file name.
3743 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3744 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3745 pcre_exec(). This saves a malloc/free per call. The default value of
3746 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3747 when configuring.
3749 32. The default maximum size of a compiled pattern is 64K. There have been a
3750 few cases of people hitting this limit. The code now uses macros to handle the
3751 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3752 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3753 configuring. Tests 2 and 5 work only with 2-byte links because they output
3754 debugging information about compiled patterns.
3756 33. Internal code re-arrangements:
3758 (a) Moved the debugging function for printing out a compiled regex into
3759 its own source file (printint.c) and used #include to pull it into
3760 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3761 separate copies.
3763 (b) Defined the list of op-code names for debugging as a macro in
3764 internal.h so that it is next to the definition of the opcodes.
3766 (c) Defined a table of op-code lengths for simpler skipping along compiled
3767 code. This is again a macro in internal.h so that it is next to the
3768 definition of the opcodes.
3770 34. Added support for recursive calls to individual subpatterns, along the
3771 lines of Robin Houston's patch (but implemented somewhat differently).
3773 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3774 allow it to read and process whole directories in Win32. This code was
3775 contributed by Lionel Fourquaux; it has not been tested by me.
3777 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3778 used to name a group. Names consist of alphanumerics and underscores, and must
3779 be unique. Back references use the syntax (?P=name) and recursive calls use
3780 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3781 numbers. The function pcre_fullinfo() can be used after compilation to extract
3782 a name/number map. There are three relevant calls:
3784 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3785 PCRE_INFO_NAMECOUNT yields the number of entries
3786 PCRE_INFO_NAMETABLE yields a pointer to the map.
3788 The map is a vector of fixed-size entries. The size of each entry depends on
3789 the length of the longest name used. The first two bytes of each entry are the
3790 group number, most significant byte first. There follows the corresponding
3791 name, zero terminated. The names are in alphabetical order.
3793 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3794 case instead of 255. Making it the same both with and without UTF-8 support
3795 means that the same test output works with both.
3797 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3798 calling malloc() with a zero argument.
3800 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3801 optimization. I've improved things by keeping a bitmap of backreferences with
3802 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3803 fact referenced, the optimization can be applied. It is unlikely that a
3804 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3805 the match to follow \n) will appear inside brackets with a number greater than
3806 31, but if it does, any back reference > 31 suppresses the optimization.
3808 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3809 of disabling numbered capturing parentheses. Any opening parenthesis that is
3810 not followed by ? behaves as if it were followed by ?: but named parentheses
3811 can still be used for capturing (and they will acquire numbers in the usual
3812 way).
3814 41. Redesigned the return codes from the match() function into yes/no/error so
3815 that errors can be passed back from deep inside the nested calls. A malloc
3816 failure while inside a recursive subpattern call now causes the
3817 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3819 42. It is now possible to set a limit on the number of times the match()
3820 function is called in a call to pcre_exec(). This facility makes it possible to
3821 limit the amount of recursion and backtracking, though not in a directly
3822 obvious way, because the match() function is used in a number of different
3823 circumstances. The count starts from zero for each position in the subject
3824 string (for non-anchored patterns). The default limit is, for compatibility, a
3825 large number, namely 10 000 000. You can change this in two ways:
3827 (a) When configuring PCRE before making, you can use --with-match-limit=n
3828 to set a default value for the compiled library.
3830 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3831 a different value is set. See 45 below.
3833 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3835 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3836 of things that can be changed at compile time. The first argument specifies
3837 what is wanted and the second points to where the information is to be placed.
3838 The current list of available information is:
3842 The output is an integer that is set to one if UTF-8 support is available;
3843 otherwise it is set to zero.
3847 The output is an integer that it set to the value of the code that is used for
3848 newline. It is either LF (10) or CR (13).
3852 The output is an integer that contains the number of bytes used for internal
3853 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3857 The output is an integer that contains the threshold above which the POSIX
3858 interface uses malloc() for output vectors. See item 31 above.
3862 The output is an unsigned integer that contains the default limit of the number
3863 of match() calls in a pcre_exec() execution. See 42 above.
3865 44. pcretest has been upgraded by the addition of the -C option. This causes it
3866 to extract all the available output from the new pcre_config() function, and to
3867 output it. The program then exits immediately.
3869 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3870 order to support additional features. One way would have been to define
3871 pcre_exec2() (for example) with extra arguments, but this would not have been
3872 extensible, and would also have required all calls to the original function to
3873 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3874 is used for passing in "extra" data from pcre_study().
3876 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3877 contains the following fields:
3879 flags a bitmap indicating which of the following fields are set
3880 study_data opaque data from pcre_study()
3881 match_limit a way of specifying a limit on match() calls for a specific
3882 call to pcre_exec()
3883 callout_data data for callouts (see 49 below)
3885 The flag bits are also defined in pcre.h, and are
3891 The pcre_study() function now returns one of these new pcre_extra blocks, with
3892 the actual study data pointed to by the study_data field, and the
3893 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3894 before. That is, this change is entirely upwards-compatible and requires no
3895 change to existing code.
3897 If you want to pass in additional data to pcre_exec(), you can either place it
3898 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3899 block.
3901 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3902 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3903 times with different match limits, until it finds the minimum value needed for
3904 pcre_exec() to complete. The value is then output. This can be instructive; for
3905 most simple matches the number is quite small, but for pathological cases it
3906 gets very large very quickly.
3908 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3909 returns the size of the data block pointed to by the study_data field in a
3910 pcre_extra block, that is, the value that was passed as the argument to
3911 pcre_malloc() when PCRE was getting memory in which to place the information
3912 created by pcre_study(). The fourth argument should point to a size_t variable.
3913 pcretest has been extended so that this information is shown after a successful
3914 pcre_study() call when information about the compiled regex is being displayed.
3916 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3917 because what follows is always an absolute path. (Later: it turns out that this
3918 is more than cosmetic for MinGW, because it doesn't like empty path
3919 components.)
3921 49. Some changes have been made to the callout feature (see 28 above):
3923 (i) A callout function now has three choices for what it returns:
3925 0 => success, carry on matching
3926 > 0 => failure at this point, but backtrack if possible
3927 < 0 => serious error, return this value from pcre_exec()
3929 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3930 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3931 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3932 use by callout functions. It will never be used by PCRE itself.
3934 (ii) The pcre_extra structure (see 45 above) has a void * field called
3935 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3936 pcre_callout_block structure has a field of the same name. The contents of
3937 the field passed in the pcre_extra structure are passed to the callout
3938 function in the corresponding field in the callout block. This makes it
3939 easier to use the same callout-containing regex from multiple threads. For
3940 testing, the pcretest program has a new data escape
3942 \C*n pass the number n (may be negative) as callout_data
3944 If the callout function in pcretest receives a non-zero value as
3945 callout_data, it returns that value.
3947 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3948 there were some redundant $(CFLAGS) in commands that are now specified as
3949 $(LINK), which already includes $(CFLAGS).
3951 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3952 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3953 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3954 one-byte characters throughout. Note that case-insensitive matching applies
3955 only to characters whose values are less than 256. PCRE doesn't support the
3956 notion of cases for higher-valued characters.
3958 (i) A character class whose characters are all within 0-255 is handled as
3959 a bit map, and the map is inverted for negative classes. Previously, a
3960 character > 255 always failed to match such a class; however it should
3961 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3963 (ii) A negated character class with a single character < 255 is coded as
3964 "not this character" (OP_NOT). This wasn't working properly when the test
3965 character was multibyte, either singly or repeated.
3967 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3968 mode, for example: \x{100}{2,3}.
3970 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3971 singly or repeated) now correctly test multibyte characters. However,
3972 PCRE doesn't recognize any characters with values greater than 255 as
3973 digits, spaces, or word characters. Such characters always match \D, \S,
3974 and \W, and never match \d, \s, or \w.
3976 (v) Classes may now contain characters and character ranges with values
3977 greater than 255. For example: [ab\x{100}-\x{400}].
3979 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3980 PCRE in UTF-8 mode.
3982 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3983 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3984 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3985 value.)
3987 53. The single man page has become too large. I have therefore split it up into
3988 a number of separate man pages. These also give rise to individual HTML pages;
3989 these are now put in a separate directory, and there is an index.html page that
3990 lists them all. Some hyperlinking between the pages has been installed.
3992 54. Added convenience functions for handling named capturing parentheses.
3994 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3995 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3996 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3997 are faulted.
3999 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4000 calling configure. These values are used when compiling the dftables.c program
4001 which is run to generate the source of the default character tables. They
4002 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4003 you will need to set these values.
4005 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4008 Version 3.9 02-Jan-02
4009 ---------------------
4011 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4013 2. If --disable-static was given, the building process failed when trying to
4014 build pcretest and pcregrep. (For some reason it was using libtool to compile
4015 them, which is not right, as they aren't part of the library.)
4018 Version 3.8 18-Dec-01
4019 ---------------------
4021 1. The experimental UTF-8 code was completely screwed up. It was packing the
4022 bytes in the wrong order. How dumb can you get?
4025 Version 3.7 29-Oct-01
4026 ---------------------
4028 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4029 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4030 this didn't happen under Solaris 8, where I normally test things.
4032 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4033 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4034 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4035 this fix an hour or so after the initial 3.7 release.)
4038 Version 3.6 23-Oct-01
4039 ---------------------
4041 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4042 offsets passed as NULL with zero offset count.
4044 2. The config.guess and config.sub files had not been updated when I moved to
4045 the latest autoconf.
4048 Version 3.5 15-Aug-01
4049 ---------------------
4051 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4052 had been forgotten.
4054 2. By using declared but undefined structures, we can avoid using "void"
4055 definitions in pcre.h while keeping the internal definitions of the structures
4056 private.
4058 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4059 user point of view, this means that both static and shared libraries are built
4060 by default, but this can be individually controlled. More of the work of
4061 handling this static/shared cases is now inside libtool instead of PCRE's make
4062 file.
4064 4. The pcretest utility is now installed along with pcregrep because it is
4065 useful for users (to test regexs) and by doing this, it automatically gets
4066 relinked by libtool. The documentation has been turned into a man page, so
4067 there are now .1, .txt, and .html versions in /doc.
4069 5. Upgrades to pcregrep:
4070 (i) Added long-form option names like gnu grep.
4071 (ii) Added --help to list all options with an explanatory phrase.
4072 (iii) Added -r, --recursive to recurse into sub-directories.
4073 (iv) Added -f, --file to read patterns from a file.
4075 6. pcre_exec() was referring to its "code" argument before testing that
4076 argument for NULL (and giving an error if it was NULL).
4078 7. Upgraded Makefile.in to allow for compiling in a different directory from
4079 the source directory.
4081 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4082 options bits, the pointer it was passed was to an int instead of to an unsigned
4083 long int. This mattered only on 64-bit systems.
4085 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4086 generated) instead of pcre.in, which it its source. Also made the same change
4087 in several of the .c files.
4089 10. A new release of gcc defines printf() as a macro, which broke pcretest
4090 because it had an ifdef in the middle of a string argument for printf(). Fixed
4091 by using separate calls to printf().
4093 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4094 script, to force use of CR or LF instead of \n in the source. On non-Unix
4095 systems, the value can be set in config.h.
4097 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4098 absolute limit. Changed the text of the error message to make this clear, and
4099 likewise updated the man page.
4101 13. The limit of 99 on the number of capturing subpatterns has been removed.
4102 The new limit is 65535, which I hope will not be a "real" limit.
4105 Version 3.4 22-Aug-00
4106 ---------------------
4108 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4110 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4113 Version 3.3 01-Aug-00
4114 ---------------------
4116 1. If an octal character was given, but the value was greater than \377, it
4117 was not getting masked to the least significant bits, as documented. This could
4118 lead to crashes in some systems.
4120 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4121 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4123 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4124 These just pass their arguments on to (pcre_free)(), but they are provided
4125 because some uses of PCRE bind it to non-C systems that can call its functions,
4126 but cannot call free() or pcre_free() directly.
4128 4. Add "make test" as a synonym for "make check". Corrected some comments in
4129 the Makefile.
4131 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4132 Makefile.
4134 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4135 command called pgrep for grepping around the active processes.
4137 7. Added the beginnings of support for UTF-8 character strings.
4139 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4140 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4141 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4142 out for the ar command.)
4145 Version 3.2 12-May-00
4146 ---------------------
4148 This is purely a bug fixing release.
4150 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4151 of ZA. This was just one example of several cases that could provoke this bug,
4152 which was introduced by change 9 of version 2.00. The code for breaking
4153 infinite loops after an iteration that matches an empty string was't working
4154 correctly.
4156 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4157 when matched against abbab (for example). After matching an empty string, it
4158 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4159 caused it to match further down the string than it should.
4161 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4162 was there because it doesn't seem to be needed, and it causes trouble on some
4163 systems, as it is not a Standard C header. It has been removed.
4165 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4166 were reported on the Macintosh. The changes were from
4168 while ((c = *(++ptr)) != 0 && c != '\n');
4169 to
4170 while ((c = *(++ptr)) != 0 && c != '\n') ;
4172 Totally extraordinary, but if that's what it takes...
4174 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4175 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4176 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4177 assumes the way PCRE uses memmove() (always moving upwards).
4179 6. PCRE is being used in one environment where strchr() is not available. There
4180 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4181 faster code anyway.
4184 Version 3.1 09-Feb-00
4185 ---------------------
4187 The only change in this release is the fixing of some bugs in Makefile.in for
4188 the "install" target:
4190 (1) It was failing to install pcreposix.h.
4192 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4195 Version 3.0 01-Feb-00
4196 ---------------------
4198 1. Add support for the /+ modifier to perltest (to output $` like it does in
4199 pcretest).
4201 2. Add support for the /g modifier to perltest.
4203 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4204 matches null strings.
4206 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4207 pattern. Perl treats empty patterns specially - it reuses the most recent
4208 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4209 effect.
4211 5. The POSIX interface was broken in that it was just handing over the POSIX
4212 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4213 required a bigger vector, with some working space on the end. This means that
4214 the POSIX wrapper now has to get and free some memory, and copy the results.
4216 6. Added some simple autoconf support, placing the test data and the
4217 documentation in separate directories, re-organizing some of the
4218 information files, and making it build pcre-config (a GNU standard). Also added
4219 libtool support for building PCRE as a shared library, which is now the
4220 default.
4222 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4223 09 are not valid octal constants. Single digits will be used for minor values
4224 less than 10.
4226 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4227 existing programs that set these in the POSIX interface can use PCRE without
4228 modification.
4230 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4231 return all that pcre_info() returns, plus additional data. The pcre_info()
4232 function is retained for compatibility, but is considered to be obsolete.
4234 10. Added experimental recursion feature (?R) to handle one common case that
4235 Perl 5.6 will be able to do with (?p{...}).
4237 11. Added support for POSIX character classes like [:alpha:], which Perl is
4238 adopting.
4241 Version 2.08 31-Aug-99
4242 ----------------------
4244 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4245 trying to match at the startoffset position, but instead was moving forward to
4246 the next newline as if a previous match had failed.
4248 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4249 and could get into a loop if a null string was matched other than at the start
4250 of the subject.
4252 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4253 be distinguished at compile time, and for completeness also added PCRE_DATE.
4255 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4256 in GnuWin32 environments.
4259 Version 2.07 29-Jul-99
4260 ----------------------
4262 1. The documentation is now supplied in plain text form and HTML as well as in
4263 the form of man page sources.
4265 2. C++ compilers don't like assigning (void *) values to other pointer types.
4266 In particular this affects malloc(). Although there is no problem in Standard
4267 C, I've put in casts to keep C++ compilers happy.
4269 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4270 should be (const char *).
4272 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4273 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4274 However, I haven't made this a standard facility. The documentation doesn't
4275 mention it, and the Makefile doesn't support it.
4277 5. The Makefile now contains an "install" target, with editable destinations at
4278 the top of the file. The pcretest program is not installed.
4280 6. pgrep -V now gives the PCRE version number and date.
4282 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4283 causing the entire string to be ignored, instead of just the last character.
4285 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4286 non-matching string, it can take a very, very long time, even for strings of
4287 quite modest length, because of the nested recursion. PCRE now does better in
4288 some of these cases. It does this by remembering the last required literal
4289 character in the pattern, and pre-searching the subject to ensure it is present
4290 before running the real match. In other words, it applies a heuristic to detect
4291 some types of certain failure quickly, and in the above example, if presented
4292 with a string that has no trailing " it gives "no match" very quickly.
4294 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4295 other alternatives are tried instead.
4298 Version 2.06 09-Jun-99
4299 ----------------------
4301 1. Change pcretest's output for amount of store used to show just the code
4302 space, because the remainder (the data block) varies in size between 32-bit and
4303 64-bit systems.
4305 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4306 start matching at. This allows lookbehinds to work when searching for multiple
4307 occurrences in a string.
4309 3. Added additional options to pcretest for testing multiple occurrences:
4311 /+ outputs the rest of the string that follows a match
4312 /g loops for multiple occurrences, using the new startoffset argument
4313 /G loops for multiple occurrences by passing an incremented pointer
4315 4. PCRE wasn't doing the "first character" optimization for patterns starting
4316 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4317 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4318 the letter 'x'. On long subject strings, this gives a significant speed-up.
4321 Version 2.05 21-Apr-99
4322 ----------------------
4324 1. Changed the type of magic_number from int to long int so that it works
4325 properly on 16-bit systems.
4327 2. Fixed a bug which caused patterns starting with .* not to work correctly
4328 when the subject string contained newline characters. PCRE was assuming
4329 anchoring for such patterns in all cases, which is not correct because .* will
4330 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4331 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4332 must be retried after every newline in the subject.
4335 Version 2.04 18-Feb-99
4336 ----------------------
4338 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4339 computation of the store needed to hold the pattern was incorrect (too large).
4340 If such patterns were nested a few deep, this could multiply and become a real
4341 problem.
4343 2. Added /M option to pcretest to show the memory requirement of a specific
4344 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4346 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4347 compiled in such a way that the backtracking after subsequent failure was
4348 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4349 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4352 Version 2.03 02-Feb-99
4353 ----------------------
4355 1. Fixed typo and small mistake in man page.
4357 2. Added 4th condition (GPL supersedes if conflict) and created separate
4358 LICENCE file containing the conditions.
4360 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4361 Perl, that is the internal \ allows the delimiter to be included in the
4362 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4363 the final delimiter, add \ to the end of the pattern (to test the error).
4365 4. Added the convenience functions for extracting substrings after a successful
4366 match. Updated pcretest to make it able to test these functions.
4369 Version 2.02 14-Jan-99
4370 ----------------------
4372 1. Initialized the working variables associated with each extraction so that
4373 their saving and restoring doesn't refer to uninitialized store.
4375 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4376 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4377 fix the problem.
4379 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4380 calls, and wasn't printing the correct value for compiling calls. Increased the
4381 default value of LOOPREPEAT, and the number of significant figures in the
4382 times.
4384 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4386 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4387 a building problem on Windows NT with a FAT file system.
4390 Version 2.01 21-Oct-98
4391 ----------------------
4393 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4394 to character tables built by pcre_maketables() in the current locale. If NULL
4395 is passed, the default tables are used.
4398 Version 2.00 24-Sep-98
4399 ----------------------
4401 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4402 it any more.
4404 2. Allow quantification of (?>) groups, and make it work correctly.
4406 3. The first character computation wasn't working for (?>) groups.
4408 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4409 end of the subject) and add 5.005's \z, which really does match only at the
4410 very end of the subject.
4412 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4414 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4415 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4416 localized options. All options to pcre_study() were also removed.
4418 7. Add other new features from 5.005:
4420 $(?<= positive lookbehind
4421 $(?<! negative lookbehind
4422 (?imsx-imsx) added the unsetting capability
4423 such a setting is global if at outer level; local otherwise
4424 (?imsx-imsx:) non-capturing groups with option setting
4425 (?(cond)re|re) conditional pattern matching
4427 A backreference to itself in a repeated group matches the previous
4428 captured string.
4430 8. General tidying up of studying (both automatic and via "study")
4431 consequential on the addition of new assertions.
4433 9. As in 5.005, unlimited repeated groups that could match an empty substring
4434 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4435 runtime if any iteration does actually match an empty substring.
4437 10. Include the RunTest script in the distribution.
4439 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4440 discrepancies, some of which were old and were also with respect to 5.004. They
4441 have now been fixed.
4444 Version 1.09 28-Apr-98
4445 ----------------------
4447 1. A negated single character class followed by a quantifier with a minimum
4448 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4449 program crashes, or just wrong answers. This did not apply to negated classes
4450 containing more than one character, or to minima other than one.
4453 Version 1.08 27-Mar-98
4454 ----------------------
4456 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4458 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4459 latter must appear before anything that relies on it in the pattern.
4462 Version 1.07 16-Feb-98
4463 ----------------------
4465 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4466 repeat of a potentially empty string).
4469 Version 1.06 23-Jan-98
4470 ----------------------
4472 1. Added Markus Oberhumer's little patches for C++.
4474 2. Literal strings longer than 255 characters were broken.
4477 Version 1.05 23-Dec-97
4478 ----------------------
4480 1. Negated character classes containing more than one character were failing if
4481 PCRE_CASELESS was set at run time.
4484 Version 1.04 19-Dec-97
4485 ----------------------
4487 1. Corrected the man page, where some "const" qualifiers had been omitted.
4489 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4490 input syntax.
4492 3. Fixed memory leak which occurred when a regex with back references was
4493 matched with an offsets vector that wasn't big enough. The temporary memory
4494 that is used in this case wasn't being freed if the match failed.
4496 4. Tidied pcretest to ensure it frees memory that it gets.
4498 5. Temporary memory was being obtained in the case where the passed offsets
4499 vector was exactly big enough.
4501 6. Corrected definition of offsetof() from change 5 below.
4503 7. I had screwed up change 6 below and broken the rules for the use of
4504 setjmp(). Now fixed.
4507 Version 1.03 18-Dec-97
4508 ----------------------
4510 1. A erroneous regex with a missing opening parenthesis was correctly
4511 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4512 on some systems.
4514 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4515 it was reported that one broken compiler failed on the former because "code" is
4516 also an independent variable.
4518 3. The erroneous regex a[]b caused an array overrun reference.
4520 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4521 fail on data ending with that character. (It was going on too far, and checking
4522 the next character, typically a binary zero.) This was specific to the
4523 optimized code for single-character negative classes.
4525 5. Added a contributed patch from the TIN world which does the following:
4527 + Add an undef for memmove, in case the the system defines a macro for it.
4529 + Add a definition of offsetof(), in case there isn't one. (I don't know
4530 the reason behind this - offsetof() is part of the ANSI standard - but
4531 it does no harm).
4533 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4534 most of the places where whitespace preceded '#'. I have given up and
4535 allowed the remaining 2 cases to be at the margin.
4537 + Rename some variables in pcre to eliminate shadowing. This seems very
4538 pedantic, but does no harm, of course.
4540 6. Moved the call to setjmp() into its own function, to get rid of warnings
4541 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4543 7. Constructs such as \d{8,} were compiling into the equivalent of
4544 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4545 outcome, but in this particular case used more store than had been allocated,
4546 which caused the bug to be discovered because it threw up an internal error.
4548 8. The debugging code in both pcre and pcretest for outputting the compiled
4549 form of a regex was going wrong in the case of back references followed by
4550 curly-bracketed repeats.
4553 Version 1.02 12-Dec-97
4554 ----------------------
4556 1. Typos in pcre.3 and comments in the source fixed.
4558 2. Applied a contributed patch to get rid of places where it used to remove
4559 'const' from variables, and fixed some signed/unsigned and uninitialized
4560 variable warnings.
4562 3. Added the "runtest" target to Makefile.
4564 4. Set default compiler flag to -O2 rather than just -O.
4567 Version 1.01 19-Nov-97
4568 ----------------------
4570 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4571 like /([ab]*)*/, that is, for classes with more than one character in them.
4573 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4574 as /((?>a*))*/ (a PCRE_EXTRA facility).
4577 Version 1.00 18-Nov-97
4578 ----------------------
4580 1. Added compile-time macros to support systems such as SunOS4 which don't have
4581 memmove() or strerror() but have other things that can be used instead.
4583 2. Arranged that "make clean" removes the executables.
4586 Version 0.99 27-Oct-97
4587 ----------------------
4589 1. Fixed bug in code for optimizing classes with only one character. It was
4590 initializing a 32-byte map regardless, which could cause it to run off the end
4591 of the memory it had got.
4593 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4596 Version 0.98 22-Oct-97
4597 ----------------------
4599 1. Fixed bug in code for handling temporary memory usage when there are more
4600 back references than supplied space in the ovector. This could cause segfaults.
4603 Version 0.97 21-Oct-97
4604 ----------------------
4606 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4608 2. Optimized negated single characters not to use a bit map.
4610 3. Brought error texts together as macro definitions; clarified some of them;
4611 fixed one that was wrong - it said "range out of order" when it meant "invalid
4612 escape sequence".
4614 4. Changed some char * arguments to const char *.
4616 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4618 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4619 pcretest.
4622 Version 0.96 16-Oct-97
4623 ----------------------
4625 1. Added a simple "pgrep" utility to the distribution.
4627 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4628 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4629 where "ddd" means "one or more decimal digits".
4631 3. Fixed serious bug. If a pattern had a back reference, but the call to
4632 pcre_exec() didn't supply a large enough ovector to record the related
4633 identifying subpattern, the match always failed. PCRE now remembers the number
4634 of the largest back reference, and gets some temporary memory in which to save
4635 the offsets during matching if necessary, in order to ensure that
4636 backreferences always work.
4638 4. Increased the compatibility with Perl in a number of ways:
4640 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4641 to request this handling. The option can be set at compile or exec time.
4643 (b) $ matches before a terminating newline by default; an option
4644 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4645 mode). The option can be set at compile or exec time.
4647 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4648 the same as Perl's. If the decimal number it represents is less than 10
4649 or there aren't that many previous left capturing parentheses, an octal
4650 escape is read. Inside a character class, it's always an octal escape,
4651 even if it is a single digit.
4653 (d) An escaped but undefined alphabetic character is taken as a literal,
4654 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4655 escapes.
4657 (e) {0} is now permitted. (The previous item is removed from the compiled
4658 pattern).
4660 5. Changed all the names of code files so that the basic parts are no longer
4661 than 10 characters, and abolished the teeny "globals.c" file.
4663 6. Changed the handling of character classes; they are now done with a 32-byte
4664 bit map always.
4666 7. Added the -d and /D options to pcretest to make it possible to look at the
4667 internals of compilation without having to recompile pcre.
4670 Version 0.95 23-Sep-97
4671 ----------------------
4673 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4674 \x20 at the start of a run of normal characters. These were being treated as
4675 real characters, instead of the source characters being re-checked.
4678 Version 0.94 18-Sep-97
4679 ----------------------
4681 1. The functions are now thread-safe, with the caveat that the global variables
4682 containing pointers to malloc() and free() or alternative functions are the
4683 same for all threads.
4685 2. Get pcre_study() to generate a bitmap of initial characters for non-
4686 anchored patterns when this is possible, and use it if passed to pcre_exec().
4689 Version 0.93 15-Sep-97
4690 ----------------------
4692 1. /(b)|(:+)/ was computing an incorrect first character.
4694 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4695 but not actually doing anything yet.
4697 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4698 as Perl does (e.g. [-az] or [az-]).
4700 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4701 all possible positions.
4703 5. Split up into different modules to avoid including unneeded functions in a
4704 compiled binary. However, compile and exec are still in one module. The "study"
4705 function is split off.
4707 6. The character tables are now in a separate module whose source is generated
4708 by an auxiliary program - but can then be edited by hand if required. There are
4709 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4710 toupper() in the code.
4712 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4713 make them global. Abolish the function for setting them, as the caller can now
4714 set them directly.
4717 Version 0.92 11-Sep-97
4718 ----------------------
4720 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4721 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4723 2. Caseless matching was not working in character classes if the characters in
4724 the pattern were in upper case.
4726 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4728 4. Make PCRE_ANCHORED public and accept as a compile option.
4730 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4731 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4732 pass them.
4734 6. Give an error if bad option bits passed at compile or run time.
4736 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4737 pcretest to cause it to pass that flag.
4739 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4740 options, and the first character, if set.
4742 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4745 Version 0.91 10-Sep-97
4746 ----------------------
4748 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4749 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4751 2. PCRE was looping on encountering an indefinitely repeated back reference to
4752 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4753 Perl does - treats the match as successful.
4755 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5