ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 931 - (show annotations)
Fri Feb 24 13:21:02 2012 UTC (9 years, 7 months ago) by ph10
File size: 227723 byte(s)
Make "RunTest list" list the tests without running them.
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.)
51 Version 8.30 04-February-2012
52 -----------------------------
54 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
55 name is defined in ctype.h.
57 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
58 only in quite long subpatterns.
60 3. Removed the function pcre_info(), which has been obsolete and deprecated
61 since it was replaced by pcre_fullinfo() in February 2000.
63 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
64 match a (*MARK), and the match failed at the start of the subject, a
65 reference to memory before the start of the subject could occur. This bug
66 was introduced by fix 17 of release 8.21.
68 5. A reference to an unset group with zero minimum repetition was giving
69 totally wrong answers (in non-JavaScript-compatibility mode). For example,
70 /(another)?(\1?)test/ matched against "hello world test". This bug was
71 introduced in release 8.13.
73 6. Add support for 16-bit character strings (a large amount of work involving
74 many changes and refactorings).
76 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
77 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
78 from a file.
80 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
81 rounding is not applied in this particular case).
83 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
84 if they appear, or are escaped, in patterns.
86 10. Get rid of a number of -Wunused-but-set-variable warnings.
88 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
89 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
90 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
91 also returns the mark "x". This bug applied to capturing parentheses,
92 non-capturing parentheses, and atomic parentheses. It also applied to some
93 assertions.
95 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
96 information out of configure.ac instead of relying on pcre.h.generic, which
97 is not stored in the repository.
99 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
100 -lreadline.
102 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
104 15. Applied Graycode's patch to put the top-level frame on the stack rather
105 than the heap when not using the stack for recursion. This gives a
106 performance improvement in many cases when recursion is not deep.
108 16. Experimental code added to "pcretest -C" to output the stack frame size.
111 Version 8.21 12-Dec-2011
112 ------------------------
114 1. Updating the JIT compiler.
116 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
117 are added as well.
119 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
120 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
121 calling _pcre_jit_exec. Some extra comments are added.
123 4. (*MARK) settings inside atomic groups that do not contain any capturing
124 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
125 was introduced by change 18 for 8.20.
127 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
128 ECMA-262 standard.
130 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
131 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
132 This bug was probably introduced by change 9 of 8.13.
134 7. While fixing 6 above, I noticed that a number of other items were being
135 incorrectly rejected as "not fixed length". This arose partly because newer
136 opcodes had not been added to the fixed-length checking code. I have (a)
137 corrected the bug and added tests for these items, and (b) arranged for an
138 error to occur if an unknown opcode is encountered while checking for fixed
139 length instead of just assuming "not fixed length". The items that were
140 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
141 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
142 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
144 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
145 being incorrectly compiled and would have given unpredicatble results.
147 9. A possessively repeated subpattern with minimum repeat count greater than
148 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
149 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
150 the first (A) could occur when it should not.
152 10. Add a cast and remove a redundant test from the code.
154 11. JIT should use pcre_malloc/pcre_free for allocation.
156 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
157 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
158 is anything other than /usr, -L is still shown).
160 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
162 14. Perl does not support \N without a following name in a [] class; PCRE now
163 also gives an error.
165 15. If a forward reference was repeated with an upper limit of around 2000,
166 it caused the error "internal error: overran compiling workspace". The
167 maximum number of forward references (including repeats) was limited by the
168 internal workspace, and dependent on the LINK_SIZE. The code has been
169 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
170 the default depends on LINK_SIZE. There is a new upper limit (for safety)
171 of around 200,000 forward references. While doing this, I also speeded up
172 the filling in of repeated forward references.
174 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
175 incorrectly expecting the subject to contain another "a" after the start.
177 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
178 in the match, the SKIP should be ignored. This was not happening; instead
179 the SKIP was being treated as NOMATCH. For patterns such as
180 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
181 tested.
183 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
184 now much more compatible with Perl, in particular in cases where the result
185 is a non-match for a non-anchored pattern. For example, if
186 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
187 "m", where previously it did not return a name. A side effect of this
188 change is that for partial matches, the last encountered mark name is
189 returned, as for non matches. A number of tests that were previously not
190 Perl-compatible have been moved into the Perl-compatible test files. The
191 refactoring has had the pleasing side effect of removing one argument from
192 the match() function, thus reducing its stack requirements.
194 19. If the /S+ option was used in pcretest to study a pattern using JIT,
195 subsequent uses of /S (without +) incorrectly behaved like /S+.
197 21. Retrieve executable code size support for the JIT compiler and fixing
198 some warnings.
200 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
201 not work when the shorter character appeared right at the end of the
202 subject string.
204 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
205 systems.
207 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
208 output it when the /M option is used in pcretest.
210 25. The CheckMan script was not being included in the distribution. Also, added
211 an explicit "perl" to run Perl scripts from the PrepareRelease script
212 because this is reportedly needed in Windows.
214 26. If study data was being save in a file and studying had not found a set of
215 "starts with" bytes for the pattern, the data written to the file (though
216 never used) was taken from uninitialized memory and so caused valgrind to
217 complain.
219 27. Updated RunTest.bat as provided by Sheri Pierce.
221 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
223 29. Computation of memory usage for the table of capturing group names was
224 giving an unnecessarily large value.
227 Version 8.20 21-Oct-2011
228 ------------------------
230 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
231 a POSIX class. After further experiments with Perl, which convinced me that
232 Perl has bugs and confusions, a closing square bracket is no longer allowed
233 in a POSIX name. This bug also affected patterns with classes that started
234 with full stops.
236 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
237 captured substring, but while checking the failing first alternative,
238 substring 1 is temporarily captured. If the output vector supplied to
239 pcre_exec() was not big enough for this capture, the yield of the function
240 was still zero ("insufficient space for captured substrings"). This cannot
241 be totally fixed without adding another stack variable, which seems a lot
242 of expense for a edge case. However, I have improved the situation in cases
243 such as /(a)(b)x|abc/ matched against "abc", where the return code
244 indicates that fewer than the maximum number of slots in the ovector have
245 been set.
247 3. Related to (2) above: when there are more back references in a pattern than
248 slots in the output vector, pcre_exec() uses temporary memory during
249 matching, and copies in the captures as far as possible afterwards. It was
250 using the entire output vector, but this conflicts with the specification
251 that only 2/3 is used for passing back captured substrings. Now it uses
252 only the first 2/3, for compatibility. This is, of course, another edge
253 case.
255 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
256 main code base, and can be used by building with --enable-jit. When this is
257 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
258 runtime --no-jit option is given.
260 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
261 ovector, the return from the function was zero, implying that there were
262 other matches that did not fit. The correct "exactly full" value is now
263 returned.
265 6. If a subpattern that was called recursively or as a subroutine contained
266 (*PRUNE) or any other control that caused it to give a non-standard return,
267 invalid errors such as "Error -26 (nested recursion at the same subject
268 position)" or even infinite loops could occur.
270 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
271 computing the minimum length on reaching *ACCEPT, and so ended up with the
272 wrong value of 1 rather than 0. Further investigation indicates that
273 computing a minimum subject length in the presence of *ACCEPT is difficult
274 (think back references, subroutine calls), and so I have changed the code
275 so that no minimum is registered for a pattern that contains *ACCEPT.
277 8. If (*THEN) was present in the first (true) branch of a conditional group,
278 it was not handled as intended. [But see 16 below.]
280 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
281 Sheri Pierce.
283 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
284 the first byte in a match must be "a".
286 11. Change 17 for 8.13 increased the recursion depth for patterns like
287 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
288 pattern contains any instances of (*THEN). If it does not, the old
289 optimizations are restored. It would be nice to do this on a per-group
290 basis, but at the moment that is not feasible.
292 12. In some environments, the output of pcretest -C is CRLF terminated. This
293 broke RunTest's code that checks for the link size. A single white space
294 character after the value is now allowed for.
296 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
297 For "fr", it uses the Windows-specific input and output files.
299 14. If (*THEN) appeared in a group that was called recursively or as a
300 subroutine, it did not work as intended. [But see next item.]
302 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
303 pattern fragments (but not containing any | characters). If A and B are
304 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
305 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
306 D. In other words, Perl considers parentheses that do not contain any |
307 characters to be part of a surrounding alternative, whereas PCRE was
308 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
309 differently. PCRE now behaves in the same way as Perl, except in the case
310 of subroutine/recursion calls such as (?1) which have in any case always
311 been different (but PCRE had them first :-).
313 16. Related to 15 above: Perl does not treat the | in a conditional group as
314 creating alternatives. Such a group is treated in the same way as an
315 ordinary group without any | characters when processing (*THEN). PCRE has
316 been changed to match Perl's behaviour.
318 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
319 RunGrepTest script failed.
321 18. Change 22 for version 13 caused atomic groups to use more stack. This is
322 inevitable for groups that contain captures, but it can lead to a lot of
323 stack use in large patterns. The old behaviour has been restored for atomic
324 groups that do not contain any capturing parentheses.
326 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
327 suppress the check for a minimum subject length at run time. (If it was
328 given to pcre_exec() or pcre_dfa_exec() it did work.)
330 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
331 fail to work when decoding hex characters in data strings in EBCDIC
332 environments.
334 21. It appears that in at least one Mac OS environment, the isxdigit() function
335 is implemented as a macro that evaluates to its argument more than once,
336 contravening the C 90 Standard (I haven't checked a later standard). There
337 was an instance in pcretest which caused it to go wrong when processing
338 \x{...} escapes in subject strings. The has been rewritten to avoid using
339 things like p++ in the argument of isxdigit().
342 Version 8.13 16-Aug-2011
343 ------------------------
345 1. The Unicode data tables have been updated to Unicode 6.0.0.
347 2. Two minor typos in pcre_internal.h have been fixed.
349 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
350 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
351 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
353 4. There were a number of related bugs in the code for matching backrefences
354 caselessly in UTF-8 mode when codes for the characters concerned were
355 different numbers of bytes. For example, U+023A and U+2C65 are an upper
356 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
357 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
358 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
359 2-byte code at the end of the subject (it thought there wasn't enough data
360 left).
362 5. Comprehensive information about what went wrong is now returned by
363 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
364 as the output vector has at least 2 elements. The offset of the start of
365 the failing character and a reason code are placed in the vector.
367 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
368 now returned is for the first byte of the failing character, instead of the
369 last byte inspected. This is an incompatible change, but I hope it is small
370 enough not to be a problem. It makes the returned offset consistent with
371 pcre_exec() and pcre_dfa_exec().
373 7. pcretest now gives a text phrase as well as the error number when
374 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
375 failure, the offset and reason code are output.
377 8. When \R was used with a maximizing quantifier it failed to skip backwards
378 over a \r\n pair if the subsequent match failed. Instead, it just skipped
379 back over a single character (\n). This seems wrong (because it treated the
380 two characters as a single entity when going forwards), conflicts with the
381 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
382 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
383 has been changed.
385 9. Some internal refactoring has changed the processing so that the handling
386 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
387 time (the PCRE_DOTALL option was changed this way some time ago: version
388 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
389 which was always a bit of a fudge. It also means that there is one less
390 argument for the match() function, which reduces its stack requirements
391 slightly. This change also fixes an incompatibility with Perl: the pattern
392 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
394 10. More internal refactoring has drastically reduced the number of recursive
395 calls to match() for possessively repeated groups such as (abc)++ when
396 using pcre_exec().
398 11. While implementing 10, a number of bugs in the handling of groups were
399 discovered and fixed:
401 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
402 (a|)*(?1) gave a compile-time internal error.
403 ((a|)+)+ did not notice that the outer group could match an empty string.
404 (^a|^)+ was not marked as anchored.
405 (.*a|.*)+ was not marked as matching at start or after a newline.
407 12. Yet more internal refactoring has removed another argument from the match()
408 function. Special calls to this function are now indicated by setting a
409 value in a variable in the "match data" data block.
411 13. Be more explicit in pcre_study() instead of relying on "default" for
412 opcodes that mean there is no starting character; this means that when new
413 ones are added and accidentally left out of pcre_study(), testing should
414 pick them up.
416 14. The -s option of pcretest has been documented for ages as being an old
417 synonym of -m (show memory usage). I have changed it to mean "force study
418 for every regex", that is, assume /S for every regex. This is similar to -i
419 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
420 using it. It makes it easier to run collections of tests with and without
421 study enabled, and thereby test pcre_study() more easily. All the standard
422 tests are now run with and without -s (but some patterns can be marked as
423 "never study" - see 20 below).
425 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
426 restoration of the capturing data to the outer values was not happening
427 correctly.
429 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
430 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
431 pattern had matched an empty string, and so incorrectly returned a no
432 match.
434 17. There was optimizing code for the last branch of non-capturing parentheses,
435 and also for the obeyed branch of a conditional subexpression, which used
436 tail recursion to cut down on stack usage. Unfortunately, now that there is
437 the possibility of (*THEN) occurring in these branches, tail recursion is
438 no longer possible because the return has to be checked for (*THEN). These
439 two optimizations have therefore been removed. [But see 8.20/11 above.]
441 18. If a pattern containing \R was studied, it was assumed that \R always
442 matched two bytes, thus causing the minimum subject length to be
443 incorrectly computed because \R can also match just one byte.
445 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
446 was incorrectly computed.
448 20. If /S is present twice on a test pattern in pcretest input, it now
449 *disables* studying, thereby overriding the use of -s on the command line
450 (see 14 above). This is necessary for one or two tests to keep the output
451 identical in both cases.
453 21. When (*ACCEPT) was used in an assertion that matched an empty string and
454 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
456 22. When an atomic group that contained a capturing parenthesis was
457 successfully matched, but the branch in which it appeared failed, the
458 capturing was not being forgotten if a higher numbered group was later
459 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
460 group 1 to "a", when in fact it should be unset. This applied to multi-
461 branched capturing and non-capturing groups, repeated or not, and also to
462 positive assertions (capturing in negative assertions does not happen
463 in PCRE) and also to nested atomic groups.
465 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
466 subject after a captured substring, to make it easier to tell which of a
467 number of identical substrings has been captured.
469 24. The way atomic groups are processed by pcre_exec() has been changed so that
470 if they are repeated, backtracking one repetition now resets captured
471 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
472 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
473 "aaa". Previously, it would have been "a". As part of this code
474 refactoring, the way recursive calls are handled has also been changed.
476 25. If an assertion condition captured any substrings, they were not passed
477 back unless some other capturing happened later. For example, if
478 (?(?=(a))a) was matched against "a", no capturing was returned.
480 26. When studying a pattern that contained subroutine calls or assertions,
481 the code for finding the minimum length of a possible match was handling
482 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
483 group 1 called group 2 while simultaneously a separate group 2 called group
484 1). A stack overflow occurred in this case. I have fixed this by limiting
485 the recursion depth to 10.
487 27. Updated RunTest.bat in the distribution to the version supplied by Tom
488 Fortmann. This supports explicit test numbers on the command line, and has
489 argument validation and error reporting.
491 28. An instance of \X with an unlimited repeat could fail if at any point the
492 first character it looked at was a mark character.
494 29. Some minor code refactoring concerning Unicode properties and scripts
495 should reduce the stack requirement of match() slightly.
497 30. Added the '=' option to pcretest to check the setting of unused capturing
498 slots at the end of the pattern, which are documented as being -1, but are
499 not included in the return count.
501 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
502 compiled something random. Now it gives a compile-time error (as does
503 Perl).
505 32. A *MARK encountered during the processing of a positive assertion is now
506 recorded and passed back (compatible with Perl).
508 33. If --only-matching or --colour was set on a pcregrep call whose pattern
509 had alternative anchored branches, the search for a second match in a line
510 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
511 matched the line "0102" twice. The same bug affected patterns that started
512 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
513 twice.
515 34. Previously, PCRE did not allow quantification of assertions. However, Perl
516 does, and because of capturing effects, quantifying parenthesized
517 assertions may at times be useful. Quantifiers are now allowed for
518 parenthesized assertions.
520 35. A minor code tidy in pcre_compile() when checking options for \R usage.
522 36. \g was being checked for fancy things in a character class, when it should
523 just be a literal "g".
525 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
526 appearance of a nested POSIX class supersedes an apparent external class.
527 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
528 unescaped square brackets may also appear as part of class names. For
529 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
530 more like Perl. (But see 8.20/1 above.)
532 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
533 was because it thought it was \N{name}, which is not supported).
535 39. Add minix to OS list not supporting the -S option in pcretest.
537 40. PCRE tries to detect cases of infinite recursion at compile time, but it
538 cannot analyze patterns in sufficient detail to catch mutual recursions
539 such as ((?1))((?2)). There is now a runtime test that gives an error if a
540 subgroup is called recursively as a subpattern for a second time at the
541 same position in the subject string. In previous releases this might have
542 been caught by the recursion limit, or it might have run out of stack.
544 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
545 happen only once. PCRE was, however incorrectly giving a compile time error
546 "recursive call could loop indefinitely" because it cannot analyze the
547 pattern in sufficient detail. The compile time test no longer happens when
548 PCRE is compiling a conditional subpattern, but actual runaway loops are
549 now caught at runtime (see 40 above).
551 42. It seems that Perl allows any characters other than a closing parenthesis
552 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
553 has been changed to be the same.
555 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
556 as not to get warnings when autogen.sh is called. Also changed
557 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
559 44. To help people who use pcregrep to scan files containing exceedingly long
560 lines, the following changes have been made:
562 (a) The default value of the buffer size parameter has been increased from
563 8K to 20K. (The actual buffer used is three times this size.)
565 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
566 PCRE is built.
568 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
569 to be set at run time.
571 (d) Numerical values in pcregrep options can be followed by K or M, for
572 example --buffer-size=50K.
574 (e) If a line being scanned overflows pcregrep's buffer, an error is now
575 given and the return code is set to 2.
577 45. Add a pointer to the latest mark to the callout data block.
579 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
580 partial match of an empty string instead of no match. This was specific to
581 the use of ".".
583 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
584 complete match instead of a partial match. This bug was dependent on both
585 the PCRE_UTF8 and PCRE_DOTALL options being set.
587 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
588 starting byte set, because \b was not being ignored.
591 Version 8.12 15-Jan-2011
592 ------------------------
594 1. Fixed some typos in the markup of the man pages, and wrote a script that
595 checks for such things as part of the documentation building process.
597 2. On a big-endian 64-bit system, pcregrep did not correctly process the
598 --match-limit and --recursion-limit options (added for 8.11). In
599 particular, this made one of the standard tests fail. (The integer value
600 went into the wrong half of a long int.)
602 3. If the --colour option was given to pcregrep with -v (invert match), it
603 did strange things, either producing crazy output, or crashing. It should,
604 of course, ignore a request for colour when reporting lines that do not
605 match.
607 4. Another pcregrep bug caused similar problems if --colour was specified with
608 -M (multiline) and the pattern match finished with a line ending.
610 5. In pcregrep, when a pattern that ended with a literal newline sequence was
611 matched in multiline mode, the following line was shown as part of the
612 match. This seems wrong, so I have changed it.
614 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
615 the check for further matches in the same line (so they could be coloured)
616 to overrun the end of the current line. If another match was found, it was
617 incorrectly shown (and then shown again when found in the next line).
619 7. If pcregrep was compiled under Windows, there was a reference to the
620 function pcregrep_exit() before it was defined. I am assuming this was
621 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
622 reported by a user. I've moved the definition above the reference.
625 Version 8.11 10-Dec-2010
626 ------------------------
628 1. (*THEN) was not working properly if there were untried alternatives prior
629 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
630 backtracked to try for "b" instead of moving to the next alternative branch
631 at the same level (in this case, to look for "c"). The Perl documentation
632 is clear that when (*THEN) is backtracked onto, it goes to the "next
633 alternative in the innermost enclosing group".
635 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
636 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
637 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
638 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
639 (*THEN).
641 3. If \s appeared in a character class, it removed the VT character from
642 the class, even if it had been included by some previous item, for example
643 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
644 of \s, but is part of the POSIX "space" class.)
646 4. A partial match never returns an empty string (because you can always
647 match an empty string at the end of the subject); however the checking for
648 an empty string was starting at the "start of match" point. This has been
649 changed to the "earliest inspected character" point, because the returned
650 data for a partial match starts at this character. This means that, for
651 example, /(?<=abc)def/ gives a partial match for the subject "abc"
652 (previously it gave "no match").
654 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
655 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
656 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
657 has an implication that the given string is incomplete (because a partial
658 match is preferred over a full match). For this reason, these items now
659 give a partial match in this situation. [Aside: previously, the one case
660 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
661 match rather than a full match, which was wrong by the old rules, but is
662 now correct.]
664 6. There was a bug in the handling of #-introduced comments, recognized when
665 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
666 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
667 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
668 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
669 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
670 places in pcre_compile().
672 7. Related to (6) above, when pcre_compile() was skipping #-introduced
673 comments when looking ahead for named forward references to subpatterns,
674 the only newline sequence it recognized was NL. It now handles newlines
675 according to the set newline convention.
677 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
678 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
679 cater for a lack of strerror(). These oversights have been fixed.
681 9. Added --match-limit and --recursion-limit to pcregrep.
683 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
685 11. When the -o option was used, pcregrep was setting a return code of 1, even
686 when matches were found, and --line-buffered was not being honoured.
688 12. Added an optional parentheses number to the -o and --only-matching options
689 of pcregrep.
691 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
692 can match an empty string. The code to do it in pcretest and pcredemo
693 needed fixing:
695 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
696 only one byte after an empty string match just before CRLF (this case
697 just got forgotten; "any" and "anycrlf" were OK).
699 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
700 mode when an empty string match preceded an ASCII character followed by
701 a non-ASCII character. (The code for advancing by one character rather
702 than one byte was nonsense.)
704 (c) The pcredemo.c sample program did not have any code at all to handle
705 the cases when CRLF is a valid newline sequence.
707 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
708 as a starting offset was within the subject string. There is now a new
709 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
710 negative or greater than the length of the string. In order to test this,
711 pcretest is extended to allow the setting of negative starting offsets.
713 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
714 starting offset points to the beginning of a UTF-8 character was
715 unnecessarily clumsy. I tidied it up.
717 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
718 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
720 17. Nobody had reported that the --include_dir option, which was added in
721 release 7.7 should have been called --include-dir (hyphen, not underscore)
722 for compatibility with GNU grep. I have changed it to --include-dir, but
723 left --include_dir as an undocumented synonym, and the same for
724 --exclude-dir, though that is not available in GNU grep, at least as of
725 release 2.5.4.
727 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
728 characters from a string of bytes) have been redefined so as not to use
729 loops, in order to improve performance in some environments. At the same
730 time, I abstracted some of the common code into auxiliary macros to save
731 repetition (this should not affect the compiled code).
733 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
734 compile-time error is now given if \c is not followed by an ASCII
735 character, that is, a byte less than 128. (In EBCDIC mode, the code is
736 different, and any byte value is allowed.)
738 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
739 START_OPTIMIZE option, which is now allowed at compile time - but just
740 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
741 to pcregrep and other applications that have no direct access to PCRE
742 options. The new /Y option in pcretest sets this option when calling
743 pcre_compile().
745 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
746 back references. Groups containing recursive back references were forced to
747 be atomic by that change, but in the case of named groups, the amount of
748 memory required was incorrectly computed, leading to "Failed: internal
749 error: code overflow". This has been fixed.
751 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
752 pcretest.c, to avoid build problems in some Borland environments.
755 Version 8.10 25-Jun-2010
756 ------------------------
758 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
759 THEN.
761 2. (*ACCEPT) was not working when inside an atomic group.
763 3. Inside a character class, \B is treated as a literal by default, but
764 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
765 causes the error). The code is unchanged, but I tidied the documentation.
767 4. Inside a character class, PCRE always treated \R and \X as literals,
768 whereas Perl faults them if its -w option is set. I have changed PCRE so
769 that it faults them when PCRE_EXTRA is set.
771 5. Added support for \N, which always matches any character other than
772 newline. (It is the same as "." when PCRE_DOTALL is not set.)
774 6. When compiling pcregrep with newer versions of gcc which may have
775 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
776 declared with attribute warn_unused_result" were given. Just casting the
777 result to (void) does not stop the warnings; a more elaborate fudge is
778 needed. I've used a macro to implement this.
780 7. Minor change to pcretest.c to avoid a compiler warning.
782 8. Added four artifical Unicode properties to help with an option to make
783 \s etc use properties (see next item). The new properties are: Xan
784 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
786 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
787 use Unicode properties. (*UCP) at the start of a pattern can be used to set
788 this option. Modified pcretest to add /W to test this facility. Added
789 REG_UCP to make it available via the POSIX interface.
791 10. Added --line-buffered to pcregrep.
793 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
794 studied, and the match started with a letter with a code point greater than
795 127 whose first byte was different to the first byte of the other case of
796 the letter, the other case of this starting letter was not recognized
797 (#976).
799 12. If a pattern that was studied started with a repeated Unicode property
800 test, for example, \p{Nd}+, there was the theoretical possibility of
801 setting up an incorrect bitmap of starting bytes, but fortunately it could
802 not have actually happened in practice until change 8 above was made (it
803 added property types that matched character-matching opcodes).
805 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
806 possible starting bytes for non-anchored patterns.
808 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
809 \R, and also a number of cases that involve Unicode properties, both
810 explicit and implicit when PCRE_UCP is set.
812 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
813 input, it could crash or give wrong results if characters with values
814 greater than 0xc0 were present in the subject string. (Detail: it assumed
815 UTF-8 input when processing these items.)
817 16. Added a lot of (int) casts to avoid compiler warnings in systems where
818 size_t is 64-bit (#991).
820 17. Added a check for running out of memory when PCRE is compiled with
821 --disable-stack-for-recursion (#990).
823 18. If the last data line in a file for pcretest does not have a newline on
824 the end, a newline was missing in the output.
826 19. The default pcre_chartables.c file recognizes only ASCII characters (values
827 less than 128) in its various bitmaps. However, there is a facility for
828 generating tables according to the current locale when PCRE is compiled. It
829 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
830 characters, are recognized by isspace() and therefore were getting set in
831 these tables, and indeed these tables seem to approximate to ISO 8859. This
832 caused a problem in UTF-8 mode when pcre_study() was used to create a list
833 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
834 which of course cannot start UTF-8 characters. I have changed the code so
835 that only real ASCII characters (less than 128) and the correct starting
836 bytes for UTF-8 encodings are set for characters greater than 127 when in
837 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
838 altogether.)
840 20. Added the /T option to pcretest so as to be able to run tests with non-
841 standard character tables, thus making it possible to include the tests
842 used for 19 above in the standard set of tests.
844 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
845 reference to a subpattern the other side of a comment that contains an
846 opening parenthesis caused either an internal compiling error, or a
847 reference to the wrong subpattern.
850 Version 8.02 19-Mar-2010
851 ------------------------
853 1. The Unicode data tables have been updated to Unicode 5.2.0.
855 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
856 configured.
858 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
859 original author of that file, following a query about its status.
861 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
862 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
864 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
865 quantifier applied to a forward-referencing subroutine call, could compile
866 incorrect code or give the error "internal error: previously-checked
867 referenced subpattern not found".
869 6. Both MS Visual Studio and Symbian OS have problems with initializing
870 variables to point to external functions. For these systems, therefore,
871 pcre_malloc etc. are now initialized to local functions that call the
872 relevant global functions.
874 7. There were two entries missing in the vectors called coptable and poptable
875 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
876 I've fixed the data, and added a kludgy way of testing at compile time that
877 the lengths are correct (equal to the number of opcodes).
879 8. Following on from 7, I added a similar kludge to check the length of the
880 eint vector in pcreposix.c.
882 9. Error texts for pcre_compile() are held as one long string to avoid too
883 much relocation at load time. To find a text, the string is searched,
884 counting zeros. There was no check for running off the end of the string,
885 which could happen if a new error number was added without updating the
886 string.
888 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
890 11. \K was not working if it appeared in an atomic group or in a group that
891 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
892 \K is "not well defined" if used in an assertion. PCRE now accepts it if
893 the assertion is positive, but not if it is negative.
895 12. Change 11 fortuitously reduced the size of the stack frame used in the
896 "match()" function of pcre_exec.c by one pointer. Forthcoming
897 implementation of support for (*MARK) will need an extra pointer on the
898 stack; I have reserved it now, so that the stack frame size does not
899 decrease.
901 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
902 item in branch that calls a recursion is a subroutine call - as in the
903 second branch in the above example - was incorrectly given the compile-
904 time error "recursive call could loop indefinitely" because pcre_compile()
905 was not correctly checking the subroutine for matching a non-empty string.
907 14. The checks for overrunning compiling workspace could trigger after an
908 overrun had occurred. This is a "should never occur" error, but it can be
909 triggered by pathological patterns such as hundreds of nested parentheses.
910 The checks now trigger 100 bytes before the end of the workspace.
912 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
915 Version 8.01 19-Jan-2010
916 ------------------------
918 1. If a pattern contained a conditional subpattern with only one branch (in
919 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
920 computed the wrong minimum data length (which is of course zero for such
921 subpatterns). This could cause incorrect "no match" results.
923 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
924 the pattern is reset in the first branch, pcre_compile() failed with
925 "internal error: code overflow at offset...". This happened only when
926 the reset was to the original external option setting. (An optimization
927 abstracts leading options settings into an external setting, which was the
928 cause of this.)
930 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
931 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
932 assertion pattern did not match (meaning that the assertion was true), it
933 was incorrectly treated as false if the SKIP had been reached during the
934 matching. This also applied to assertions used as conditions.
936 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
937 assertion subpattern, including such a pattern used as a condition,
938 unpredictable results occurred, instead of the error return
941 5. The C++ GlobalReplace function was not working like Perl for the special
942 situation when an empty string is matched. It now does the fancy magic
943 stuff that is necessary.
945 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
946 removed. (These were left over from very, very early versions of PCRE.)
948 7. Some cosmetic changes to the code to make life easier when compiling it
949 as part of something else:
951 (a) Change DEBUG to PCRE_DEBUG.
953 (b) In pcre_compile(), rename the member of the "branch_chain" structure
954 called "current" as "current_branch", to prevent a collision with the
955 Linux macro when compiled as a kernel module.
957 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
958 prevent a collision with the Linux macro when compiled as a kernel
959 module.
961 8. In pcre_compile() there are some checks for integer overflows that used to
962 cast potentially large values to (double). This has been changed to that
963 when building, a check for int64_t is made, and if it is found, it is used
964 instead, thus avoiding the use of floating point arithmetic. (There is no
965 other use of FP in PCRE.) If int64_t is not found, the fallback is to
966 double.
968 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
969 2005 (difference between two addresses compared to an unsigned value).
971 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
972 custom one, because of the following reported problem in Windows:
974 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
975 under Win32.
976 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
977 therefore missing the function definition.
978 - The compiler thus generates a "C" signature for the test function.
979 - The linker fails to find the "C" function.
980 - PCRE fails to configure if asked to do so against libbz2.
982 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
983 messages were output:
985 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
986 rerunning libtoolize, to keep the correct libtool macros in-tree.
987 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
989 I have done both of these things.
991 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
992 most of the time, it *can* run out if it is given a pattern that contains a
993 runaway infinite recursion. I updated the discussion in the pcrestack man
994 page.
996 13. Now that we have gone to the x.xx style of version numbers, the minor
997 version may start with zero. Using 08 or 09 is a bad idea because users
998 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
999 interpreted as invalid octal numbers. I've updated the previous comment in
1000 configure.ac, and also added a check that gives an error if 08 or 09 are
1001 used.
1003 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1004 causing partial matching to fail when the end of the subject matched \W
1005 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1007 15. There were some discrepancies between the declarations in pcre_internal.h
1008 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1009 their definitions. The declarations used "const uschar *" and the
1010 definitions used USPTR. Even though USPTR is normally defined as "const
1011 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1012 reported that: "This difference in casting confuses some C++ compilers, for
1013 example, SunCC recognizes above declarations as different functions and
1014 generates broken code for hbpcre." I have changed the declarations to use
1015 USPTR.
1017 16. GNU libtool is named differently on some systems. The autogen.sh script now
1018 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1019 (FreeBSD).
1021 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1022 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1023 comment: "Figure out how to create a longlong from a string: strtoll and
1024 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1025 instance, but it only takes 2 args instead of 3!"
1027 18. A subtle bug concerned with back references has been fixed by a change of
1028 specification, with a corresponding code fix. A pattern such as
1029 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1030 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1031 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1032 same bug. Such groups have to be quantified to be useful, or contained
1033 inside another quantified group. (If there's no repetition, the reference
1034 can never match.) The problem arises because, having left the group and
1035 moved on to the rest of the pattern, a later failure that backtracks into
1036 the group uses the captured value from the final iteration of the group
1037 rather than the correct earlier one. I have fixed this in PCRE by forcing
1038 any group that contains a reference to itself to be an atomic group; that
1039 is, there cannot be any backtracking into it once it has completed. This is
1040 similar to recursive and subroutine calls.
1043 Version 8.00 19-Oct-09
1044 ----------------------
1046 1. The table for translating pcre_compile() error codes into POSIX error codes
1047 was out-of-date, and there was no check on the pcre_compile() error code
1048 being within the table. This could lead to an OK return being given in
1049 error.
1051 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1052 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1053 in a Windows environment.
1055 3. The pcregrep --count option prints the count for each file even when it is
1056 zero, as does GNU grep. However, pcregrep was also printing all files when
1057 --files-with-matches was added. Now, when both options are given, it prints
1058 counts only for those files that have at least one match. (GNU grep just
1059 prints the file name in this circumstance, but including the count seems
1060 more useful - otherwise, why use --count?) Also ensured that the
1061 combination -clh just lists non-zero counts, with no names.
1063 4. The long form of the pcregrep -F option was incorrectly implemented as
1064 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1065 but it seems right to fix it, and I didn't think it was worth preserving
1066 the old behaviour.
1068 5. The command line items --regex=pattern and --regexp=pattern were not
1069 recognized by pcregrep, which required --regex pattern or --regexp pattern
1070 (with a space rather than an '='). The man page documented the '=' forms,
1071 which are compatible with GNU grep; these now work.
1073 6. No libpcreposix.pc file was created for pkg-config; there was just
1074 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1076 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1077 when UCP support is not needed, by modifying the Python script that
1078 generates it from Unicode data files. This should not matter if the module
1079 is correctly used as a library, but I received one complaint about 50K of
1080 unwanted data. My guess is that the person linked everything into his
1081 program rather than using a library. Anyway, it does no harm.
1083 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1084 was a minimum greater than 1 for a wide character in a possessive
1085 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1086 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1087 character. Chaos in the form of incorrect output or a compiling loop could
1088 result.
1090 9. The restrictions on what a pattern can contain when partial matching is
1091 requested for pcre_exec() have been removed. All patterns can now be
1092 partially matched by this function. In addition, if there are at least two
1093 slots in the offset vector, the offset of the earliest inspected character
1094 for the match and the offset of the end of the subject are set in them when
1095 PCRE_ERROR_PARTIAL is returned.
1097 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1098 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1099 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1100 and may be more useful for multi-segment matching.
1102 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1103 used to be given if ever the end of the subject was reached; now it is
1104 given only if matching could not proceed because another character was
1105 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1106 string "Z", which now yields "no match" instead of "partial match". In the
1107 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1108 final character ended with (*FAIL).
1110 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1111 if the pattern had a "must contain" character that was already found in the
1112 earlier partial match, unless partial matching was again requested. For
1113 example, with the pattern /dog.(body)?/, the "must contain" character is
1114 "g". If the first part-match was for the string "dog", restarting with
1115 "sbody" failed. This bug has been fixed.
1117 13. The string returned by pcre_dfa_exec() after a partial match has been
1118 changed so that it starts at the first inspected character rather than the
1119 first character of the match. This makes a difference only if the pattern
1120 starts with a lookbehind assertion or \b or \B (\K is not supported by
1121 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1122 matching functions compatible, and I think it's the right thing to do.
1124 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1125 so that the demonstration program is easily available in environments where
1126 PCRE has not been installed from source.
1128 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1129 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1130 library.
1132 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1133 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1134 is not the first non-POSIX option to be added. Clearly some people find
1135 these options useful.
1137 17. If a caller to the POSIX matching function regexec() passes a non-zero
1138 value for nmatch with a NULL value for pmatch, the value of
1139 nmatch is forced to zero.
1141 18. RunGrepTest did not have a test for the availability of the -u option of
1142 the diff command, as RunTest does. It now checks in the same way as
1143 RunTest, and also checks for the -b option.
1145 19. If an odd number of negated classes containing just a single character
1146 interposed, within parentheses, between a forward reference to a named
1147 subpattern and the definition of the subpattern, compilation crashed with
1148 an internal error, complaining that it could not find the referenced
1149 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1150 [The bug was that it was starting one character too far in when skipping
1151 over the character class, thus treating the ] as data rather than
1152 terminating the class. This meant it could skip too much.]
1154 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1155 /g option in pcretest when the pattern contains \K, which makes it possible
1156 to have an empty string match not at the start, even when the pattern is
1157 anchored. Updated pcretest and pcredemo to use this option.
1159 21. If the maximum number of capturing subpatterns in a recursion was greater
1160 than the maximum at the outer level, the higher number was returned, but
1161 with unset values at the outer level. The correct (outer level) value is
1162 now given.
1164 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1165 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1166 make it do so. The string so far is captured, making this feature
1167 compatible with Perl.
1169 23. The tests have been re-organized, adding tests 11 and 12, to make it
1170 possible to check the Perl 5.10 features against Perl 5.10.
1172 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1173 pattern matches a fixed length string. PCRE did not allow this; now it
1174 does. Neither allows recursion.
1176 25. I finally figured out how to implement a request to provide the minimum
1177 length of subject string that was needed in order to match a given pattern.
1178 (It was back references and recursion that I had previously got hung up
1179 on.) This code has now been added to pcre_study(); it finds a lower bound
1180 to the length of subject needed. It is not necessarily the greatest lower
1181 bound, but using it to avoid searching strings that are too short does give
1182 some useful speed-ups. The value is available to calling programs via
1183 pcre_fullinfo().
1185 26. While implementing 25, I discovered to my embarrassment that pcretest had
1186 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1187 study optimizations had never been tested with that matching function.
1188 Oops. What is worse, even when it was passed study data, there was a bug in
1189 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1190 were also very few tests of studied patterns with pcre_dfa_exec().
1192 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1193 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1194 on the other side of the coin, they are no longer allowed to have different
1195 names, because these cannot be distinguished in PCRE, and this has caused
1196 confusion. (This is a difference from Perl.)
1198 28. When duplicate subpattern names are present (necessarily with different
1199 numbers, as required by 27 above), and a test is made by name in a
1200 conditional pattern, either for a subpattern having been matched, or for
1201 recursion in such a pattern, all the associated numbered subpatterns are
1202 tested, and the overall condition is true if the condition is true for any
1203 one of them. This is the way Perl works, and is also more like the way
1204 testing by number works.
1207 Version 7.9 11-Apr-09
1208 ---------------------
1210 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1211 (pcretest), all targets were linked against these libraries. This included
1212 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1213 libraries. This caused unwanted dependencies to be created. This problem
1214 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1215 pcretest is linked with readline.
1217 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1218 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1219 moved outside it again, because FALSE and TRUE are already defined in AIX,
1220 but BOOL is not.
1222 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1223 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1225 4. The pcregrep documentation said spaces were inserted as well as colons (or
1226 hyphens) following file names and line numbers when outputting matching
1227 lines. This is not true; no spaces are inserted. I have also clarified the
1228 wording for the --colour (or --color) option.
1230 5. In pcregrep, when --colour was used with -o, the list of matching strings
1231 was not coloured; this is different to GNU grep, so I have changed it to be
1232 the same.
1234 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1235 each matching line was coloured. Now it goes on to look for further matches
1236 of any of the test patterns, which is the same behaviour as GNU grep.
1238 7. A pattern that could match an empty string could cause pcregrep to loop; it
1239 doesn't make sense to accept an empty string match in pcregrep, so I have
1240 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1241 seems to be how GNU grep behaves.
1243 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1244 start or after a newline", because the conditional assertion was not being
1245 correctly handled. The rule now is that both the assertion and what follows
1246 in the first alternative must satisfy the test.
1248 9. If auto-callout was enabled in a pattern with a conditional group whose
1249 condition was an assertion, PCRE could crash during matching, both with
1250 pcre_exec() and pcre_dfa_exec().
1252 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1253 used for matching.
1255 11. Unicode property support in character classes was not working for
1256 characters (bytes) greater than 127 when not in UTF-8 mode.
1258 12. Added the -M command line option to pcretest.
1260 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1262 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1264 16. Added comments and documentation about mis-use of no_arg in the C++
1265 wrapper.
1267 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1268 from Martin Jerabek that uses macro names for all relevant character and
1269 string constants.
1271 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1272 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1273 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1274 these, but not everybody uses configure.
1276 19. A conditional group that had only one branch was not being correctly
1277 recognized as an item that could match an empty string. This meant that an
1278 enclosing group might also not be so recognized, causing infinite looping
1279 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1280 with the subject "ab", where knowledge that the repeated group can match
1281 nothing is needed in order to break the loop.
1283 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1284 exec(), but without supplying a callout function, matching went wrong.
1286 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1287 leak if the size of the offset vector was greater than 30. When the vector
1288 is smaller, the saved offsets during recursion go onto a local stack
1289 vector, but for larger vectors malloc() is used. It was failing to free
1290 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1291 error, in fact).
1293 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1294 heapframe that is used only when UTF-8 support is enabled. This caused no
1295 problem, but was untidy.
1297 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1298 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1299 included within another project.
1301 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1302 slightly modified by me:
1304 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1305 not building pcregrep.
1307 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1308 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1310 25. Forward references, both numeric and by name, in patterns that made use of
1311 duplicate group numbers, could behave incorrectly or give incorrect errors,
1312 because when scanning forward to find the reference group, PCRE was not
1313 taking into account the duplicate group numbers. A pattern such as
1314 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1316 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1317 the feature of a custom pointer more persuasive (as requested by a user).
1319 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1320 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1321 pre-defined.
1323 28. Added support for (*UTF8) at the start of a pattern.
1325 29. Arrange for flags added by the "release type" setting in CMake to be shown
1326 in the configuration summary.
1329 Version 7.8 05-Sep-08
1330 ---------------------
1332 1. Replaced UCP searching code with optimized version as implemented for Ad
1333 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1334 stage table and inline lookup instead of a function, giving speed ups of 2
1335 to 5 times on some simple patterns that I tested. Permission was given to
1336 distribute the MultiStage2.py script that generates the tables (it's not in
1337 the tarball, but is in the Subversion repository).
1339 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1340 scripts.
1342 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1343 a group with a zero qualifier. The result of the study could be incorrect,
1344 or the function might crash, depending on the pattern.
1346 4. Caseless matching was not working for non-ASCII characters in back
1347 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1348 It now works when Unicode Property Support is available.
1350 5. In pcretest, an escape such as \x{de} in the data was always generating
1351 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1352 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1353 truncation.
1355 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1357 7. Added two (int) casts to pcregrep when printing the difference of two
1358 pointers, in case they are 64-bit values.
1360 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1361 test 2 if it fails.
1363 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1364 and a #define of that name to empty if it is not externally set. This is to
1365 allow users of MSVC to set it if necessary.
1367 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1368 the convenience functions in the pcre_get.c source file.
1370 11. An option change at the start of a pattern that had top-level alternatives
1371 could cause overwriting and/or a crash. This command provoked a crash in
1372 some environments:
1374 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1376 This potential security problem was recorded as CVE-2008-2371.
1378 12. For a pattern where the match had to start at the beginning or immediately
1379 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1380 pcre_dfa_exec() could read past the end of the passed subject if there was
1381 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1382 pcretest so that it places the subject at the end of its malloc-ed buffer.
1384 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1385 exec() might read past the end of the data buffer in UTF-8 mode.
1387 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1388 the data contained the byte 0x85 as part of a UTF-8 character within its
1389 first line. This applied both to normal and DFA matching.
1391 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1392 /^[^d]*?$/8 failed to match "abc".
1394 16. Added a missing copyright notice to pcrecpp_internal.h.
1396 17. Make it more clear in the documentation that values returned from
1397 pcre_exec() in ovector are byte offsets, not character counts.
1399 18. Tidied a few places to stop certain compilers from issuing warnings.
1401 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1402 supplied by Stefan Weber. I made a further small update for 7.8 because
1403 there is a change of source arrangements: the pcre_searchfuncs.c module is
1404 replaced by pcre_ucd.c.
1407 Version 7.7 07-May-08
1408 ---------------------
1410 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1411 a string to a long long, pretend we don't even have a long long." This is
1412 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1414 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1415 pre-7.6 versions, which defined a global no_arg variable instead of putting
1416 it in the RE class. (See also #8 below.)
1418 3. Remove a line of dead code, identified by coverity and reported by Nuno
1419 Lopes.
1421 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1423 (1) The include/exclude patterns were being applied to the whole pathnames
1424 of files, instead of just to the final components.
1426 (2) If there was more than one level of directory, the subdirectories were
1427 skipped unless they satisfied the include/exclude conditions. This is
1428 inconsistent with GNU grep (and could even be seen as contrary to the
1429 pcregrep specification - which I improved to make it absolutely clear).
1430 The action now is always to scan all levels of directory, and just
1431 apply the include/exclude patterns to regular files.
1433 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1434 --exclude_dir in the tests to avoid scanning .svn directories.
1436 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1437 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1438 doesn't support NULs in patterns.
1440 7. Added some missing "const"s to declarations of static tables in
1441 pcre_compile.c and pcre_dfa_exec.c.
1443 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1444 caused by fix #2 above. (Subsequently also a second patch to fix the
1445 first patch. And a third patch - this was a messy problem.)
1447 9. Applied Craig's patch to remove the use of push_back().
1449 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1450 matching function regexec().
1452 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1453 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1454 references. PCRE supports relative numbers with this syntax (I don't think
1455 Oniguruma does).
1457 12. Previously, a group with a zero repeat such as (...){0} was completely
1458 omitted from the compiled regex. However, this means that if the group
1459 was called as a subroutine from elsewhere in the pattern, things went wrong
1460 (an internal error was given). Such groups are now left in the compiled
1461 pattern, with a new opcode that causes them to be skipped at execution
1462 time.
1464 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1465 to the way PCRE behaves:
1467 (a) A lone ] character is dis-allowed (Perl treats it as data).
1469 (b) A back reference to an unmatched subpattern matches an empty string
1470 (Perl fails the current match path).
1472 (c) A data ] in a character class must be notated as \] because if the
1473 first data character in a class is ], it defines an empty class. (In
1474 Perl it is not possible to have an empty class.) The empty class []
1475 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1476 The negative empty class [^] matches any one character, independently
1477 of the DOTALL setting.
1479 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1480 non-existent subpattern following a character class starting with ']' and
1481 containing () gave an internal compiling error instead of "reference to
1482 non-existent subpattern". Fortunately, when the pattern did exist, the
1483 compiled code was correct. (When scanning forwards to check for the
1484 existencd of the subpattern, it was treating the data ']' as terminating
1485 the class, so got the count wrong. When actually compiling, the reference
1486 was subsequently set up correctly.)
1488 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1489 it was being rejected as not supported by pcre_dfa_exec(), even though
1490 other assertions are supported. I have made pcre_dfa_exec() support
1491 (*FAIL).
1493 16. The implementation of 13c above involved the invention of a new opcode,
1494 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1495 cannot be changed at match time, I realized I could make a small
1496 improvement to matching performance by compiling OP_ALLANY instead of
1497 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1498 on the OP_ANY path.
1500 17. Compiling pcretest on Windows with readline support failed without the
1501 following two fixes: (1) Make the unistd.h include conditional on
1502 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1504 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1505 ncurses library to be included for pcretest when ReadLine support is
1506 requested, but also to allow for it to be overridden. This patch came from
1507 Daniel Bergström.
1509 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1510 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1511 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1512 spotting this.
1515 Version 7.6 28-Jan-08
1516 ---------------------
1518 1. A character class containing a very large number of characters with
1519 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1520 overflow.
1522 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1523 HAVE_LONG_LONG is not defined.
1525 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1526 bring them up to date and include new features. This patch includes:
1528 - Fixed PH's badly added libz and libbz2 support.
1529 - Fixed a problem with static linking.
1530 - Added pcredemo. [But later removed - see 7 below.]
1531 - Fixed dftables problem and added an option.
1532 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1534 - Added readline support for pcretest.
1535 - Added an listing of the option settings after cmake has run.
1537 4. A user submitted a patch to Makefile that makes it easy to create
1538 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1539 Makefile.am that cause it to include this special target, without
1540 affecting anything else. Note that the same mingw target plus all
1541 the other distribution libraries and programs are now supported
1542 when configuring with CMake (see 6 below) instead of with
1543 Configure/Make.
1545 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1546 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1547 exported in the Windows port". It has not yet been confirmed that the patch
1548 solves the problem, but it does no harm.
1550 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1551 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1552 with CMake, and also correct the comment about stack recursion.
1554 7. Remove the automatic building of pcredemo from the ./configure system and
1555 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1556 of a program that users should build themselves after PCRE is installed, so
1557 building it automatically is not really right. What is more, it gave
1558 trouble in some build environments.
1560 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1563 Version 7.5 10-Jan-08
1564 ---------------------
1566 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1567 values in parens when parsing an RE using the C++ wrapper."
1569 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1570 Characters greater than 255 were excluded from the class instead of being
1571 included.
1573 3. The same bug as (2) above applied to negated POSIX classes such as
1574 [:^space:].
1576 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1577 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1578 I have changed it.
1580 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1581 first named subpattern) and a construct such as (?&a) would reference the
1582 first named subpattern whose name started with "a" (in other words, the
1583 length check was missing). Both these problems are fixed. "Subpattern name
1584 expected" is now given for (?&) (a zero-length name), and this patch also
1585 makes it give the same error for \k'' (previously it complained that that
1586 was a reference to a non-existent subpattern).
1588 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1589 this is right because (?- can be followed by option settings as well as by
1590 digits. I have, however, made the messages clearer.
1592 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1593 than the number used in the conditional) now cause a compile-time error.
1594 This is actually not compatible with Perl, which accepts such patterns, but
1595 treats the conditional as always being FALSE (as PCRE used to), but it
1596 seems to me that giving a diagnostic is better.
1598 8. Change "alphameric" to the more common word "alphanumeric" in comments
1599 and messages.
1601 9. Fix two occurrences of "backslash" in comments that should have been
1602 "backspace".
1604 10. Remove two redundant lines of code that can never be obeyed (their function
1605 was moved elsewhere).
1607 11. The program that makes PCRE's Unicode character property table had a bug
1608 which caused it to generate incorrect table entries for sequences of
1609 characters that have the same character type, but are in different scripts.
1610 It amalgamated them into a single range, with the script of the first of
1611 them. In other words, some characters were in the wrong script. There were
1612 thirteen such cases, affecting characters in the following ranges:
1614 U+002b0 - U+002c1
1615 U+0060c - U+0060d
1616 U+0061e - U+00612
1617 U+0064b - U+0065e
1618 U+0074d - U+0076d
1619 U+01800 - U+01805
1620 U+01d00 - U+01d77
1621 U+01d9b - U+01dbf
1622 U+0200b - U+0200f
1623 U+030fc - U+030fe
1624 U+03260 - U+0327f
1625 U+0fb46 - U+0fbb1
1626 U+10450 - U+1049d
1628 12. The -o option (show only the matching part of a line) for pcregrep was not
1629 compatible with GNU grep in that, if there was more than one match in a
1630 line, it showed only the first of them. It now behaves in the same way as
1631 GNU grep.
1633 13. If the -o and -v options were combined for pcregrep, it printed a blank
1634 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1635 does the same. The return code can be used to tell if there were any
1636 non-matching lines.
1638 14. Added --file-offsets and --line-offsets to pcregrep.
1640 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1641 infinitely looping recursion. The bug was that positive lookaheads were not
1642 being skipped when checking for a possible empty match (negative lookaheads
1643 and both kinds of lookbehind were skipped).
1645 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1646 inclusion of <windows.h> to before rather than after the definition of
1647 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1649 17. Specifying a possessive quantifier with a specific limit for a Unicode
1650 character property caused pcre_compile() to compile bad code, which led at
1651 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1652 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1653 caused the error; without that there was no problem.
1655 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1657 19. Added --enable-pcretest-libreadline.
1659 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1660 RE::GlobalReplace(). As a result, the number of replacements returned was
1661 double what it should be. I removed one of the increments, but Craig sent a
1662 later patch that removed the other one (the right fix) and added unit tests
1663 that check the return values (which was not done before).
1665 21. Several CMake things:
1667 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1668 the names libpcre and libpcreposix, not just pcre and pcreposix.
1670 (2) The above change means that pcretest and pcregrep are now correctly
1671 linked with the newly-built libraries, not previously installed ones.
1675 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1676 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1677 UTF-8 newline character). The key issue is that the pattern starts .*;
1678 this means that the match must be either at the beginning, or after a
1679 newline. The bug was in the code for advancing after a failed match and
1680 checking that the new position followed a newline. It was not taking
1681 account of UTF-8 characters correctly.
1683 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1684 character classes. PCRE was not treating the sequence [:...:] as a
1685 character class unless the ... were all letters. Perl, however, seems to
1686 allow any characters between [: and :], though of course it rejects as
1687 unknown any "names" that contain non-letters, because all the known class
1688 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1689 for example, whereas PCRE did not - it did not recognize a POSIX character
1690 class. This seemed a bit dangerous, so the code has been changed to be
1691 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1692 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1693 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1694 Perl does, and where it didn't before.
1696 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1697 Windows environments %n is disabled by default.
1700 Version 7.4 21-Sep-07
1701 ---------------------
1703 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1704 means that a class such as [\s] counted as "explicit reference to CR or
1705 LF". That isn't really right - the whole point of the change was to try to
1706 help when there was an actual mention of one of the two characters. So now
1707 the change happens only if \r or \n (or a literal CR or LF) character is
1708 encountered.
1710 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1711 of both had grown to the point where there were only 3 bits left.
1712 Fortunately, there was spare space in the data structure, and so I have
1713 moved the internal flags into a new 16-bit field to free up more option
1714 bits.
1716 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1717 but did not set the internal JCHANGED flag - either of these is enough to
1718 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1719 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1720 start sets both bits.
1722 4. Added options (at build time, compile time, exec time) to change \R from
1723 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1725 5. doc/pcresyntax.html was missing from the distribution.
1727 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1728 compatibility, even though it is no longer used.
1730 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1731 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1732 windows.h file is present (where different names are used). [This was
1733 reversed later after testing - see 16 below.]
1735 8. Changed all #include <config.h> to #include "config.h". There were also
1736 some further <pcre.h> cases that I changed to "pcre.h".
1738 9. When pcregrep was used with the --colour option, it missed the line ending
1739 sequence off the lines that it output.
1741 10. It was pointed out to me that arrays of string pointers cause lots of
1742 relocations when a shared library is dynamically loaded. A technique of
1743 using a single long string with a table of offsets can drastically reduce
1744 these. I have refactored PCRE in four places to do this. The result is
1745 dramatic:
1747 Originally: 290
1748 After changing UCP table: 187
1749 After changing error message table: 43
1750 After changing table of "verbs" 36
1751 After changing table of Posix names 22
1753 Thanks to the folks working on Gregex for glib for this insight.
1755 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1756 unicode-properties was also set.
1758 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1760 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1761 checked only for CRLF.
1763 14. Added casts to pcretest.c to avoid compiler warnings.
1765 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1767 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1768 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1769 entirely. This removes changes made in 7 above.
1771 17. The CMake files have been updated, and there is now more information about
1772 building with CMake in the NON-UNIX-USE document.
1775 Version 7.3 28-Aug-07
1776 ---------------------
1778 1. In the rejigging of the build system that eventually resulted in 7.1, the
1779 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1780 brackets there is not right, since it causes compilers to look for an
1781 installed pcre.h, not the version that is in the source that is being
1782 compiled (which of course may be different). I have changed it back to:
1784 #include "pcre.h"
1786 I have a vague recollection that the change was concerned with compiling in
1787 different directories, but in the new build system, that is taken care of
1788 by the VPATH setting the Makefile.
1790 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1791 when the subject happened to end in the byte 0x85 (e.g. if the last
1792 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1793 characters but of course it shouldn't be taken as a newline when it is part
1794 of another character. The bug was that, for an unlimited repeat of . in
1795 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1796 characters when looking for a newline.
1798 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1800 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1801 in debug output.
1803 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1804 long printing in the pcrecpp unittest when running under MinGW.
1806 6. ESC_K was left out of the EBCDIC table.
1808 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1809 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1810 limit also applies to "virtual nesting" when a pattern is recursive, and in
1811 this case 1000 isn't so big. I have been able to remove this limit at the
1812 expense of backing off one optimization in certain circumstances. Normally,
1813 when pcre_exec() would call its internal match() function recursively and
1814 immediately return the result unconditionally, it uses a "tail recursion"
1815 feature to save stack. However, when a subpattern that can match an empty
1816 string has an unlimited repetition quantifier, it no longer makes this
1817 optimization. That gives it a stack frame in which to save the data for
1818 checking that an empty string has been matched. Previously this was taken
1819 from the 1000-entry workspace that had been reserved. So now there is no
1820 explicit limit, but more stack is used.
1822 8. Applied Daniel's patches to solve problems with the import/export magic
1823 syntax that is required for Windows, and which was going wrong for the
1824 pcreposix and pcrecpp parts of the library. These were overlooked when this
1825 problem was solved for the main library.
1827 9. There were some crude static tests to avoid integer overflow when computing
1828 the size of patterns that contain repeated groups with explicit upper
1829 limits. As the maximum quantifier is 65535, the maximum group length was
1830 set at 30,000 so that the product of these two numbers did not overflow a
1831 32-bit integer. However, it turns out that people want to use groups that
1832 are longer than 30,000 bytes (though not repeat them that many times).
1833 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1834 made it possible to implement the integer overflow checks in a much more
1835 dynamic way, which I have now done. The artificial limitation on group
1836 length has been removed - we now have only the limit on the total length of
1837 the compiled pattern, which depends on the LINK_SIZE setting.
1839 10. Fixed a bug in the documentation for get/copy named substring when
1840 duplicate names are permitted. If none of the named substrings are set, the
1841 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1842 empty string.
1844 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1845 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1846 because the ] is interpreted as the first data character and the
1847 terminating ] is not found. PCRE has been made compatible with Perl in this
1848 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1849 cause memory overwriting.
1851 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1852 string has been matched (to stop an infinite loop). It was not recognizing
1853 a conditional subpattern that could match an empty string if that
1854 subpattern was within another subpattern. For example, it looped when
1855 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1856 condition was not nested. This bug has been fixed.
1858 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1859 past the start of the subject in the presence of bytes with the top bit
1860 set, for example "\x8aBCD".
1862 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1863 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1865 14. Optimized (?!) to (*FAIL).
1867 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1868 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1869 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1870 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1871 does: it's just the validity check that is more restrictive.
1873 16. Inserted checks for integer overflows during escape sequence (backslash)
1874 processing, and also fixed erroneous offset values for syntax errors during
1875 backslash processing.
1877 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1878 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1880 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1881 caused an overrun.
1883 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1884 something other than just ASCII characters) inside a group that had an
1885 unlimited repeat caused a loop at compile time (while checking to see
1886 whether the group could match an empty string).
1888 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1889 [\P{Any}] did so. (Error in the code for printing property names.)
1891 21. An orphan \E inside a character class could cause a crash.
1893 22. A repeated capturing bracket such as (A)? could cause a wild memory
1894 reference during compilation.
1896 23. There are several functions in pcre_compile() that scan along a compiled
1897 expression for various reasons (e.g. to see if it's fixed length for look
1898 behind). There were bugs in these functions when a repeated \p or \P was
1899 present in the pattern. These operators have additional parameters compared
1900 with \d, etc, and these were not being taken into account when moving along
1901 the compiled data. Specifically:
1903 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1904 length.
1906 (b) An item such as \pL+ within a repeated group could cause crashes or
1907 loops.
1909 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1910 "reference to non-existent subpattern" error.
1912 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1914 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1915 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1917 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1919 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1920 character were causing crashes (broken optimization).
1922 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1923 \p or \P) caused a compile-time loop.
1925 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1926 break. For example, the unstudied pattern [\r\n]A does not match the string
1927 "\r\nA" because change 7.0/46 below moves the current point on by two
1928 characters after failing to match at the start. However, the pattern \nA
1929 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1930 the same is true. There doesn't seem any very clean way out of this, but
1931 what I have chosen to do makes the common cases work: PCRE now takes note
1932 of whether there can be an explicit match for \r or \n anywhere in the
1933 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1934 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1935 pattern has explicit CR or LF references.
1937 29. Added (*CR) etc for changing newline setting at start of pattern.
1940 Version 7.2 19-Jun-07
1941 ---------------------
1943 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1944 which is apparently normally available under Windows.
1946 2. Re-jig the pcregrep tests with different newline settings in an attempt
1947 to make them independent of the local environment's newline setting.
1949 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1951 4. Some of the "internals" tests were previously cut out when the link size
1952 was not 2, because the output contained actual offsets. The recent new
1953 "Z" feature of pcretest means that these can be cut out, making the tests
1954 usable with all link sizes.
1956 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1957 stack recursion. This gives a massive performance boost under BSD, but just
1958 a small improvement under Linux. However, it saves one field in the frame
1959 in all cases.
1961 6. Added more features from the forthcoming Perl 5.10:
1963 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1964 recursion call. It refers to the nth most recently opened parentheses.
1966 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1967 to be opened parentheses.
1969 (c) Conditions that refer to capturing parentheses can be specified
1970 relatively, for example, (?(-2)... or (?(+3)...
1972 (d) \K resets the start of the current match so that everything before
1973 is not part of it.
1975 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1977 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1978 reference syntax.
1980 (g) (?| introduces a group in which the numbering of parentheses in each
1981 alternative starts with the same number.
1983 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1985 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1988 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1989 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1990 for detecting groups that can match an empty string.
1992 9. A pattern with a very large number of alternatives (more than several
1993 hundred) was running out of internal workspace during the pre-compile
1994 phase, where pcre_compile() figures out how much memory will be needed. A
1995 bit of new cunning has reduced the workspace needed for groups with
1996 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1997 workspace instead of running out of the 4096 that are available.
1999 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2001 11. Applied patch from Google to remove an optimization that didn't quite work.
2002 The report of the bug said:
2004 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2005 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2006 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2008 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2009 it matched the wrong number of bytes.
2012 Version 7.1 24-Apr-07
2013 ---------------------
2015 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2016 that is more "standard", making use of automake and other Autotools. There
2017 is some re-arrangement of the files and adjustment of comments consequent
2018 on this.
2020 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2021 for recursive directory scanning broke on some systems because the files
2022 are not scanned in any specific order and on different systems the order
2023 was different. A call to "sort" has been inserted into RunGrepTest for the
2024 approprate test as a short-term fix. In the longer term there may be an
2025 alternative.
2027 3. I had an email from Eric Raymond about problems translating some of PCRE's
2028 man pages to HTML (despite the fact that I distribute HTML pages, some
2029 people do their own conversions for various reasons). The problems
2030 concerned the use of low-level troff macros .br and .in. I have therefore
2031 removed all such uses from the man pages (some were redundant, some could
2032 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2033 HTML has been updated to handle .nf/.fi and to complain if it encounters
2034 .br or .in.
2036 4. Updated comments in configure.ac that get placed in config.h.in and also
2037 arranged for config.h to be included in the distribution, with the name
2038 config.h.generic, for the benefit of those who have to compile without
2039 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2041 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2042 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2043 makevp.bat for the current PCRE, using the additional files
2044 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2046 6. A Windows user reported a minor discrepancy with test 2, which turned out
2047 to be caused by a trailing space on an input line that had got lost in his
2048 copy. The trailing space was an accident, so I've just removed it.
2050 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2051 that is needed.
2053 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2054 as "const" (a) because they are and (b) because it helps the PHP
2055 maintainers who have recently made a script to detect big data structures
2056 in the php code that should be moved to the .rodata section. I remembered
2057 to update Builducptable as well, so it won't revert if ucptable.h is ever
2058 re-created.
2060 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2061 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2062 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2063 support is not required. This saves 1.5-2K of code, which is important in
2064 some applications.
2066 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2067 so as not to refer to the tables, even though these functions will never be
2068 called when UTF-8 support is disabled. Otherwise there are problems with a
2069 shared library.
2071 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2073 (a) It was defining its arguments as char * instead of void *.
2075 (b) It was assuming that all moves were upwards in memory; this was true
2076 a long time ago when I wrote it, but is no longer the case.
2078 The emulated memove() is provided for those environments that have neither
2079 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2080 is clearly not the case, as these two bugs were recently reported.
2082 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2083 and Detrail to create the HTML documentation, the .txt form of the man
2084 pages, and it removes trailing spaces from listed files. It also creates
2085 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2086 case, it wraps all the #defines with #ifndefs. This script should be run
2087 before "make dist".
2089 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2090 with Unicode property support.
2092 (a) For a maximizing quantifier, if the two different cases of the
2093 character were of different lengths in their UTF-8 codings (there are
2094 some cases like this - I found 11), and the matching function had to
2095 back up over a mixture of the two cases, it incorrectly assumed they
2096 were both the same length.
2098 (b) When PCRE was configured to use the heap rather than the stack for
2099 recursion during matching, it was not correctly preserving the data for
2100 the other case of a UTF-8 character when checking ahead for a match
2101 while processing a minimizing repeat. If the check also involved
2102 matching a wide character, but failed, corruption could cause an
2103 erroneous result when trying to check for a repeat of the original
2104 character.
2106 13. Some tidying changes to the testing mechanism:
2108 (a) The RunTest script now detects the internal link size and whether there
2109 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2110 values substituted by "configure". (The RunGrepTest script already did
2111 this for UTF-8.) The configure.ac script no longer substitutes the
2112 relevant variables.
2114 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2115 with length and offset values. This means that the output is different
2116 for different internal link sizes. Test 2 is skipped for link sizes
2117 other than 2 because of this, bypassing the problem. Unfortunately,
2118 there was also a test in test 3 (the locale tests) that used /B and
2119 failed for link sizes other than 2. Rather than cut the whole test out,
2120 I have added a new /Z option to pcretest that replaces the length and
2121 offset values with spaces. This is now used to make test 3 independent
2122 of link size. (Test 2 will be tidied up later.)
2124 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2125 segmentation fault instead of returning the appropriate error message.
2127 15. In multiline mode when the newline sequence was set to "any", the pattern
2128 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2129 This doesn't seem right; it now treats the CRLF combination as the line
2130 ending, and so does not match in that case. It's only a pattern such as ^$
2131 that would hit this one: something like ^ABC$ would have failed after \r
2132 and then tried again after \r\n.
2134 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2135 in an attempt to make files that differ only in their line terminators
2136 compare equal. This works on Linux.
2138 17. Under certain error circumstances pcregrep might try to free random memory
2139 as it exited. This is now fixed, thanks to valgrind.
2141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2143 was because its rules for how to advance for /g after matching an empty
2144 string at the end of a line did not allow for this case. They now check for
2145 it specially.
2147 20. pcretest is supposed to handle patterns and data of any length, by
2148 extending its buffers when necessary. It was getting this wrong when the
2149 buffer for a data line had to be extended.
2151 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2152 CRLF as a newline sequence.
2154 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2155 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2156 I have nevertheless tidied it up.
2158 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2160 24. Added a man page for pcre-config.
2163 Version 7.0 19-Dec-06
2164 ---------------------
2166 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2167 moving to gcc 4.1.1.
2169 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2170 sys/time.h, which is documented as needed for this function. It doesn't
2171 seem to matter on Linux, but it showed up on some releases of OS X.
2173 3. It seems that there are systems where bytes whose values are greater than
2174 127 match isprint() in the "C" locale. The "C" locale should be the
2175 default when a C program starts up. In most systems, only ASCII printing
2176 characters match isprint(). This difference caused the output from pcretest
2177 to vary, making some of the tests fail. I have changed pcretest so that:
2179 (a) When it is outputting text in the compiled version of a pattern, bytes
2180 other than 32-126 are always shown as hex escapes.
2182 (b) When it is outputting text that is a matched part of a subject string,
2183 it does the same, unless a different locale has been set for the match
2184 (using the /L modifier). In this case, it uses isprint() to decide.
2186 4. Fixed a major bug that caused incorrect computation of the amount of memory
2187 required for a compiled pattern when options that changed within the
2188 pattern affected the logic of the preliminary scan that determines the
2189 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2190 that the computed length was too small. The symptoms of this bug were
2191 either the PCRE error "internal error: code overflow" from pcre_compile(),
2192 or a glibc crash with a message such as "pcretest: free(): invalid next
2193 size (fast)". Examples of patterns that provoked this bug (shown in
2194 pcretest format) are:
2196 /(?-x: )/x
2197 /(?x)(?-x: \s*#\s*)/
2198 /((?i)[\x{c0}])/8
2199 /(?i:[\x{c0}])/8
2201 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2202 is now done differently.
2204 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2205 wrapper classes; (b) implement a new function in the C++ scanner that is
2206 more efficient than the old way of doing things because it avoids levels of
2207 recursion in the regex matching; (c) add a paragraph to the documentation
2208 for the FullMatch() function.
2210 6. The escape sequence \n was being treated as whatever was defined as
2211 "newline". Not only was this contrary to the documentation, which states
2212 that \n is character 10 (hex 0A), but it also went horribly wrong when
2213 "newline" was defined as CRLF. This has been fixed.
2215 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2216 was being set to -1 for the "end of line" case (supposedly a value that no
2217 character can have). Though this value is never used (the check for end of
2218 line is "zero bytes in current character"), it caused compiler complaints.
2219 I've changed it to 0xffffffff.
2221 8. In pcre_version.c, the version string was being built by a sequence of
2222 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2223 string (as it is for production releases) called a macro with an empty
2224 argument. The C standard says the result of this is undefined. The gcc
2225 compiler treats it as an empty string (which was what was wanted) but it is
2226 reported that Visual C gives an error. The source has been hacked around to
2227 avoid this problem.
2229 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2230 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2231 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2232 of them did).
2234 10. Originally, pcretest opened its input and output without "b"; then I was
2235 told that "b" was needed in some environments, so it was added for release
2236 5.0 to both the input and output. (It makes no difference on Unix-like
2237 systems.) Later I was told that it is wrong for the input on Windows. I've
2238 now abstracted the modes into two macros, to make it easier to fiddle with
2239 them, and removed "b" from the input mode under Windows.
2241 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2243 12. Added -help and --help to pcretest as an official way of being reminded
2244 of the options.
2246 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2247 and pcrecpp.cc because they annoy compilers at high warning levels.
2249 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2251 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2252 scripts are not C programs :-) and which was not noticed because it works
2253 on Linux.
2255 16. pcretest is supposed to handle any length of pattern and data line (as one
2256 line or as a continued sequence of lines) by extending its input buffer if
2257 necessary. This feature was broken for very long pattern lines, leading to
2258 a string of junk being passed to pcre_compile() if the pattern was longer
2259 than about 50K.
2261 17. I have done a major re-factoring of the way pcre_compile() computes the
2262 amount of memory needed for a compiled pattern. Previously, there was code
2263 that made a preliminary scan of the pattern in order to do this. That was
2264 OK when PCRE was new, but as the facilities have expanded, it has become
2265 harder and harder to keep it in step with the real compile phase, and there
2266 have been a number of bugs (see for example, 4 above). I have now found a
2267 cunning way of running the real compile function in a "fake" mode that
2268 enables it to compute how much memory it would need, while actually only
2269 ever using a few hundred bytes of working memory and without too many
2270 tests of the mode. This should make future maintenance and development
2271 easier. A side effect of this work is that the limit of 200 on the nesting
2272 depth of parentheses has been removed (though this was never a serious
2273 limitation, I suspect). However, there is a downside: pcre_compile() now
2274 runs more slowly than before (30% or more, depending on the pattern). I
2275 hope this isn't a big issue. There is no effect on runtime performance.
2277 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2278 newline (only possible for the last line of a file) and it was a
2279 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2281 19. Added additional timing features to pcretest. (1) The -tm option now times
2282 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2283 separate command line item, by a number that specifies the number of
2284 repeats to use when timing. The default is 50000; this gives better
2285 precision, but takes uncomfortably long for very large patterns.
2287 20. Extended pcre_study() to be more clever in cases where a branch of a
2288 subpattern has no definite first character. For example, (a*|b*)[cd] would
2289 previously give no result from pcre_study(). Now it recognizes that the
2290 first character must be a, b, c, or d.
2292 21. There was an incorrect error "recursive call could loop indefinitely" if
2293 a subpattern (or the entire pattern) that was being tested for matching an
2294 empty string contained only one non-empty item after a nested subpattern.
2295 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2296 incorrectly, because the \d was being skipped in the check.
2298 22. The pcretest program now has a new pattern option /B and a command line
2299 option -b, which is equivalent to adding /B to every pattern. This causes
2300 it to show the compiled bytecode, without the additional information that
2301 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2302 is the same as /B/I).
2304 23. A new optimization is now able automatically to treat some sequences such
2305 as a*b as a*+b. More specifically, if something simple (such as a character
2306 or a simple class like \d) has an unlimited quantifier, and is followed by
2307 something that cannot possibly match the quantified thing, the quantifier
2308 is automatically "possessified".
2310 24. A recursive reference to a subpattern whose number was greater than 39
2311 went wrong under certain circumstances in UTF-8 mode. This bug could also
2312 have affected the operation of pcre_study().
2314 25. Realized that a little bit of performance could be had by replacing
2315 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2317 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2319 27. Possessive quantifiers such as a++ were previously implemented by turning
2320 them into atomic groups such as ($>a+). Now they have their own opcodes,
2321 which improves performance. This includes the automatically created ones
2322 from 23 above.
2324 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2325 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2326 the first matched character to be a colon. This applied both to named and
2327 numbered groups.
2329 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2331 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2332 building PCRE on Windows easier, so I have included it in the distribution.
2334 31. There is now a check in pcretest against a ridiculously large number being
2335 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2336 loop, the loop is abandoned.
2338 32. Forward references to subpatterns in conditions such as (?(2)...) where
2339 subpattern 2 is defined later cause pcre_compile() to search forwards in
2340 the pattern for the relevant set of parentheses. This search went wrong
2341 when there were unescaped parentheses in a character class, parentheses
2342 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2344 33. "Subroutine" calls and backreferences were previously restricted to
2345 referencing subpatterns earlier in the regex. This restriction has now
2346 been removed.
2348 34. Added a number of extra features that are going to be in Perl 5.10. On the
2349 whole, these are just syntactic alternatives for features that PCRE had
2350 previously implemented using the Python syntax or my own invention. The
2351 other formats are all retained for compatibility.
2353 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2354 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2355 also .NET compatible.
2357 (b) A recursion or subroutine call to a named group can now be defined as
2358 (?&name) as well as (?P>name).
2360 (c) A backreference to a named group can now be defined as \k<name> or
2361 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2362 5.10, are also .NET compatible.
2364 (d) A conditional reference to a named group can now use the syntax
2365 (?(<name>) or (?('name') as well as (?(name).
2367 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2368 groups (named and numbered) that are never evaluated inline, but can be
2369 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2370 is always false. There may be only one alternative in such a group.
2372 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2373 as the simple (?(R). The condition is true only if the most recent
2374 recursion is that of the given number or name. It does not search out
2375 through the entire recursion stack.
2377 (g) The escape \gN or \g{N} has been added, where N is a positive or
2378 negative number, specifying an absolute or relative reference.
2380 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2381 some "unreachable code" warnings.
2383 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2384 things, this adds five new scripts.
2386 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2387 There were also incompatibilities regarding the handling of \Q..\E inside
2388 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2389 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2391 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2392 matches an empty string, and forcibly breaks the loop. There were bugs in
2393 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2394 against aaaa the result was just "a" rather than "aaaa", for example. Two
2395 separate and independent bugs (that affected different cases) have been
2396 fixed.
2398 39. Refactored the code to abolish the use of different opcodes for small
2399 capturing bracket numbers. This is a tidy that I avoided doing when I
2400 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2401 The new approach is not only tidier, it makes it possible to reduce the
2402 memory needed to fix the previous bug (38).
2404 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2405 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2406 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2407 mode.
2409 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2410 report.
2412 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2413 copy construction and assignment in the C++ wrapper.
2415 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2416 couple of bugs that could have given wrong results in the "--newline=crlf"
2417 case.
2419 44. Added a number of casts and did some reorganization of signed/unsigned int
2420 variables following suggestions from Dair Grant. Also renamed the variable
2421 "this" as "item" because it is a C++ keyword.
2423 45. Arranged for dftables to add
2425 #include "pcre_internal.h"
2427 to pcre_chartables.c because without it, gcc 4.x may remove the array
2428 definition from the final binary if PCRE is built into a static library and
2429 dead code stripping is activated.
2431 46. For an unanchored pattern, if a match attempt fails at the start of a
2432 newline sequence, and the newline setting is CRLF or ANY, and the next two
2433 characters are CRLF, advance by two characters instead of one.
2436 Version 6.7 04-Jul-06
2437 ---------------------
2439 1. In order to handle tests when input lines are enormously long, pcretest has
2440 been re-factored so that it automatically extends its buffers when
2441 necessary. The code is crude, but this _is_ just a test program. The
2442 default size has been increased from 32K to 50K.
2444 2. The code in pcre_study() was using the value of the re argument before
2445 testing it for NULL. (Of course, in any sensible call of the function, it
2446 won't be NULL.)
2448 3. The memmove() emulation function in pcre_internal.h, which is used on
2449 systems that lack both memmove() and bcopy() - that is, hardly ever -
2450 was missing a "static" storage class specifier.
2452 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2453 containing an extended class (one that cannot be represented by a bitmap
2454 because it contains high-valued characters or Unicode property items, e.g.
2455 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2456 pattern, but PCRE should not loop if you do not (it no longer does).
2457 [Detail: two cases were found: (a) a repeated subpattern containing an
2458 extended class; (b) a recursive reference to a subpattern that followed a
2459 previous extended class. It wasn't skipping over the extended class
2460 correctly when UTF-8 mode was not set.]
2462 5. A negated single-character class was not being recognized as fixed-length
2463 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2464 compile error "lookbehind assertion is not fixed length".
2466 6. The RunPerlTest auxiliary script was showing an unexpected difference
2467 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2468 write a Perl script that can interpret lines of an input file either as
2469 byte characters or as UTF-8, which is what "perltest" was being required to
2470 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2471 can't do is switch easily at run time between having the "use utf8;" pragma
2472 or not. In the end, I fudged it by using the RunPerlTest script to insert
2473 "use utf8;" explicitly for the UTF-8 tests.
2475 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2476 the end of the subject string, contrary to the documentation and to what
2477 Perl does. This was true of both matching functions. Now it matches only at
2478 the start of the subject and immediately after *internal* newlines.
2480 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2481 a pointer to an int instead of a pointer to an unsigned long int. This
2482 caused problems on 64-bit systems.
2484 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2485 instance of the 'standard' template library not being so standard".
2487 10. There was no check on the number of named subpatterns nor the maximum
2488 length of a subpattern name. The product of these values is used to compute
2489 the size of the memory block for a compiled pattern. By supplying a very
2490 long subpattern name and a large number of named subpatterns, the size
2491 computation could be caused to overflow. This is now prevented by limiting
2492 the length of names to 32 characters, and the number of named subpatterns
2493 to 10,000.
2495 11. Subpatterns that are repeated with specific counts have to be replicated in
2496 the compiled pattern. The size of memory for this was computed from the
2497 length of the subpattern and the repeat count. The latter is limited to
2498 65535, but there was no limit on the former, meaning that integer overflow
2499 could in principle occur. The compiled length of a repeated subpattern is
2500 now limited to 30,000 bytes in order to prevent this.
2502 12. Added the optional facility to have named substrings with the same name.
2504 13. Added the ability to use a named substring as a condition, using the
2505 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2506 are numbers (not recommended). Forward references are permitted.
2508 14. Added forward references in named backreferences (if you see what I mean).
2510 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2511 pattern could run off the end of the subject. For example, the pattern
2512 "(?s)(.{1,5})"8 did this with the subject "ab".
2514 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2515 PCRE_CASELESS was set when matching characters that were quantified with ?
2516 or *.
2518 17. A character class other than a single negated character that had a minimum
2519 but no maximum quantifier - for example [ab]{6,} - was not handled
2520 correctly by pce_dfa_exec(). It would match only one character.
2522 18. A valid (though odd) pattern that looked like a POSIX character
2523 class but used an invalid character after [ (for example [[,abc,]]) caused
2524 pcre_compile() to give the error "Failed: internal error: code overflow" or
2525 in some cases to crash with a glibc free() error. This could even happen if
2526 the pattern terminated after [[ but there just happened to be a sequence of
2527 letters, a binary zero, and a closing ] in the memory that followed.
2529 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2530 over the years. Originally (before any Unicode support), just the bottom 8
2531 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2532 output from "man perlunicode" includes this:
2534 The regular expression compiler produces polymorphic opcodes. That
2535 is, the pattern adapts to the data and automatically switches to
2536 the Unicode character scheme when presented with Unicode data--or
2537 instead uses a traditional byte scheme when presented with byte
2538 data.
2540 Sadly, a wide octal escape does not cause a switch, and in a string with
2541 no other multibyte characters, these octal escapes are treated as before.
2542 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2543 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2544 Unicode string.
2546 I have not perpetrated such confusion in PCRE. Up till now, it took just
2547 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2548 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2549 translate to the appropriate multibyte character.
2551 29. Applied some refactoring to reduce the number of warnings from Microsoft
2552 and Borland compilers. This has included removing the fudge introduced
2553 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2554 a warning about an unused variable.
2556 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2557 characters since release 4.0, because Perl (from release 5.004) does not.
2558 [Or at least, is documented not to: some releases seem to be in conflict
2559 with the documentation.] However, when a pattern was studied with
2560 pcre_study() and all its branches started with \s, PCRE still included VT
2561 as a possible starting character. Of course, this did no harm; it just
2562 caused an unnecessary match attempt.
2564 22. Removed a now-redundant internal flag bit that recorded the fact that case
2565 dependency changed within the pattern. This was once needed for "required
2566 byte" processing, but is no longer used. This recovers a now-scarce options
2567 bit. Also moved the least significant internal flag bit to the most-
2568 significant bit of the word, which was not previously used (hangover from
2569 the days when it was an int rather than a uint) to free up another bit for
2570 the future.
2572 23. Added support for CRLF line endings as well as CR and LF. As well as the
2573 default being selectable at build time, it can now be changed at runtime
2574 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2575 specify that it is scanning data with non-default line endings.
2577 24. Changed the definition of CXXLINK to make it agree with the definition of
2578 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2580 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2581 recursions. This makes a big different to stack usage for some patterns.
2583 26. If a subpattern containing a named recursion or subroutine reference such
2584 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2585 the space required for the compiled pattern went wrong and gave too small a
2586 value. Depending on the environment, this could lead to "Failed: internal
2587 error: code overflow at offset 49" or "glibc detected double free or
2588 corruption" errors.
2590 27. Applied patches from Google (a) to support the new newline modes and (b) to
2591 advance over multibyte UTF-8 characters in GlobalReplace.
2593 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2594 difference for some implementation of PCRE in some Windows version.
2596 29. Added some extra testing facilities to pcretest:
2598 \q<number> in a data line sets the "match limit" value
2599 \Q<number> in a data line sets the "match recursion limt" value
2600 -S <number> sets the stack size, where <number> is in megabytes
2602 The -S option isn't available for Windows.
2605 Version 6.6 06-Feb-06
2606 ---------------------
2608 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2609 in pcreposix.h. I have copied the definition from pcre.h.
2611 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2612 because pcre.h is no longer a built file.
2614 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2615 not normally included in the compiled code.
2618 Version 6.5 01-Feb-06
2619 ---------------------
2621 1. When using the partial match feature with pcre_dfa_exec(), it was not
2622 anchoring the second and subsequent partial matches at the new starting
2623 point. This could lead to incorrect results. For example, with the pattern
2624 /1234/, partially matching against "123" and then "a4" gave a match.
2626 2. Changes to pcregrep:
2628 (a) All non-match returns from pcre_exec() were being treated as failures
2629 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2630 error message is output. Some extra information is given for the
2632 probably the only errors that are likely to be caused by users (by
2633 specifying a regex that has nested indefinite repeats, for instance).
2634 If there are more than 20 of these errors, pcregrep is abandoned.
2636 (b) A binary zero was treated as data while matching, but terminated the
2637 output line if it was written out. This has been fixed: binary zeroes
2638 are now no different to any other data bytes.
2640 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2641 used to set a locale for matching. The --locale=xxxx long option has
2642 been added (no short equivalent) to specify a locale explicitly on the
2643 pcregrep command, overriding the environment variables.
2645 (d) When -B was used with -n, some line numbers in the output were one less
2646 than they should have been.
2648 (e) Added the -o (--only-matching) option.
2650 (f) If -A or -C was used with -c (count only), some lines of context were
2651 accidentally printed for the final match.
2653 (g) Added the -H (--with-filename) option.
2655 (h) The combination of options -rh failed to suppress file names for files
2656 that were found from directory arguments.
2658 (i) Added the -D (--devices) and -d (--directories) options.
2660 (j) Added the -F (--fixed-strings) option.
2662 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2664 (l) Added the --colo(u)r option.
2666 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2667 is not present by default.
2669 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2670 items such as (?R) or (?1), when the recursion could match a number of
2671 alternatives. If it matched one of the alternatives, but subsequently,
2672 outside the recursion, there was a failure, the code tried to back up into
2673 the recursion. However, because of the way PCRE is implemented, this is not
2674 possible, and the result was an incorrect result from the match.
2676 In order to prevent this happening, the specification of recursion has
2677 been changed so that all such subpatterns are automatically treated as
2678 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2680 4. I had overlooked the fact that, in some locales, there are characters for
2681 which isalpha() is true but neither isupper() nor islower() are true. In
2682 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2683 and ordfeminine) are like this. This affected the treatment of \w and \W
2684 when they appeared in character classes, but not when they appeared outside
2685 a character class. The bit map for "word" characters is now created
2686 separately from the results of isalnum() instead of just taking it from the
2687 upper, lower, and digit maps. (Plus the underscore character, of course.)
2689 5. The above bug also affected the handling of POSIX character classes such as
2690 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2691 permanent tables. Instead, the bit maps for such a class were previously
2692 created as the appropriate unions of the upper, lower, and digit bitmaps.
2693 Now they are created by subtraction from the [[:word:]] class, which has
2694 its own bitmap.
2696 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2697 It is created by subtracting the vertical space characters (\x09, \x0a,
2698 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2699 subtraction was done in the overall bitmap for a character class, meaning
2700 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2701 be recognized. This bug has been fixed.
2703 7. Patches from the folks at Google:
2705 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2706 real life, but is still worth protecting against".
2708 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2709 regular expressions".
2711 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2712 have it.
2714 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2715 "configure" and the latter not, in order to fix a problem somebody had
2716 with compiling the Arg class on HP-UX.
2718 (e) Improve the error-handling of the C++ wrapper a little bit.
2720 (f) New tests for checking recursion limiting.
2722 8. The pcre_memmove() function, which is used only if the environment does not
2723 have a standard memmove() function (and is therefore rarely compiled),
2724 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2725 returning a result (though PCRE never actually uses the result).
2727 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2728 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2729 returned instead of calling malloc() with an overflowing number that would
2730 most likely cause subsequent chaos.
2732 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2734 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2735 with this option is matched, the nmatch and pmatch options of regexec() are
2736 ignored.
2738 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2739 provided in case anyone wants to the the POSIX interface with UTF-8
2740 strings.
2742 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2743 C++ linking (needed for some HP-UX environments).
2745 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2746 (unused parameter) and in the pcre_printint() function (omitted "default"
2747 switch label when the default is to do nothing).
2749 15. Added some code to make it possible, when PCRE is compiled as a C++
2750 library, to replace subject pointers for pcre_exec() with a smart pointer
2751 class, thus making it possible to process discontinuous strings.
2753 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2754 much the same function. They were added by different people who were trying
2755 to make PCRE easy to compile on non-Unix systems. It has been suggested
2756 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2757 for compiling on Windows systems. I have therefore replaced it with
2758 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2759 defaults to "extern" for C or "extern C" for C++, which works fine on
2760 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2761 SCOPE with something explicit in config.h. In addition:
2763 (a) pcreposix.h still had just "extern" instead of either of these macros;
2764 I have replaced it with PCRE_DATA_SCOPE.
2766 (b) Functions such as _pcre_xclass(), which are internal to the library,
2767 but external in the C sense, all had PCRE_EXPORT in their definitions.
2768 This is apparently wrong for the Windows case, so I have removed it.
2769 (It makes no difference on Unix-like systems.)
2771 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2772 of recursive calls to match(). This is different to MATCH_LIMIT because
2773 that limits the total number of calls to match(), not all of which increase
2774 the depth of recursion. Limiting the recursion depth limits the amount of
2775 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2776 when PCRE is compiled, and changed at run time. A patch from Google adds
2777 this functionality to the C++ interface.
2779 18. Changes to the handling of Unicode character properties:
2781 (a) Updated the table to Unicode 4.1.0.
2783 (b) Recognize characters that are not in the table as "Cn" (undefined).
2785 (c) I revised the way the table is implemented to a much improved format
2786 which includes recognition of ranges. It now supports the ranges that
2787 are defined in UnicodeData.txt, and it also amalgamates other
2788 characters into ranges. This has reduced the number of entries in the
2789 table from around 16,000 to around 3,000, thus reducing its size
2790 considerably. I realized I did not need to use a tree structure after
2791 all - a binary chop search is just as efficient. Having reduced the
2792 number of entries, I extended their size from 6 bytes to 8 bytes to
2793 allow for more data.
2795 (d) Added support for Unicode script names via properties such as \p{Han}.
2797 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2798 matching that character.
2800 20. When matching a repeated Unicode property with a minimum greater than zero,
2801 (for example \pL{2,}), PCRE could look past the end of the subject if it
2802 reached it while seeking the minimum number of characters. This could
2803 happen only if some of the characters were more than one byte long, because
2804 there is a check for at least the minimum number of bytes.
2806 21. Refactored the implementation of \p and \P so as to be more general, to
2807 allow for more different types of property in future. This has changed the
2808 compiled form incompatibly. Anybody with saved compiled patterns that use
2809 \p or \P will have to recompile them.
2811 22. Added "Any" and "L&" to the supported property types.
2813 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2814 but give a compile time error if the value is greater than 0xff.
2816 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2817 accidentally not being installed or uninstalled.
2819 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2820 made were to insert the current release number. This seemed silly, because
2821 it made things harder for people building PCRE on systems that don't run
2822 "configure". I have turned pcre.h into a distributed file, no longer built
2823 by "configure", with the version identification directly included. There is
2824 no longer a pcre.h.in file.
2826 However, this change necessitated a change to the pcre-config script as
2827 well. It is built from pcre-config.in, and one of the substitutions was the
2828 release number. I have updated configure.ac so that ./configure now finds
2829 the release number by grepping pcre.h.
2831 26. Added the ability to run the tests under valgrind.
2834 Version 6.4 05-Sep-05
2835 ---------------------
2837 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2838 "--" to be printed when multiple files were scanned, even when none of the
2839 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2840 consider it to be a bug, and have restored the previous behaviour.
2842 2. A couple of code tidies to get rid of compiler warnings.
2844 3. The pcretest program used to cheat by referring to symbols in the library
2845 whose names begin with _pcre_. These are internal symbols that are not
2846 really supposed to be visible externally, and in some environments it is
2847 possible to suppress them. The cheating is now confined to including
2848 certain files from the library's source, which is a bit cleaner.
2850 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2851 file's purpose clearer.
2853 5. Reorganized pcre_ucp_findchar().
2856 Version 6.3 15-Aug-05
2857 ---------------------
2859 1. The file libpcre.pc.in did not have general read permission in the tarball.
2861 2. There were some problems when building without C++ support:
2863 (a) If C++ support was not built, "make install" and "make test" still
2864 tried to test it.
2866 (b) There were problems when the value of CXX was explicitly set. Some
2867 changes have been made to try to fix these, and ...
2869 (c) --disable-cpp can now be used to explicitly disable C++ support.
2871 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2872 backslash in a target when C++ was disabled. This confuses some
2873 versions of "make", apparently. Using an intermediate variable solves
2874 this. (Same for CPP_LOBJ.)
2876 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2877 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2878 necessary on certain architectures.
2880 4. Added a setting of -export-symbols-regex to the link command to remove
2881 those symbols that are exported in the C sense, but actually are local
2882 within the library, and not documented. Their names all begin with
2883 "_pcre_". This is not a perfect job, because (a) we have to except some
2884 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2885 available (and never for static libraries). I have made a note to try to
2886 find a way round (a) in the future.
2889 Version 6.2 01-Aug-05
2890 ---------------------
2892 1. There was no test for integer overflow of quantifier values. A construction
2893 such as {1111111111111111} would give undefined results. What is worse, if
2894 a minimum quantifier for a parenthesized subpattern overflowed and became
2895 negative, the calculation of the memory size went wrong. This could have
2896 led to memory overwriting.
2898 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2900 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2901 operating environments where this matters.
2903 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2904 PCRE options from within the C++ wrapper.
2906 5. Named capturing subpatterns were not being correctly counted when a pattern
2907 was compiled. This caused two problems: (a) If there were more than 100
2908 such subpatterns, the calculation of the memory needed for the whole
2909 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2910 back references of the form \12, where the number was greater than 9, were
2911 not recognized as back references, even though there were sufficient
2912 previous subpatterns.
2914 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2915 versions of gcc, e.g. 2.95.4.
2918 Version 6.1 21-Jun-05
2919 ---------------------
2921 1. There was one reference to the variable "posix" in pcretest.c that was not
2922 surrounded by "#if !defined NOPOSIX".
2924 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2925 the cross-check on the old pcre_info() function, for the benefit of the
2926 cut-down version of PCRE that is currently imported into Exim.
2928 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2929 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2930 patterns that start (?i) but I don't think that matters. The use of (?i) is
2931 just an example; this all applies to the other options as well.
2933 4. Since libtool seems to echo the compile commands it is issuing, the output
2934 from "make" can be reduced a bit by putting "@" in front of each libtool
2935 compile command.
2937 5. Patch from the folks at Google for configure.in to be a bit more thorough
2938 in checking for a suitable C++ installation before trying to compile the
2939 C++ stuff. This should fix a reported problem when a compiler was present,
2940 but no suitable headers.
2942 6. The man pages all had just "PCRE" as their title. I have changed them to
2943 be the relevant file name. I have also arranged that these names are
2944 retained in the file doc/pcre.txt, which is a concatenation in text format
2945 of all the man pages except the little individual ones for each function.
2947 7. The NON-UNIX-USE file had not been updated for the different set of source
2948 files that come with release 6. I also added a few comments about the C++
2949 wrapper.
2952 Version 6.0 07-Jun-05
2953 ---------------------
2955 1. Some minor internal re-organization to help with my DFA experiments.
2957 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2958 didn't matter for the library itself when fully configured, but did matter
2959 when compiling without UCP support, or within Exim, where the ucp files are
2960 not imported.
2962 3. Refactoring of the library code to split up the various functions into
2963 different source modules. The addition of the new DFA matching code (see
2964 below) to a single monolithic source would have made it really too
2965 unwieldy, quite apart from causing all the code to be include in a
2966 statically linked application, when only some functions are used. This is
2967 relevant even without the DFA addition now that patterns can be compiled in
2968 one application and matched in another.
2970 The downside of splitting up is that there have to be some external
2971 functions and data tables that are used internally in different modules of
2972 the library but which are not part of the API. These have all had their
2973 names changed to start with "_pcre_" so that they are unlikely to clash
2974 with other external names.
2976 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2977 a different (DFA) algorithm. Although it is slower than the original
2978 function, it does have some advantages for certain types of matching
2979 problem.
2981 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2982 including restarting after a partial match.
2984 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2985 defined when compiling for Windows was sent to me. I have put it into the
2986 code, though I have no means of testing or verifying it.
2988 7. Added the pcre_refcount() auxiliary function.
2990 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2991 match before or at the first newline in the subject string. In pcretest,
2992 the /f option on a pattern can be used to set this.
2994 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2995 would behave wrongly. This has been present in PCRE since release 4.0.
2997 10. A number of changes to the pcregrep command:
2999 (a) Refactored how -x works; insert ^(...)$ instead of setting
3000 PCRE_ANCHORED and checking the length, in preparation for adding
3001 something similar for -w.
3003 (b) Added the -w (match as a word) option.
3005 (c) Refactored the way lines are read and buffered so as to have more
3006 than one at a time available.
3008 (d) Implemented a pcregrep test script.
3010 (e) Added the -M (multiline match) option. This allows patterns to match
3011 over several lines of the subject. The buffering ensures that at least
3012 8K, or the rest of the document (whichever is the shorter) is available
3013 for matching (and similarly the previous 8K for lookbehind assertions).
3015 (f) Changed the --help output so that it now says
3017 -w, --word-regex(p)
3019 instead of two lines, one with "regex" and the other with "regexp"
3020 because that confused at least one person since the short forms are the
3021 same. (This required a bit of code, as the output is generated
3022 automatically from a table. It wasn't just a text change.)
3024 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3025 option but starts with a hyphen. Could be a pattern or a path name
3026 starting with a hyphen, for instance.
3028 (h) "-" can be given as a file name to represent stdin.
3030 (i) When file names are being printed, "(standard input)" is used for
3031 the standard input, for compatibility with GNU grep. Previously
3032 "<stdin>" was used.
3034 (j) The option --label=xxx can be used to supply a name to be used for
3035 stdin when file names are being printed. There is no short form.
3037 (k) Re-factored the options decoding logic because we are going to add
3038 two more options that take data. Such options can now be given in four
3039 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3041 (l) Added the -A, -B, and -C options for requesting that lines of context
3042 around matches be printed.
3044 (m) Added the -L option to print the names of files that do not contain
3045 any matching lines, that is, the complement of -l.
3047 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3048 continue to scan other files.
3050 (o) The -s option was incorrectly implemented. For compatibility with other
3051 greps, it now suppresses the error message for a non-existent or non-
3052 accessible file (but not the return code). There is a new option called
3053 -q that suppresses the output of matching lines, which was what -s was
3054 previously doing.
3056 (p) Added --include and --exclude options to specify files for inclusion
3057 and exclusion when recursing.
3059 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3060 Hopefully, it now does.
3062 12. Missing cast in pcre_study().
3064 13. Added an "uninstall" target to the makefile.
3066 14. Replaced "extern" in the function prototypes in Makefile.in with
3067 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3068 world, but is set differently for Windows.
3070 15. Added a second compiling function called pcre_compile2(). The only
3071 difference is that it has an extra argument, which is a pointer to an
3072 integer error code. When there is a compile-time failure, this is set
3073 non-zero, in addition to the error test pointer being set to point to an
3074 error message. The new argument may be NULL if no error number is required
3075 (but then you may as well call pcre_compile(), which is now just a
3076 wrapper). This facility is provided because some applications need a
3077 numeric error indication, but it has also enabled me to tidy up the way
3078 compile-time errors are handled in the POSIX wrapper.
3080 16. Added VPATH=.libs to the makefile; this should help when building with one
3081 prefix path and installing with another. (Or so I'm told by someone who
3082 knows more about this stuff than I do.)
3084 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3085 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3086 match everything, including newlines. This is not POSIX-compatible, but
3087 somebody wanted the feature. From pcretest it can be activated by using
3088 both the P and the s flags.
3090 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3092 19. libpcre.pc was being incorrectly installed as executable.
3094 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3095 it now also looks for '\r' so that it will work unmodified on Windows.
3097 21. Added Google's contributed C++ wrapper to the distribution.
3099 22. Added some untidy missing memory free() calls in pcretest, to keep
3100 Electric Fence happy when testing.
3104 Version 5.0 13-Sep-04
3105 ---------------------
3107 1. Internal change: literal characters are no longer packed up into items
3108 containing multiple characters in a single byte-string. Each character
3109 is now matched using a separate opcode. However, there may be more than one
3110 byte in the character in UTF-8 mode.
3112 2. The pcre_callout_block structure has two new fields: pattern_position and
3113 next_item_length. These contain the offset in the pattern to the next match
3114 item, and its length, respectively.
3116 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3117 insertion of callouts before each pattern item. Added the /C option to
3118 pcretest to make use of this.
3120 4. On the advice of a Windows user, the lines
3122 #if defined(_WIN32) || defined(WIN32)
3123 _setmode( _fileno( stdout ), 0x8000 );
3124 #endif /* defined(_WIN32) || defined(WIN32) */
3126 have been added to the source of pcretest. This apparently does useful
3127 magic in relation to line terminators.
3129 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3130 for the benefit of those environments where the "b" makes a difference.
3132 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3133 to know about it. I have put a hack into configure.in that adds in code
3134 to set GCC=yes if CC=icc. This seems to end up at a point in the
3135 generated configure script that is early enough to affect the setting of
3136 compiler options, which is what is needed, but I have no means of testing
3137 whether it really works. (The user who reported this had patched the
3138 generated configure script, which of course I cannot do.)
3140 LATER: After change 22 below (new libtool files), the configure script
3141 seems to know about icc (and also ecc). Therefore, I have commented out
3142 this hack in configure.in.
3144 7. Added support for pkg-config (2 patches were sent in).
3146 8. Negated POSIX character classes that used a combination of internal tables
3147 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3148 [[:^ascii]]. Typically, they would match almost any characters. The other
3149 POSIX classes were not broken in this way.
3151 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3152 to find the match, as PCRE was deluded into thinking that the match had to
3153 start at the start point or following a newline. The same bug applied to
3154 patterns with negative forward assertions or any backward assertions
3155 preceding ".*" at the start, unless the pattern required a fixed first
3156 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3158 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3159 starting at the last subject character, bytes beyond the end of the subject
3160 string were read.
3162 11. Renamed the variable "class" as "classbits" to make life easier for C++
3163 users. (Previously there was a macro definition, but it apparently wasn't
3164 enough.)
3166 12. Added the new field "tables" to the extra data so that tables can be passed
3167 in at exec time, or the internal tables can be re-selected. This allows
3168 a compiled regex to be saved and re-used at a later time by a different
3169 program that might have everything at different addresses.
3171 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3172 -R library as well as a -L library.
3174 14. The debugging options of pcretest (-d on the command line or D on a
3175 pattern) showed incorrect output for anything following an extended class
3176 that contained multibyte characters and which was followed by a quantifier.
3178 15. Added optional support for general category Unicode character properties
3179 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3180 support. It adds about 90K to the size of the library. The meanings of the
3181 inbuilt class escapes such as \d and \s have NOT been changed.
3183 16. Updated pcredemo.c to include calls to free() to release the memory for the
3184 compiled pattern.
3186 17. The generated file chartables.c was being created in the source directory
3187 instead of in the building directory. This caused the build to fail if the
3188 source directory was different from the building directory, and was
3189 read-only.
3191 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3192 file. No doubt somebody will tell me if they don't make sense... Also added
3193 Dan Mooney's comments about building on OpenVMS.
3195 19. Added support for partial matching via the PCRE_PARTIAL option for
3196 pcre_exec() and the \P data escape in pcretest.
3198 20. Extended pcretest with 3 new pattern features:
3200 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3201 write the compiled pattern to the file whose name is "rest-of-line".
3202 This is a straight binary dump of the data, with the saved pointer to
3203 the character tables forced to be NULL. The study data, if any, is
3204 written too. After writing, pcretest reads a new pattern.
3206 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3207 compiled pattern from the given file. There must not be any
3208 occurrences of "<" in the file name (pretty unlikely); if there are,
3209 pcretest will instead treat the initial "<" as a pattern delimiter.
3210 After reading in the pattern, pcretest goes on to read data lines as
3211 usual.
3213 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3214 and 16-bit fields in a compiled pattern, to simulate a pattern that
3215 was compiled on a host of opposite endianness.
3217 21. The pcre-exec() function can now cope with patterns that were compiled on
3218 hosts of opposite endianness, with this restriction:
3220 As for any compiled expression that is saved and used later, the tables
3221 pointer field cannot be preserved; the extra_data field in the arguments
3222 to pcre_exec() should be used to pass in a tables address if a value
3223 other than the default internal tables were used at compile time.
3225 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3226 now diagnosed as an error. Previously, most of the time, a negative number
3227 would have been treated as zero, but if in addition "ovector" was passed as
3228 NULL, a crash could occur.
3230 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3231 new versions from the libtool 1.5 distribution (the last one is a copy of
3232 a file called libtool.m4). This seems to have fixed the need to patch
3233 "configure" to support Darwin 1.3 (which I used to do). However, I still
3234 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3235 workstation).
3237 24. Changed the PCRE licence to be the more standard "BSD" licence.
3240 Version 4.5 01-Dec-03
3241 ---------------------
3243 1. There has been some re-arrangement of the code for the match() function so
3244 that it can be compiled in a version that does not call itself recursively.
3245 Instead, it keeps those local variables that need separate instances for
3246 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3247 needs to "recurse". Keeping track of where control must go is done by means
3248 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3249 hide most of the details from the main code, and operates only if
3250 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3251 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3252 operating.
3254 To make it easier for callers to provide specially tailored get/free
3255 functions for this usage, two new functions, pcre_stack_malloc, and
3256 pcre_stack_free, are used. They are always called in strict stacking order,
3257 and the size of block requested is always the same.
3259 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3260 PCRE has been compiled to use the stack or the heap for recursion. The
3261 -C option of pcretest uses this to show which version is compiled.
3263 A new data escape \S, is added to pcretest; it causes the amounts of store
3264 obtained and freed by both kinds of malloc/free at match time to be added
3265 to the output.
3267 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3268 what's available on my current Linux desktop machine.
3270 3. When matching a UTF-8 string, the test for a valid string at the start has
3271 been extended. If start_offset is not zero, PCRE now checks that it points
3272 to a byte that is the start of a UTF-8 character. If not, it returns
3273 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3274 this is necessary because there may be backward assertions in the pattern.
3275 When matching the same subject several times, it may save resources to use
3276 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3278 4. The code for checking the validity of UTF-8 strings has been tightened so
3279 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3280 containing "overlong sequences".
3282 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3283 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3284 should have been "&", but it just so happened that all the cases this let
3285 through by mistake were picked up later in the function.
3287 6. I had used a variable called "isblank" - this is a C99 function, causing
3288 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3290 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3291 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3292 so the version is shown; (c) stop "make test" echoing "./RunTest".
3294 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3296 9. The prototype for memmove() for systems that don't have it was using
3297 size_t, but the inclusion of the header that defines size_t was later. I've
3298 moved the #includes for the C headers earlier to avoid this.
3300 10. Added some adjustments to the code to make it easier to compiler on certain
3301 special systems:
3303 (a) Some "const" qualifiers were missing.
3304 (b) Added the macro EXPORT before all exported functions; by default this
3305 is defined to be empty.
3306 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3307 that it reads its output file name as an argument instead of writing
3308 to the standard output and assuming this can be redirected.
3310 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3311 class containing characters with values greater than 255, PCRE compilation
3312 went into a loop.
3314 12. A recursive reference to a subpattern that was within another subpattern
3315 that had a minimum quantifier of zero caused PCRE to crash. For example,
3316 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3317 recursion. If the recursively-called subpattern itself had a zero repeat,
3318 that was OK.
3320 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3321 buffer into which it was copied (for escape processing) was still set at
3322 1024, so long lines caused crashes.
3324 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3325 "internal error: code overflow...". This applied to any character class
3326 that was followed by a possessive quantifier.
3328 15. Modified the Makefile to add libpcre.la as a prerequisite for
3329 libpcreposix.la because I was told this is needed for a parallel build to
3330 work.
3332 16. If a pattern that contained .* following optional items at the start was
3333 studied, the wrong optimizing data was generated, leading to matching
3334 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3335 matching string must start with a or b or c. The correct conclusion for
3336 this pattern is that a match can start with any character.
3339 Version 4.4 13-Aug-03
3340 ---------------------
3342 1. In UTF-8 mode, a character class containing characters with values between
3343 127 and 255 was not handled correctly if the compiled pattern was studied.
3344 In fixing this, I have also improved the studying algorithm for such
3345 classes (slightly).
3347 2. Three internal functions had redundant arguments passed to them. Removal
3348 might give a very teeny performance improvement.
3350 3. Documentation bug: the value of the capture_top field in a callout is *one
3351 more than* the number of the hightest numbered captured substring.
3353 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3354 in incorrectly linking with a previously installed version. They now link
3355 explicitly with libpcre.la.
3357 5. configure.in no longer needs to recognize Cygwin specially.
3359 6. A problem in pcre.in for Windows platforms is fixed.
3361 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3362 pcretest, it used to include the size of the study block as part of its
3363 output. Unfortunately, the structure contains a field that has a different
3364 size on different hardware architectures. This meant that the tests that
3365 showed this size failed. As the block is currently always of a fixed size,
3366 this information isn't actually particularly useful in pcretest output, so
3367 I have just removed it.
3369 8. Three pre-processor statements accidentally did not start in column 1.
3370 Sadly, there are *still* compilers around that complain, even though
3371 standard C has not required this for well over a decade. Sigh.
3373 9. In pcretest, the code for checking callouts passed small integers in the
3374 callout_data field, which is a void * field. However, some picky compilers
3375 complained about the casts involved for this on 64-bit systems. Now
3376 pcretest passes the address of the small integer instead, which should get
3377 rid of the warnings.
3379 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3380 both compile and run time, and gives an error if an invalid UTF-8 sequence
3381 is found. There is a option for disabling this check in cases where the
3382 string is known to be correct and/or the maximum performance is wanted.
3384 11. In response to a bug report, I changed one line in Makefile.in from
3386 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3387 to
3388 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3390 to look similar to other lines, but I have no way of telling whether this
3391 is the right thing to do, as I do not use Windows. No doubt I'll get told
3392 if it's wrong...
3395 Version 4.3 21-May-03
3396 ---------------------
3398 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3399 Makefile.
3401 2. Some refactoring to improve the quality of the code:
3403 (i) The utf8_table... variables are now declared "const".
3405 (ii) The code for \cx, which used the "case flipping" table to upper case
3406 lower case letters, now just substracts 32. This is ASCII-specific,
3407 but the whole concept of \cx is ASCII-specific, so it seems
3408 reasonable.
3410 (iii) PCRE was using its character types table to recognize decimal and
3411 hexadecimal digits in the pattern. This is silly, because it handles
3412 only 0-9, a-f, and A-F, but the character types table is locale-
3413 specific, which means strange things might happen. A private
3414 table is now used for this - though it costs 256 bytes, a table is
3415 much faster than multiple explicit tests. Of course, the standard
3416 character types table is still used for matching digits in subject
3417 strings against \d.
3419 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3420 ending in _t are). So I've renamed it as ESC_tee.
3422 3. The first argument for regexec() in the POSIX wrapper should have been
3423 defined as "const".
3425 4. Changed pcretest to use malloc() for its buffers so that they can be
3426 Electric Fenced for debugging.
3428 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3429 to read one or more bytes before the start of the subject string. Often this
3430 had no effect on PCRE's behaviour, but in some circumstances it could
3431 provoke a segmentation fault.
3433 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3434 to try to read one or more bytes before the start of the subject string.
3436 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3437 UTF-8 support could misbehave in various ways if the subject string
3438 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3439 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3440 back over UTF-8 characters.)
3443 Version 4.2 14-Apr-03
3444 ---------------------
3446 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3448 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3449 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3450 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3451 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3452 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3454 Note: automatic setting of the BUILD variables is not yet working
3455 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3456 compile-time but not at link-time
3457 [LINK]: use for linking executables only
3458 make different versions for Windows and non-Windows
3459 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3460 libraries
3461 [LINK_FOR_BUILD]: new variable
3462 [OBJEXT]: use throughout
3463 [EXEEXT]: use throughout
3464 <winshared>: new target
3465 <wininstall>: new target
3466 <dftables.o>: use native compiler
3467 <dftables>: use native linker
3468 <install>: handle Windows platform correctly
3469 <clean>: ditto
3470 <check>: ditto
3471 copy DLL to top builddir before testing
3473 As part of these changes, -no-undefined was removed again. This was reported
3474 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3475 in any case.
3477 3. Some tidies to get rid of compiler warnings:
3479 . In the match_data structure, match_limit was an unsigned long int, whereas
3480 match_call_count was an int. I've made them both unsigned long ints.
3482 . In pcretest the fact that a const uschar * doesn't automatically cast to
3483 a void * provoked a warning.
3485 . Turning on some more compiler warnings threw up some "shadow" variables
3486 and a few more missing casts.
3488 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3489 option, a class that contained a single character with a value between 128
3490 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3492 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3493 option, a class that contained several characters, but with at least one
3494 whose value was between 128 and 255 caused PCRE to crash.
3497 Version 4.1 12-Mar-03
3498 ---------------------
3500 1. Compiling with gcc -pedantic found a couple of places where casts were
3501 needed, and a string in dftables.c that was longer than standard compilers are
3502 required to support.
3504 2. Compiling with Sun's compiler found a few more places where the code could
3505 be tidied up in order to avoid warnings.
3507 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3508 first of these names is deprecated in the latest Autoconf in favour of the name
3509 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3510 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3511 analogy I have changed it to CFLAGS_FOR_BUILD.
3513 4. Added -no-undefined to the linking command in the Makefile, because this is
3514 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3515 linking step for the pcreposix library.
3517 5. PCRE was failing to diagnose the case of two named groups with the same
3518 name.
3520 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3521 literal character that is needed in the subject for a match, and scans along to
3522 ensure that it is present before embarking on the full matching process. This
3523 saves time in cases of nested unlimited repeats that are never going to match.
3524 Problem: the scan can take a lot of time if the subject is very long (e.g.
3525 megabytes), thus penalizing straightforward matches. It is now done only if the
3526 amount of subject to be scanned is less than 1000 bytes.
3528 7. A lesser problem with the same optimization is that it was recording the
3529 first character of an anchored pattern as "needed", thus provoking a search
3530 right along the subject, even when the first match of the pattern was going to
3531 fail. The "needed" character is now not set for anchored patterns, unless it
3532 follows something in the pattern that is of non-fixed length. Thus, it still
3533 fulfils its original purpose of finding quick non-matches in cases of nested
3534 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3537 Version 4.0 17-Feb-03
3538 ---------------------
3540 1. If a comment in an extended regex that started immediately after a meta-item
3541 extended to the end of string, PCRE compiled incorrect data. This could lead to
3542 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3544 2. Moved to autoconf 2.53 and libtool 1.4.2.
3546 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3547 the special perltest8 script is no longer needed - all the tests can be run
3548 from a single perltest script.
3550 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3551 by \s. It has now been removed in PCRE. This means it isn't recognized as
3552 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3553 class [:space:] *does* include VT, thereby creating a mess.
3555 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3556 space and tab.
3558 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3559 its new features into the main test script, reducing the number of scripts.
3561 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3562 were backward compatible, and made the (?i) apply to the whole pattern, as if
3563 /i were given. Now it behaves more logically, and applies the option setting
3564 only to what follows. PCRE has been changed to follow suit. However, if it
3565 finds options settings right at the start of the pattern, it extracts them into
3566 the global options, as before. Thus, they show up in the info data.
3568 8. Added support for the \Q...\E escape sequence. Characters in between are
3569 treated as literals. This is slightly different from Perl in that $ and @ are
3570 also handled as literals inside the quotes. In Perl, they will cause variable
3571 interpolation. Note the following examples:
3573 Pattern PCRE matches Perl matches
3575 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3576 \Qabc\$xyz\E abc\$xyz abc\$xyz
3577 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3579 For compatibility with Perl, \Q...\E sequences are recognized inside character
3580 classes as well as outside them.
3582 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3583 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3584 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3585 signed/unsigned warnings.
3587 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3588 option for pcretest, so I've replaced it by a simple function that does just
3589 that job.
3591 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3592 "pcregrep -".
3594 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3595 Java package. This provides some syntactic sugar for simple cases of what my
3596 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3597 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3598 item, you can use this simplified notation. Note that only makes sense with
3599 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3600 greediness, whatever the setting of the PCRE_UNGREEDY option.
3602 13. A change of greediness default within a pattern was not taking effect at
3603 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3604 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3605 was abstracted outside.
3607 14. PCRE now supports the \G assertion. It is true when the current matching
3608 position is at the start point of the match. This differs from \A when the
3609 starting offset is non-zero. Used with the /g option of pcretest (or similar
3610 code), it works in the same way as it does for Perl's /g option. If all
3611 alternatives of a regex begin with \G, the expression is anchored to the start
3612 match position, and the "anchored" flag is set in the compiled expression.
3614 15. Some bugs concerning the handling of certain option changes within patterns
3615 have been fixed. These applied to options other than (?ims). For example,
3616 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3617 been the other way round. Some of this was related to change 7 above.
3619 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3620 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3621 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3622 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3624 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3625 mode. Unlike ".", it always matches newline, whatever the setting of
3626 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3627 assertions. Perl allows it, but it doesn't (in general) work because it can't
3628 calculate the length of the lookbehind. At least, that's the case for Perl
3629 5.8.0 - I've been told they are going to document that it doesn't work in
3630 future.
3632 18. Added an error diagnosis for escapes that PCRE does not support: these are
3633 \L, \l, \N, \P, \p, \U, \u, and \X.
3635 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3636 reading past the end of the pattern in cases such as /[abcd/.
3638 20. PCRE was getting more memory than necessary for patterns with classes that
3639 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3641 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3642 compiling PCRE for use with Virtual Pascal.
3644 22. Small fix to the Makefile to make it work properly if the build is done
3645 outside the source tree.
3647 23. Added a new extension: a condition to go with recursion. If a conditional
3648 subpattern starts with (?(R) the "true" branch is used if recursion has
3649 happened, whereas the "false" branch is used only at the top level.
3651 24. When there was a very long string of literal characters (over 255 bytes
3652 without UTF support, over 250 bytes with UTF support), the computation of how
3653 much memory was required could be incorrect, leading to segfaults or other
3654 strange effects.
3656 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3657 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3658 there was a subsequent back reference to those brackets. This meant that, for
3659 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3660 possible to check for precisely this case. All we can do is abandon the
3661 optimization if .* occurs inside capturing brackets when there are any back
3662 references whatsoever. (See below for a better fix that came later.)
3664 26. The handling of the optimization for finding the first character of a
3665 non-anchored pattern, and for finding a character that is required later in the
3666 match were failing in some cases. This didn't break the matching; it just
3667 failed to optimize when it could. The way this is done has been re-implemented.
3669 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3671 28. Added a new feature that provides some of the functionality that Perl
3672 provides with (?{...}). The facility is termed a "callout". The way it is done
3673 in PCRE is for the caller to provide an optional function, by setting
3674 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3675 global variable. By default it is unset, which disables all calling out. To get
3676 the function called, the regex must include (?C) at appropriate points. This
3677 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3678 This provides a means of identifying different callout points. When PCRE
3679 reaches such a point in the regex, if pcre_callout has been set, the external
3680 function is called. It is provided with data in a structure called
3681 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3682 matching continues; if it returns a non-zero value, the match at the current
3683 point fails. However, backtracking will occur if possible. [This was changed
3684 later and other features added - see item 49 below.]
3686 29. pcretest is upgraded to test the callout functionality. It provides a
3687 callout function that displays information. By default, it shows the start of
3688 the match and the current position in the text. There are some new data escapes
3689 to vary what happens:
3691 \C+ in addition, show current contents of captured substrings
3692 \C- do not supply a callout function
3693 \C!n return 1 when callout number n is reached
3694 \C!n!m return 1 when callout number n is reached for the mth time
3696 30. If pcregrep was called with the -l option and just a single file name, it
3697 output "<stdin>" if a match was found, instead of the file name.
3699 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3700 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3701 pcre_exec(). This saves a malloc/free per call. The default value of
3702 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3703 when configuring.
3705 32. The default maximum size of a compiled pattern is 64K. There have been a
3706 few cases of people hitting this limit. The code now uses macros to handle the
3707 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3708 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3709 configuring. Tests 2 and 5 work only with 2-byte links because they output
3710 debugging information about compiled patterns.
3712 33. Internal code re-arrangements:
3714 (a) Moved the debugging function for printing out a compiled regex into
3715 its own source file (printint.c) and used #include to pull it into
3716 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3717 separate copies.
3719 (b) Defined the list of op-code names for debugging as a macro in
3720 internal.h so that it is next to the definition of the opcodes.
3722 (c) Defined a table of op-code lengths for simpler skipping along compiled
3723 code. This is again a macro in internal.h so that it is next to the
3724 definition of the opcodes.
3726 34. Added support for recursive calls to individual subpatterns, along the
3727 lines of Robin Houston's patch (but implemented somewhat differently).
3729 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3730 allow it to read and process whole directories in Win32. This code was
3731 contributed by Lionel Fourquaux; it has not been tested by me.
3733 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3734 used to name a group. Names consist of alphanumerics and underscores, and must
3735 be unique. Back references use the syntax (?P=name) and recursive calls use
3736 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3737 numbers. The function pcre_fullinfo() can be used after compilation to extract
3738 a name/number map. There are three relevant calls:
3740 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3741 PCRE_INFO_NAMECOUNT yields the number of entries
3742 PCRE_INFO_NAMETABLE yields a pointer to the map.
3744 The map is a vector of fixed-size entries. The size of each entry depends on
3745 the length of the longest name used. The first two bytes of each entry are the
3746 group number, most significant byte first. There follows the corresponding
3747 name, zero terminated. The names are in alphabetical order.
3749 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3750 case instead of 255. Making it the same both with and without UTF-8 support
3751 means that the same test output works with both.
3753 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3754 calling malloc() with a zero argument.
3756 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3757 optimization. I've improved things by keeping a bitmap of backreferences with
3758 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3759 fact referenced, the optimization can be applied. It is unlikely that a
3760 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3761 the match to follow \n) will appear inside brackets with a number greater than
3762 31, but if it does, any back reference > 31 suppresses the optimization.
3764 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3765 of disabling numbered capturing parentheses. Any opening parenthesis that is
3766 not followed by ? behaves as if it were followed by ?: but named parentheses
3767 can still be used for capturing (and they will acquire numbers in the usual
3768 way).
3770 41. Redesigned the return codes from the match() function into yes/no/error so
3771 that errors can be passed back from deep inside the nested calls. A malloc
3772 failure while inside a recursive subpattern call now causes the
3773 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3775 42. It is now possible to set a limit on the number of times the match()
3776 function is called in a call to pcre_exec(). This facility makes it possible to
3777 limit the amount of recursion and backtracking, though not in a directly
3778 obvious way, because the match() function is used in a number of different
3779 circumstances. The count starts from zero for each position in the subject
3780 string (for non-anchored patterns). The default limit is, for compatibility, a
3781 large number, namely 10 000 000. You can change this in two ways:
3783 (a) When configuring PCRE before making, you can use --with-match-limit=n
3784 to set a default value for the compiled library.
3786 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3787 a different value is set. See 45 below.
3789 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3791 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3792 of things that can be changed at compile time. The first argument specifies
3793 what is wanted and the second points to where the information is to be placed.
3794 The current list of available information is:
3798 The output is an integer that is set to one if UTF-8 support is available;
3799 otherwise it is set to zero.
3803 The output is an integer that it set to the value of the code that is used for
3804 newline. It is either LF (10) or CR (13).
3808 The output is an integer that contains the number of bytes used for internal
3809 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3813 The output is an integer that contains the threshold above which the POSIX
3814 interface uses malloc() for output vectors. See item 31 above.
3818 The output is an unsigned integer that contains the default limit of the number
3819 of match() calls in a pcre_exec() execution. See 42 above.
3821 44. pcretest has been upgraded by the addition of the -C option. This causes it
3822 to extract all the available output from the new pcre_config() function, and to
3823 output it. The program then exits immediately.
3825 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3826 order to support additional features. One way would have been to define
3827 pcre_exec2() (for example) with extra arguments, but this would not have been
3828 extensible, and would also have required all calls to the original function to
3829 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3830 is used for passing in "extra" data from pcre_study().
3832 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3833 contains the following fields:
3835 flags a bitmap indicating which of the following fields are set
3836 study_data opaque data from pcre_study()
3837 match_limit a way of specifying a limit on match() calls for a specific
3838 call to pcre_exec()
3839 callout_data data for callouts (see 49 below)
3841 The flag bits are also defined in pcre.h, and are
3847 The pcre_study() function now returns one of these new pcre_extra blocks, with
3848 the actual study data pointed to by the study_data field, and the
3849 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3850 before. That is, this change is entirely upwards-compatible and requires no
3851 change to existing code.
3853 If you want to pass in additional data to pcre_exec(), you can either place it
3854 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3855 block.
3857 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3858 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3859 times with different match limits, until it finds the minimum value needed for
3860 pcre_exec() to complete. The value is then output. This can be instructive; for
3861 most simple matches the number is quite small, but for pathological cases it
3862 gets very large very quickly.
3864 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3865 returns the size of the data block pointed to by the study_data field in a
3866 pcre_extra block, that is, the value that was passed as the argument to
3867 pcre_malloc() when PCRE was getting memory in which to place the information
3868 created by pcre_study(). The fourth argument should point to a size_t variable.
3869 pcretest has been extended so that this information is shown after a successful
3870 pcre_study() call when information about the compiled regex is being displayed.
3872 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3873 because what follows is always an absolute path. (Later: it turns out that this
3874 is more than cosmetic for MinGW, because it doesn't like empty path
3875 components.)
3877 49. Some changes have been made to the callout feature (see 28 above):
3879 (i) A callout function now has three choices for what it returns:
3881 0 => success, carry on matching
3882 > 0 => failure at this point, but backtrack if possible
3883 < 0 => serious error, return this value from pcre_exec()
3885 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3886 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3887 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3888 use by callout functions. It will never be used by PCRE itself.
3890 (ii) The pcre_extra structure (see 45 above) has a void * field called
3891 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3892 pcre_callout_block structure has a field of the same name. The contents of
3893 the field passed in the pcre_extra structure are passed to the callout
3894 function in the corresponding field in the callout block. This makes it
3895 easier to use the same callout-containing regex from multiple threads. For
3896 testing, the pcretest program has a new data escape
3898 \C*n pass the number n (may be negative) as callout_data
3900 If the callout function in pcretest receives a non-zero value as
3901 callout_data, it returns that value.
3903 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3904 there were some redundant $(CFLAGS) in commands that are now specified as
3905 $(LINK), which already includes $(CFLAGS).
3907 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3908 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3909 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3910 one-byte characters throughout. Note that case-insensitive matching applies
3911 only to characters whose values are less than 256. PCRE doesn't support the
3912 notion of cases for higher-valued characters.
3914 (i) A character class whose characters are all within 0-255 is handled as
3915 a bit map, and the map is inverted for negative classes. Previously, a
3916 character > 255 always failed to match such a class; however it should
3917 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3919 (ii) A negated character class with a single character < 255 is coded as
3920 "not this character" (OP_NOT). This wasn't working properly when the test
3921 character was multibyte, either singly or repeated.
3923 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3924 mode, for example: \x{100}{2,3}.
3926 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3927 singly or repeated) now correctly test multibyte characters. However,
3928 PCRE doesn't recognize any characters with values greater than 255 as
3929 digits, spaces, or word characters. Such characters always match \D, \S,
3930 and \W, and never match \d, \s, or \w.
3932 (v) Classes may now contain characters and character ranges with values
3933 greater than 255. For example: [ab\x{100}-\x{400}].
3935 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3936 PCRE in UTF-8 mode.
3938 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3939 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3940 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3941 value.)
3943 53. The single man page has become too large. I have therefore split it up into
3944 a number of separate man pages. These also give rise to individual HTML pages;
3945 these are now put in a separate directory, and there is an index.html page that
3946 lists them all. Some hyperlinking between the pages has been installed.
3948 54. Added convenience functions for handling named capturing parentheses.
3950 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3951 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3952 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3953 are faulted.
3955 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3956 calling configure. These values are used when compiling the dftables.c program
3957 which is run to generate the source of the default character tables. They
3958 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3959 you will need to set these values.
3961 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3964 Version 3.9 02-Jan-02
3965 ---------------------
3967 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3969 2. If --disable-static was given, the building process failed when trying to
3970 build pcretest and pcregrep. (For some reason it was using libtool to compile
3971 them, which is not right, as they aren't part of the library.)
3974 Version 3.8 18-Dec-01
3975 ---------------------
3977 1. The experimental UTF-8 code was completely screwed up. It was packing the
3978 bytes in the wrong order. How dumb can you get?
3981 Version 3.7 29-Oct-01
3982 ---------------------
3984 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3985 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3986 this didn't happen under Solaris 8, where I normally test things.
3988 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3989 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3990 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3991 this fix an hour or so after the initial 3.7 release.)
3994 Version 3.6 23-Oct-01
3995 ---------------------
3997 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3998 offsets passed as NULL with zero offset count.
4000 2. The config.guess and config.sub files had not been updated when I moved to
4001 the latest autoconf.
4004 Version 3.5 15-Aug-01
4005 ---------------------
4007 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4008 had been forgotten.
4010 2. By using declared but undefined structures, we can avoid using "void"
4011 definitions in pcre.h while keeping the internal definitions of the structures
4012 private.
4014 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4015 user point of view, this means that both static and shared libraries are built
4016 by default, but this can be individually controlled. More of the work of
4017 handling this static/shared cases is now inside libtool instead of PCRE's make
4018 file.
4020 4. The pcretest utility is now installed along with pcregrep because it is
4021 useful for users (to test regexs) and by doing this, it automatically gets
4022 relinked by libtool. The documentation has been turned into a man page, so
4023 there are now .1, .txt, and .html versions in /doc.
4025 5. Upgrades to pcregrep:
4026 (i) Added long-form option names like gnu grep.
4027 (ii) Added --help to list all options with an explanatory phrase.
4028 (iii) Added -r, --recursive to recurse into sub-directories.
4029 (iv) Added -f, --file to read patterns from a file.
4031 6. pcre_exec() was referring to its "code" argument before testing that
4032 argument for NULL (and giving an error if it was NULL).
4034 7. Upgraded Makefile.in to allow for compiling in a different directory from
4035 the source directory.
4037 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4038 options bits, the pointer it was passed was to an int instead of to an unsigned
4039 long int. This mattered only on 64-bit systems.
4041 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4042 generated) instead of pcre.in, which it its source. Also made the same change
4043 in several of the .c files.
4045 10. A new release of gcc defines printf() as a macro, which broke pcretest
4046 because it had an ifdef in the middle of a string argument for printf(). Fixed
4047 by using separate calls to printf().
4049 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4050 script, to force use of CR or LF instead of \n in the source. On non-Unix
4051 systems, the value can be set in config.h.
4053 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4054 absolute limit. Changed the text of the error message to make this clear, and
4055 likewise updated the man page.
4057 13. The limit of 99 on the number of capturing subpatterns has been removed.
4058 The new limit is 65535, which I hope will not be a "real" limit.
4061 Version 3.4 22-Aug-00
4062 ---------------------
4064 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4066 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4069 Version 3.3 01-Aug-00
4070 ---------------------
4072 1. If an octal character was given, but the value was greater than \377, it
4073 was not getting masked to the least significant bits, as documented. This could
4074 lead to crashes in some systems.
4076 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4077 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4079 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4080 These just pass their arguments on to (pcre_free)(), but they are provided
4081 because some uses of PCRE bind it to non-C systems that can call its functions,
4082 but cannot call free() or pcre_free() directly.
4084 4. Add "make test" as a synonym for "make check". Corrected some comments in
4085 the Makefile.
4087 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4088 Makefile.
4090 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4091 command called pgrep for grepping around the active processes.
4093 7. Added the beginnings of support for UTF-8 character strings.
4095 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4096 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4097 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4098 out for the ar command.)
4101 Version 3.2 12-May-00
4102 ---------------------
4104 This is purely a bug fixing release.
4106 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4107 of ZA. This was just one example of several cases that could provoke this bug,
4108 which was introduced by change 9 of version 2.00. The code for breaking
4109 infinite loops after an iteration that matches an empty string was't working
4110 correctly.
4112 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4113 when matched against abbab (for example). After matching an empty string, it
4114 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4115 caused it to match further down the string than it should.
4117 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4118 was there because it doesn't seem to be needed, and it causes trouble on some
4119 systems, as it is not a Standard C header. It has been removed.
4121 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4122 were reported on the Macintosh. The changes were from
4124 while ((c = *(++ptr)) != 0 && c != '\n');
4125 to
4126 while ((c = *(++ptr)) != 0 && c != '\n') ;
4128 Totally extraordinary, but if that's what it takes...
4130 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4131 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4132 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4133 assumes the way PCRE uses memmove() (always moving upwards).
4135 6. PCRE is being used in one environment where strchr() is not available. There
4136 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4137 faster code anyway.
4140 Version 3.1 09-Feb-00
4141 ---------------------
4143 The only change in this release is the fixing of some bugs in Makefile.in for
4144 the "install" target:
4146 (1) It was failing to install pcreposix.h.
4148 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4151 Version 3.0 01-Feb-00
4152 ---------------------
4154 1. Add support for the /+ modifier to perltest (to output $` like it does in
4155 pcretest).
4157 2. Add support for the /g modifier to perltest.
4159 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4160 matches null strings.
4162 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4163 pattern. Perl treats empty patterns specially - it reuses the most recent
4164 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4165 effect.
4167 5. The POSIX interface was broken in that it was just handing over the POSIX
4168 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4169 required a bigger vector, with some working space on the end. This means that
4170 the POSIX wrapper now has to get and free some memory, and copy the results.
4172 6. Added some simple autoconf support, placing the test data and the
4173 documentation in separate directories, re-organizing some of the
4174 information files, and making it build pcre-config (a GNU standard). Also added
4175 libtool support for building PCRE as a shared library, which is now the
4176 default.
4178 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4179 09 are not valid octal constants. Single digits will be used for minor values
4180 less than 10.
4182 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4183 existing programs that set these in the POSIX interface can use PCRE without
4184 modification.
4186 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4187 return all that pcre_info() returns, plus additional data. The pcre_info()
4188 function is retained for compatibility, but is considered to be obsolete.
4190 10. Added experimental recursion feature (?R) to handle one common case that
4191 Perl 5.6 will be able to do with (?p{...}).
4193 11. Added support for POSIX character classes like [:alpha:], which Perl is
4194 adopting.
4197 Version 2.08 31-Aug-99
4198 ----------------------
4200 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4201 trying to match at the startoffset position, but instead was moving forward to
4202 the next newline as if a previous match had failed.
4204 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4205 and could get into a loop if a null string was matched other than at the start
4206 of the subject.
4208 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4209 be distinguished at compile time, and for completeness also added PCRE_DATE.
4211 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4212 in GnuWin32 environments.
4215 Version 2.07 29-Jul-99
4216 ----------------------
4218 1. The documentation is now supplied in plain text form and HTML as well as in
4219 the form of man page sources.
4221 2. C++ compilers don't like assigning (void *) values to other pointer types.
4222 In particular this affects malloc(). Although there is no problem in Standard
4223 C, I've put in casts to keep C++ compilers happy.
4225 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4226 should be (const char *).
4228 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4229 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4230 However, I haven't made this a standard facility. The documentation doesn't
4231 mention it, and the Makefile doesn't support it.
4233 5. The Makefile now contains an "install" target, with editable destinations at
4234 the top of the file. The pcretest program is not installed.
4236 6. pgrep -V now gives the PCRE version number and date.
4238 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4239 causing the entire string to be ignored, instead of just the last character.
4241 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4242 non-matching string, it can take a very, very long time, even for strings of
4243 quite modest length, because of the nested recursion. PCRE now does better in
4244 some of these cases. It does this by remembering the last required literal
4245 character in the pattern, and pre-searching the subject to ensure it is present
4246 before running the real match. In other words, it applies a heuristic to detect
4247 some types of certain failure quickly, and in the above example, if presented
4248 with a string that has no trailing " it gives "no match" very quickly.
4250 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4251 other alternatives are tried instead.
4254 Version 2.06 09-Jun-99
4255 ----------------------
4257 1. Change pcretest's output for amount of store used to show just the code
4258 space, because the remainder (the data block) varies in size between 32-bit and
4259 64-bit systems.
4261 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4262 start matching at. This allows lookbehinds to work when searching for multiple
4263 occurrences in a string.
4265 3. Added additional options to pcretest for testing multiple occurrences:
4267 /+ outputs the rest of the string that follows a match
4268 /g loops for multiple occurrences, using the new startoffset argument
4269 /G loops for multiple occurrences by passing an incremented pointer
4271 4. PCRE wasn't doing the "first character" optimization for patterns starting
4272 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4273 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4274 the letter 'x'. On long subject strings, this gives a significant speed-up.
4277 Version 2.05 21-Apr-99
4278 ----------------------
4280 1. Changed the type of magic_number from int to long int so that it works
4281 properly on 16-bit systems.
4283 2. Fixed a bug which caused patterns starting with .* not to work correctly
4284 when the subject string contained newline characters. PCRE was assuming
4285 anchoring for such patterns in all cases, which is not correct because .* will
4286 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4287 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4288 must be retried after every newline in the subject.
4291 Version 2.04 18-Feb-99
4292 ----------------------
4294 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4295 computation of the store needed to hold the pattern was incorrect (too large).
4296 If such patterns were nested a few deep, this could multiply and become a real
4297 problem.
4299 2. Added /M option to pcretest to show the memory requirement of a specific
4300 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4302 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4303 compiled in such a way that the backtracking after subsequent failure was
4304 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4305 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4308 Version 2.03 02-Feb-99
4309 ----------------------
4311 1. Fixed typo and small mistake in man page.
4313 2. Added 4th condition (GPL supersedes if conflict) and created separate
4314 LICENCE file containing the conditions.
4316 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4317 Perl, that is the internal \ allows the delimiter to be included in the
4318 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4319 the final delimiter, add \ to the end of the pattern (to test the error).
4321 4. Added the convenience functions for extracting substrings after a successful
4322 match. Updated pcretest to make it able to test these functions.
4325 Version 2.02 14-Jan-99
4326 ----------------------
4328 1. Initialized the working variables associated with each extraction so that
4329 their saving and restoring doesn't refer to uninitialized store.
4331 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4332 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4333 fix the problem.
4335 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4336 calls, and wasn't printing the correct value for compiling calls. Increased the
4337 default value of LOOPREPEAT, and the number of significant figures in the
4338 times.
4340 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4342 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4343 a building problem on Windows NT with a FAT file system.
4346 Version 2.01 21-Oct-98
4347 ----------------------
4349 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4350 to character tables built by pcre_maketables() in the current locale. If NULL
4351 is passed, the default tables are used.
4354 Version 2.00 24-Sep-98
4355 ----------------------
4357 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4358 it any more.
4360 2. Allow quantification of (?>) groups, and make it work correctly.
4362 3. The first character computation wasn't working for (?>) groups.
4364 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4365 end of the subject) and add 5.005's \z, which really does match only at the
4366 very end of the subject.
4368 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4370 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4371 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4372 localized options. All options to pcre_study() were also removed.
4374 7. Add other new features from 5.005:
4376 $(?<= positive lookbehind
4377 $(?<! negative lookbehind
4378 (?imsx-imsx) added the unsetting capability
4379 such a setting is global if at outer level; local otherwise
4380 (?imsx-imsx:) non-capturing groups with option setting
4381 (?(cond)re|re) conditional pattern matching
4383 A backreference to itself in a repeated group matches the previous
4384 captured string.
4386 8. General tidying up of studying (both automatic and via "study")
4387 consequential on the addition of new assertions.
4389 9. As in 5.005, unlimited repeated groups that could match an empty substring
4390 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4391 runtime if any iteration does actually match an empty substring.
4393 10. Include the RunTest script in the distribution.
4395 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4396 discrepancies, some of which were old and were also with respect to 5.004. They
4397 have now been fixed.
4400 Version 1.09 28-Apr-98
4401 ----------------------
4403 1. A negated single character class followed by a quantifier with a minimum
4404 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4405 program crashes, or just wrong answers. This did not apply to negated classes
4406 containing more than one character, or to minima other than one.
4409 Version 1.08 27-Mar-98
4410 ----------------------
4412 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4414 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4415 latter must appear before anything that relies on it in the pattern.
4418 Version 1.07 16-Feb-98
4419 ----------------------
4421 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4422 repeat of a potentially empty string).
4425 Version 1.06 23-Jan-98
4426 ----------------------
4428 1. Added Markus Oberhumer's little patches for C++.
4430 2. Literal strings longer than 255 characters were broken.
4433 Version 1.05 23-Dec-97
4434 ----------------------
4436 1. Negated character classes containing more than one character were failing if
4437 PCRE_CASELESS was set at run time.
4440 Version 1.04 19-Dec-97
4441 ----------------------
4443 1. Corrected the man page, where some "const" qualifiers had been omitted.
4445 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4446 input syntax.
4448 3. Fixed memory leak which occurred when a regex with back references was
4449 matched with an offsets vector that wasn't big enough. The temporary memory
4450 that is used in this case wasn't being freed if the match failed.
4452 4. Tidied pcretest to ensure it frees memory that it gets.
4454 5. Temporary memory was being obtained in the case where the passed offsets
4455 vector was exactly big enough.
4457 6. Corrected definition of offsetof() from change 5 below.
4459 7. I had screwed up change 6 below and broken the rules for the use of
4460 setjmp(). Now fixed.
4463 Version 1.03 18-Dec-97
4464 ----------------------
4466 1. A erroneous regex with a missing opening parenthesis was correctly
4467 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4468 on some systems.
4470 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4471 it was reported that one broken compiler failed on the former because "code" is
4472 also an independent variable.
4474 3. The erroneous regex a[]b caused an array overrun reference.
4476 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4477 fail on data ending with that character. (It was going on too far, and checking
4478 the next character, typically a binary zero.) This was specific to the
4479 optimized code for single-character negative classes.
4481 5. Added a contributed patch from the TIN world which does the following:
4483 + Add an undef for memmove, in case the the system defines a macro for it.
4485 + Add a definition of offsetof(), in case there isn't one. (I don't know
4486 the reason behind this - offsetof() is part of the ANSI standard - but
4487 it does no harm).
4489 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4490 most of the places where whitespace preceded '#'. I have given up and
4491 allowed the remaining 2 cases to be at the margin.
4493 + Rename some variables in pcre to eliminate shadowing. This seems very
4494 pedantic, but does no harm, of course.
4496 6. Moved the call to setjmp() into its own function, to get rid of warnings
4497 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4499 7. Constructs such as \d{8,} were compiling into the equivalent of
4500 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4501 outcome, but in this particular case used more store than had been allocated,
4502 which caused the bug to be discovered because it threw up an internal error.
4504 8. The debugging code in both pcre and pcretest for outputting the compiled
4505 form of a regex was going wrong in the case of back references followed by
4506 curly-bracketed repeats.
4509 Version 1.02 12-Dec-97
4510 ----------------------
4512 1. Typos in pcre.3 and comments in the source fixed.
4514 2. Applied a contributed patch to get rid of places where it used to remove
4515 'const' from variables, and fixed some signed/unsigned and uninitialized
4516 variable warnings.
4518 3. Added the "runtest" target to Makefile.
4520 4. Set default compiler flag to -O2 rather than just -O.
4523 Version 1.01 19-Nov-97
4524 ----------------------
4526 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4527 like /([ab]*)*/, that is, for classes with more than one character in them.
4529 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4530 as /((?>a*))*/ (a PCRE_EXTRA facility).
4533 Version 1.00 18-Nov-97
4534 ----------------------
4536 1. Added compile-time macros to support systems such as SunOS4 which don't have
4537 memmove() or strerror() but have other things that can be used instead.
4539 2. Arranged that "make clean" removes the executables.
4542 Version 0.99 27-Oct-97
4543 ----------------------
4545 1. Fixed bug in code for optimizing classes with only one character. It was
4546 initializing a 32-byte map regardless, which could cause it to run off the end
4547 of the memory it had got.
4549 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4552 Version 0.98 22-Oct-97
4553 ----------------------
4555 1. Fixed bug in code for handling temporary memory usage when there are more
4556 back references than supplied space in the ovector. This could cause segfaults.
4559 Version 0.97 21-Oct-97
4560 ----------------------
4562 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4564 2. Optimized negated single characters not to use a bit map.
4566 3. Brought error texts together as macro definitions; clarified some of them;
4567 fixed one that was wrong - it said "range out of order" when it meant "invalid
4568 escape sequence".
4570 4. Changed some char * arguments to const char *.
4572 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4574 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4575 pcretest.
4578 Version 0.96 16-Oct-97
4579 ----------------------
4581 1. Added a simple "pgrep" utility to the distribution.
4583 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4584 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4585 where "ddd" means "one or more decimal digits".
4587 3. Fixed serious bug. If a pattern had a back reference, but the call to
4588 pcre_exec() didn't supply a large enough ovector to record the related
4589 identifying subpattern, the match always failed. PCRE now remembers the number
4590 of the largest back reference, and gets some temporary memory in which to save
4591 the offsets during matching if necessary, in order to ensure that
4592 backreferences always work.
4594 4. Increased the compatibility with Perl in a number of ways:
4596 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4597 to request this handling. The option can be set at compile or exec time.
4599 (b) $ matches before a terminating newline by default; an option
4600 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4601 mode). The option can be set at compile or exec time.
4603 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4604 the same as Perl's. If the decimal number it represents is less than 10
4605 or there aren't that many previous left capturing parentheses, an octal
4606 escape is read. Inside a character class, it's always an octal escape,
4607 even if it is a single digit.
4609 (d) An escaped but undefined alphabetic character is taken as a literal,
4610 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4611 escapes.
4613 (e) {0} is now permitted. (The previous item is removed from the compiled
4614 pattern).
4616 5. Changed all the names of code files so that the basic parts are no longer
4617 than 10 characters, and abolished the teeny "globals.c" file.
4619 6. Changed the handling of character classes; they are now done with a 32-byte
4620 bit map always.
4622 7. Added the -d and /D options to pcretest to make it possible to look at the
4623 internals of compilation without having to recompile pcre.
4626 Version 0.95 23-Sep-97
4627 ----------------------
4629 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4630 \x20 at the start of a run of normal characters. These were being treated as
4631 real characters, instead of the source characters being re-checked.
4634 Version 0.94 18-Sep-97
4635 ----------------------
4637 1. The functions are now thread-safe, with the caveat that the global variables
4638 containing pointers to malloc() and free() or alternative functions are the
4639 same for all threads.
4641 2. Get pcre_study() to generate a bitmap of initial characters for non-
4642 anchored patterns when this is possible, and use it if passed to pcre_exec().
4645 Version 0.93 15-Sep-97
4646 ----------------------
4648 1. /(b)|(:+)/ was computing an incorrect first character.
4650 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4651 but not actually doing anything yet.
4653 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4654 as Perl does (e.g. [-az] or [az-]).
4656 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4657 all possible positions.
4659 5. Split up into different modules to avoid including unneeded functions in a
4660 compiled binary. However, compile and exec are still in one module. The "study"
4661 function is split off.
4663 6. The character tables are now in a separate module whose source is generated
4664 by an auxiliary program - but can then be edited by hand if required. There are
4665 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4666 toupper() in the code.
4668 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4669 make them global. Abolish the function for setting them, as the caller can now
4670 set them directly.
4673 Version 0.92 11-Sep-97
4674 ----------------------
4676 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4677 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4679 2. Caseless matching was not working in character classes if the characters in
4680 the pattern were in upper case.
4682 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4684 4. Make PCRE_ANCHORED public and accept as a compile option.
4686 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4687 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4688 pass them.
4690 6. Give an error if bad option bits passed at compile or run time.
4692 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4693 pcretest to cause it to pass that flag.
4695 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4696 options, and the first character, if set.
4698 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4701 Version 0.91 10-Sep-97
4702 ----------------------
4704 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4705 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4707 2. PCRE was looping on encountering an indefinitely repeated back reference to
4708 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4709 Perl does - treats the match as successful.
4711 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5