ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 886 - (show annotations)
Tue Jan 17 14:25:07 2012 UTC (3 years, 10 months ago) by ph10
File size: 225406 byte(s)
Error occurred while calculating annotation data.
CMake parses version information out of configure.ac instead of relying on 
1 ChangeLog for PCRE
2 ------------------
4 Version 8.30
5 ------------
7 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
8 name is defined in ctype.h.
10 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
11 only in quite long subpatterns.
13 3. Removed the function pcre_info(), which has been obsolete and deprecated
14 since it was replaced by pcre_fullinfo() in February 2000.
16 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
17 match a (*MARK), and the match failed at the start of the subject, a
18 reference to memory before the start of the subject could occur. This bug
19 was introduced by fix 17 of release 8.21.
21 5. A reference to an unset group with zero minimum repetition was giving
22 totally wrong answers (in non-JavaScript-compatibility mode). For example,
23 /(another)?(\1?)test/ matched against "hello world test". This bug was
24 introduced in release 8.13.
26 6. Add support for 16-bit character strings (a large amount of work involving
27 many changes and refactorings).
29 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
30 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
31 from a file.
33 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
34 rounding is not applied in this particular case).
36 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
37 if they appear, or are escaped, in patterns.
39 10. Get rid of a number of -Wunused-but-set-variable warnings.
41 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
42 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
43 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
44 also returns the mark "x". This bug applied to capturing parentheses,
45 non-capturing parentheses, and atomic parentheses.
47 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
48 information out of configure.ac instead of relying on pcre.h.generic, which
49 is not stored in the repository.
52 Version 8.21 12-Dec-2011
53 ------------------------
55 1. Updating the JIT compiler.
57 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
58 are added as well.
60 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
61 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
62 calling _pcre_jit_exec. Some extra comments are added.
64 4. (*MARK) settings inside atomic groups that do not contain any capturing
65 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
66 was introduced by change 18 for 8.20.
68 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
69 ECMA-262 standard.
71 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
72 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
73 This bug was probably introduced by change 9 of 8.13.
75 7. While fixing 6 above, I noticed that a number of other items were being
76 incorrectly rejected as "not fixed length". This arose partly because newer
77 opcodes had not been added to the fixed-length checking code. I have (a)
78 corrected the bug and added tests for these items, and (b) arranged for an
79 error to occur if an unknown opcode is encountered while checking for fixed
80 length instead of just assuming "not fixed length". The items that were
81 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
82 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
83 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
85 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
86 being incorrectly compiled and would have given unpredicatble results.
88 9. A possessively repeated subpattern with minimum repeat count greater than
89 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
90 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
91 the first (A) could occur when it should not.
93 10. Add a cast and remove a redundant test from the code.
95 11. JIT should use pcre_malloc/pcre_free for allocation.
97 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
98 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
99 is anything other than /usr, -L is still shown).
101 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
103 14. Perl does not support \N without a following name in a [] class; PCRE now
104 also gives an error.
106 15. If a forward reference was repeated with an upper limit of around 2000,
107 it caused the error "internal error: overran compiling workspace". The
108 maximum number of forward references (including repeats) was limited by the
109 internal workspace, and dependent on the LINK_SIZE. The code has been
110 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
111 the default depends on LINK_SIZE. There is a new upper limit (for safety)
112 of around 200,000 forward references. While doing this, I also speeded up
113 the filling in of repeated forward references.
115 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
116 incorrectly expecting the subject to contain another "a" after the start.
118 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
119 in the match, the SKIP should be ignored. This was not happening; instead
120 the SKIP was being treated as NOMATCH. For patterns such as
121 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
122 tested.
124 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
125 now much more compatible with Perl, in particular in cases where the result
126 is a non-match for a non-anchored pattern. For example, if
127 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
128 "m", where previously it did not return a name. A side effect of this
129 change is that for partial matches, the last encountered mark name is
130 returned, as for non matches. A number of tests that were previously not
131 Perl-compatible have been moved into the Perl-compatible test files. The
132 refactoring has had the pleasing side effect of removing one argument from
133 the match() function, thus reducing its stack requirements.
135 19. If the /S+ option was used in pcretest to study a pattern using JIT,
136 subsequent uses of /S (without +) incorrectly behaved like /S+.
138 21. Retrieve executable code size support for the JIT compiler and fixing
139 some warnings.
141 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
142 not work when the shorter character appeared right at the end of the
143 subject string.
145 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
146 systems.
148 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
149 output it when the /M option is used in pcretest.
151 25. The CheckMan script was not being included in the distribution. Also, added
152 an explicit "perl" to run Perl scripts from the PrepareRelease script
153 because this is reportedly needed in Windows.
155 26. If study data was being save in a file and studying had not found a set of
156 "starts with" bytes for the pattern, the data written to the file (though
157 never used) was taken from uninitialized memory and so caused valgrind to
158 complain.
160 27. Updated RunTest.bat as provided by Sheri Pierce.
162 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
164 29. Computation of memory usage for the table of capturing group names was
165 giving an unnecessarily large value.
168 Version 8.20 21-Oct-2011
169 ------------------------
171 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
172 a POSIX class. After further experiments with Perl, which convinced me that
173 Perl has bugs and confusions, a closing square bracket is no longer allowed
174 in a POSIX name. This bug also affected patterns with classes that started
175 with full stops.
177 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
178 captured substring, but while checking the failing first alternative,
179 substring 1 is temporarily captured. If the output vector supplied to
180 pcre_exec() was not big enough for this capture, the yield of the function
181 was still zero ("insufficient space for captured substrings"). This cannot
182 be totally fixed without adding another stack variable, which seems a lot
183 of expense for a edge case. However, I have improved the situation in cases
184 such as /(a)(b)x|abc/ matched against "abc", where the return code
185 indicates that fewer than the maximum number of slots in the ovector have
186 been set.
188 3. Related to (2) above: when there are more back references in a pattern than
189 slots in the output vector, pcre_exec() uses temporary memory during
190 matching, and copies in the captures as far as possible afterwards. It was
191 using the entire output vector, but this conflicts with the specification
192 that only 2/3 is used for passing back captured substrings. Now it uses
193 only the first 2/3, for compatibility. This is, of course, another edge
194 case.
196 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
197 main code base, and can be used by building with --enable-jit. When this is
198 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
199 runtime --no-jit option is given.
201 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
202 ovector, the return from the function was zero, implying that there were
203 other matches that did not fit. The correct "exactly full" value is now
204 returned.
206 6. If a subpattern that was called recursively or as a subroutine contained
207 (*PRUNE) or any other control that caused it to give a non-standard return,
208 invalid errors such as "Error -26 (nested recursion at the same subject
209 position)" or even infinite loops could occur.
211 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
212 computing the minimum length on reaching *ACCEPT, and so ended up with the
213 wrong value of 1 rather than 0. Further investigation indicates that
214 computing a minimum subject length in the presence of *ACCEPT is difficult
215 (think back references, subroutine calls), and so I have changed the code
216 so that no minimum is registered for a pattern that contains *ACCEPT.
218 8. If (*THEN) was present in the first (true) branch of a conditional group,
219 it was not handled as intended. [But see 16 below.]
221 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
222 Sheri Pierce.
224 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
225 the first byte in a match must be "a".
227 11. Change 17 for 8.13 increased the recursion depth for patterns like
228 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
229 pattern contains any instances of (*THEN). If it does not, the old
230 optimizations are restored. It would be nice to do this on a per-group
231 basis, but at the moment that is not feasible.
233 12. In some environments, the output of pcretest -C is CRLF terminated. This
234 broke RunTest's code that checks for the link size. A single white space
235 character after the value is now allowed for.
237 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
238 For "fr", it uses the Windows-specific input and output files.
240 14. If (*THEN) appeared in a group that was called recursively or as a
241 subroutine, it did not work as intended. [But see next item.]
243 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
244 pattern fragments (but not containing any | characters). If A and B are
245 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
246 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
247 D. In other words, Perl considers parentheses that do not contain any |
248 characters to be part of a surrounding alternative, whereas PCRE was
249 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
250 differently. PCRE now behaves in the same way as Perl, except in the case
251 of subroutine/recursion calls such as (?1) which have in any case always
252 been different (but PCRE had them first :-).
254 16. Related to 15 above: Perl does not treat the | in a conditional group as
255 creating alternatives. Such a group is treated in the same way as an
256 ordinary group without any | characters when processing (*THEN). PCRE has
257 been changed to match Perl's behaviour.
259 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
260 RunGrepTest script failed.
262 18. Change 22 for version 13 caused atomic groups to use more stack. This is
263 inevitable for groups that contain captures, but it can lead to a lot of
264 stack use in large patterns. The old behaviour has been restored for atomic
265 groups that do not contain any capturing parentheses.
267 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
268 suppress the check for a minimum subject length at run time. (If it was
269 given to pcre_exec() or pcre_dfa_exec() it did work.)
271 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
272 fail to work when decoding hex characters in data strings in EBCDIC
273 environments.
275 21. It appears that in at least one Mac OS environment, the isxdigit() function
276 is implemented as a macro that evaluates to its argument more than once,
277 contravening the C 90 Standard (I haven't checked a later standard). There
278 was an instance in pcretest which caused it to go wrong when processing
279 \x{...} escapes in subject strings. The has been rewritten to avoid using
280 things like p++ in the argument of isxdigit().
283 Version 8.13 16-Aug-2011
284 ------------------------
286 1. The Unicode data tables have been updated to Unicode 6.0.0.
288 2. Two minor typos in pcre_internal.h have been fixed.
290 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
291 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
292 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
294 4. There were a number of related bugs in the code for matching backrefences
295 caselessly in UTF-8 mode when codes for the characters concerned were
296 different numbers of bytes. For example, U+023A and U+2C65 are an upper
297 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
298 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
299 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
300 2-byte code at the end of the subject (it thought there wasn't enough data
301 left).
303 5. Comprehensive information about what went wrong is now returned by
304 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
305 as the output vector has at least 2 elements. The offset of the start of
306 the failing character and a reason code are placed in the vector.
308 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
309 now returned is for the first byte of the failing character, instead of the
310 last byte inspected. This is an incompatible change, but I hope it is small
311 enough not to be a problem. It makes the returned offset consistent with
312 pcre_exec() and pcre_dfa_exec().
314 7. pcretest now gives a text phrase as well as the error number when
315 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
316 failure, the offset and reason code are output.
318 8. When \R was used with a maximizing quantifier it failed to skip backwards
319 over a \r\n pair if the subsequent match failed. Instead, it just skipped
320 back over a single character (\n). This seems wrong (because it treated the
321 two characters as a single entity when going forwards), conflicts with the
322 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
323 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
324 has been changed.
326 9. Some internal refactoring has changed the processing so that the handling
327 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
328 time (the PCRE_DOTALL option was changed this way some time ago: version
329 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
330 which was always a bit of a fudge. It also means that there is one less
331 argument for the match() function, which reduces its stack requirements
332 slightly. This change also fixes an incompatibility with Perl: the pattern
333 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
335 10. More internal refactoring has drastically reduced the number of recursive
336 calls to match() for possessively repeated groups such as (abc)++ when
337 using pcre_exec().
339 11. While implementing 10, a number of bugs in the handling of groups were
340 discovered and fixed:
342 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
343 (a|)*(?1) gave a compile-time internal error.
344 ((a|)+)+ did not notice that the outer group could match an empty string.
345 (^a|^)+ was not marked as anchored.
346 (.*a|.*)+ was not marked as matching at start or after a newline.
348 12. Yet more internal refactoring has removed another argument from the match()
349 function. Special calls to this function are now indicated by setting a
350 value in a variable in the "match data" data block.
352 13. Be more explicit in pcre_study() instead of relying on "default" for
353 opcodes that mean there is no starting character; this means that when new
354 ones are added and accidentally left out of pcre_study(), testing should
355 pick them up.
357 14. The -s option of pcretest has been documented for ages as being an old
358 synonym of -m (show memory usage). I have changed it to mean "force study
359 for every regex", that is, assume /S for every regex. This is similar to -i
360 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
361 using it. It makes it easier to run collections of tests with and without
362 study enabled, and thereby test pcre_study() more easily. All the standard
363 tests are now run with and without -s (but some patterns can be marked as
364 "never study" - see 20 below).
366 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
367 restoration of the capturing data to the outer values was not happening
368 correctly.
370 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
371 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
372 pattern had matched an empty string, and so incorrectly returned a no
373 match.
375 17. There was optimizing code for the last branch of non-capturing parentheses,
376 and also for the obeyed branch of a conditional subexpression, which used
377 tail recursion to cut down on stack usage. Unfortunately, now that there is
378 the possibility of (*THEN) occurring in these branches, tail recursion is
379 no longer possible because the return has to be checked for (*THEN). These
380 two optimizations have therefore been removed. [But see 8.20/11 above.]
382 18. If a pattern containing \R was studied, it was assumed that \R always
383 matched two bytes, thus causing the minimum subject length to be
384 incorrectly computed because \R can also match just one byte.
386 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
387 was incorrectly computed.
389 20. If /S is present twice on a test pattern in pcretest input, it now
390 *disables* studying, thereby overriding the use of -s on the command line
391 (see 14 above). This is necessary for one or two tests to keep the output
392 identical in both cases.
394 21. When (*ACCEPT) was used in an assertion that matched an empty string and
395 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
397 22. When an atomic group that contained a capturing parenthesis was
398 successfully matched, but the branch in which it appeared failed, the
399 capturing was not being forgotten if a higher numbered group was later
400 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
401 group 1 to "a", when in fact it should be unset. This applied to multi-
402 branched capturing and non-capturing groups, repeated or not, and also to
403 positive assertions (capturing in negative assertions does not happen
404 in PCRE) and also to nested atomic groups.
406 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
407 subject after a captured substring, to make it easier to tell which of a
408 number of identical substrings has been captured.
410 24. The way atomic groups are processed by pcre_exec() has been changed so that
411 if they are repeated, backtracking one repetition now resets captured
412 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
413 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
414 "aaa". Previously, it would have been "a". As part of this code
415 refactoring, the way recursive calls are handled has also been changed.
417 25. If an assertion condition captured any substrings, they were not passed
418 back unless some other capturing happened later. For example, if
419 (?(?=(a))a) was matched against "a", no capturing was returned.
421 26. When studying a pattern that contained subroutine calls or assertions,
422 the code for finding the minimum length of a possible match was handling
423 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
424 group 1 called group 2 while simultaneously a separate group 2 called group
425 1). A stack overflow occurred in this case. I have fixed this by limiting
426 the recursion depth to 10.
428 27. Updated RunTest.bat in the distribution to the version supplied by Tom
429 Fortmann. This supports explicit test numbers on the command line, and has
430 argument validation and error reporting.
432 28. An instance of \X with an unlimited repeat could fail if at any point the
433 first character it looked at was a mark character.
435 29. Some minor code refactoring concerning Unicode properties and scripts
436 should reduce the stack requirement of match() slightly.
438 30. Added the '=' option to pcretest to check the setting of unused capturing
439 slots at the end of the pattern, which are documented as being -1, but are
440 not included in the return count.
442 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
443 compiled something random. Now it gives a compile-time error (as does
444 Perl).
446 32. A *MARK encountered during the processing of a positive assertion is now
447 recorded and passed back (compatible with Perl).
449 33. If --only-matching or --colour was set on a pcregrep call whose pattern
450 had alternative anchored branches, the search for a second match in a line
451 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
452 matched the line "0102" twice. The same bug affected patterns that started
453 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
454 twice.
456 34. Previously, PCRE did not allow quantification of assertions. However, Perl
457 does, and because of capturing effects, quantifying parenthesized
458 assertions may at times be useful. Quantifiers are now allowed for
459 parenthesized assertions.
461 35. A minor code tidy in pcre_compile() when checking options for \R usage.
463 36. \g was being checked for fancy things in a character class, when it should
464 just be a literal "g".
466 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
467 appearance of a nested POSIX class supersedes an apparent external class.
468 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
469 unescaped square brackets may also appear as part of class names. For
470 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
471 more like Perl. (But see 8.20/1 above.)
473 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
474 was because it thought it was \N{name}, which is not supported).
476 39. Add minix to OS list not supporting the -S option in pcretest.
478 40. PCRE tries to detect cases of infinite recursion at compile time, but it
479 cannot analyze patterns in sufficient detail to catch mutual recursions
480 such as ((?1))((?2)). There is now a runtime test that gives an error if a
481 subgroup is called recursively as a subpattern for a second time at the
482 same position in the subject string. In previous releases this might have
483 been caught by the recursion limit, or it might have run out of stack.
485 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
486 happen only once. PCRE was, however incorrectly giving a compile time error
487 "recursive call could loop indefinitely" because it cannot analyze the
488 pattern in sufficient detail. The compile time test no longer happens when
489 PCRE is compiling a conditional subpattern, but actual runaway loops are
490 now caught at runtime (see 40 above).
492 42. It seems that Perl allows any characters other than a closing parenthesis
493 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
494 has been changed to be the same.
496 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
497 as not to get warnings when autogen.sh is called. Also changed
498 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
500 44. To help people who use pcregrep to scan files containing exceedingly long
501 lines, the following changes have been made:
503 (a) The default value of the buffer size parameter has been increased from
504 8K to 20K. (The actual buffer used is three times this size.)
506 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
507 PCRE is built.
509 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
510 to be set at run time.
512 (d) Numerical values in pcregrep options can be followed by K or M, for
513 example --buffer-size=50K.
515 (e) If a line being scanned overflows pcregrep's buffer, an error is now
516 given and the return code is set to 2.
518 45. Add a pointer to the latest mark to the callout data block.
520 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
521 partial match of an empty string instead of no match. This was specific to
522 the use of ".".
524 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
525 complete match instead of a partial match. This bug was dependent on both
526 the PCRE_UTF8 and PCRE_DOTALL options being set.
528 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
529 starting byte set, because \b was not being ignored.
532 Version 8.12 15-Jan-2011
533 ------------------------
535 1. Fixed some typos in the markup of the man pages, and wrote a script that
536 checks for such things as part of the documentation building process.
538 2. On a big-endian 64-bit system, pcregrep did not correctly process the
539 --match-limit and --recursion-limit options (added for 8.11). In
540 particular, this made one of the standard tests fail. (The integer value
541 went into the wrong half of a long int.)
543 3. If the --colour option was given to pcregrep with -v (invert match), it
544 did strange things, either producing crazy output, or crashing. It should,
545 of course, ignore a request for colour when reporting lines that do not
546 match.
548 4. Another pcregrep bug caused similar problems if --colour was specified with
549 -M (multiline) and the pattern match finished with a line ending.
551 5. In pcregrep, when a pattern that ended with a literal newline sequence was
552 matched in multiline mode, the following line was shown as part of the
553 match. This seems wrong, so I have changed it.
555 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
556 the check for further matches in the same line (so they could be coloured)
557 to overrun the end of the current line. If another match was found, it was
558 incorrectly shown (and then shown again when found in the next line).
560 7. If pcregrep was compiled under Windows, there was a reference to the
561 function pcregrep_exit() before it was defined. I am assuming this was
562 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
563 reported by a user. I've moved the definition above the reference.
566 Version 8.11 10-Dec-2010
567 ------------------------
569 1. (*THEN) was not working properly if there were untried alternatives prior
570 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
571 backtracked to try for "b" instead of moving to the next alternative branch
572 at the same level (in this case, to look for "c"). The Perl documentation
573 is clear that when (*THEN) is backtracked onto, it goes to the "next
574 alternative in the innermost enclosing group".
576 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
577 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
578 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
579 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
580 (*THEN).
582 3. If \s appeared in a character class, it removed the VT character from
583 the class, even if it had been included by some previous item, for example
584 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
585 of \s, but is part of the POSIX "space" class.)
587 4. A partial match never returns an empty string (because you can always
588 match an empty string at the end of the subject); however the checking for
589 an empty string was starting at the "start of match" point. This has been
590 changed to the "earliest inspected character" point, because the returned
591 data for a partial match starts at this character. This means that, for
592 example, /(?<=abc)def/ gives a partial match for the subject "abc"
593 (previously it gave "no match").
595 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
596 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
597 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
598 has an implication that the given string is incomplete (because a partial
599 match is preferred over a full match). For this reason, these items now
600 give a partial match in this situation. [Aside: previously, the one case
601 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
602 match rather than a full match, which was wrong by the old rules, but is
603 now correct.]
605 6. There was a bug in the handling of #-introduced comments, recognized when
606 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
607 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
608 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
609 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
610 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
611 places in pcre_compile().
613 7. Related to (6) above, when pcre_compile() was skipping #-introduced
614 comments when looking ahead for named forward references to subpatterns,
615 the only newline sequence it recognized was NL. It now handles newlines
616 according to the set newline convention.
618 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
619 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
620 cater for a lack of strerror(). These oversights have been fixed.
622 9. Added --match-limit and --recursion-limit to pcregrep.
624 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
626 11. When the -o option was used, pcregrep was setting a return code of 1, even
627 when matches were found, and --line-buffered was not being honoured.
629 12. Added an optional parentheses number to the -o and --only-matching options
630 of pcregrep.
632 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
633 can match an empty string. The code to do it in pcretest and pcredemo
634 needed fixing:
636 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
637 only one byte after an empty string match just before CRLF (this case
638 just got forgotten; "any" and "anycrlf" were OK).
640 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
641 mode when an empty string match preceded an ASCII character followed by
642 a non-ASCII character. (The code for advancing by one character rather
643 than one byte was nonsense.)
645 (c) The pcredemo.c sample program did not have any code at all to handle
646 the cases when CRLF is a valid newline sequence.
648 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
649 as a starting offset was within the subject string. There is now a new
650 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
651 negative or greater than the length of the string. In order to test this,
652 pcretest is extended to allow the setting of negative starting offsets.
654 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
655 starting offset points to the beginning of a UTF-8 character was
656 unnecessarily clumsy. I tidied it up.
658 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
659 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
661 17. Nobody had reported that the --include_dir option, which was added in
662 release 7.7 should have been called --include-dir (hyphen, not underscore)
663 for compatibility with GNU grep. I have changed it to --include-dir, but
664 left --include_dir as an undocumented synonym, and the same for
665 --exclude-dir, though that is not available in GNU grep, at least as of
666 release 2.5.4.
668 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
669 characters from a string of bytes) have been redefined so as not to use
670 loops, in order to improve performance in some environments. At the same
671 time, I abstracted some of the common code into auxiliary macros to save
672 repetition (this should not affect the compiled code).
674 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
675 compile-time error is now given if \c is not followed by an ASCII
676 character, that is, a byte less than 128. (In EBCDIC mode, the code is
677 different, and any byte value is allowed.)
679 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
680 START_OPTIMIZE option, which is now allowed at compile time - but just
681 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
682 to pcregrep and other applications that have no direct access to PCRE
683 options. The new /Y option in pcretest sets this option when calling
684 pcre_compile().
686 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
687 back references. Groups containing recursive back references were forced to
688 be atomic by that change, but in the case of named groups, the amount of
689 memory required was incorrectly computed, leading to "Failed: internal
690 error: code overflow". This has been fixed.
692 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
693 pcretest.c, to avoid build problems in some Borland environments.
696 Version 8.10 25-Jun-2010
697 ------------------------
699 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
700 THEN.
702 2. (*ACCEPT) was not working when inside an atomic group.
704 3. Inside a character class, \B is treated as a literal by default, but
705 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
706 causes the error). The code is unchanged, but I tidied the documentation.
708 4. Inside a character class, PCRE always treated \R and \X as literals,
709 whereas Perl faults them if its -w option is set. I have changed PCRE so
710 that it faults them when PCRE_EXTRA is set.
712 5. Added support for \N, which always matches any character other than
713 newline. (It is the same as "." when PCRE_DOTALL is not set.)
715 6. When compiling pcregrep with newer versions of gcc which may have
716 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
717 declared with attribute warn_unused_result" were given. Just casting the
718 result to (void) does not stop the warnings; a more elaborate fudge is
719 needed. I've used a macro to implement this.
721 7. Minor change to pcretest.c to avoid a compiler warning.
723 8. Added four artifical Unicode properties to help with an option to make
724 \s etc use properties (see next item). The new properties are: Xan
725 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
727 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
728 use Unicode properties. (*UCP) at the start of a pattern can be used to set
729 this option. Modified pcretest to add /W to test this facility. Added
730 REG_UCP to make it available via the POSIX interface.
732 10. Added --line-buffered to pcregrep.
734 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
735 studied, and the match started with a letter with a code point greater than
736 127 whose first byte was different to the first byte of the other case of
737 the letter, the other case of this starting letter was not recognized
738 (#976).
740 12. If a pattern that was studied started with a repeated Unicode property
741 test, for example, \p{Nd}+, there was the theoretical possibility of
742 setting up an incorrect bitmap of starting bytes, but fortunately it could
743 not have actually happened in practice until change 8 above was made (it
744 added property types that matched character-matching opcodes).
746 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
747 possible starting bytes for non-anchored patterns.
749 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
750 \R, and also a number of cases that involve Unicode properties, both
751 explicit and implicit when PCRE_UCP is set.
753 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
754 input, it could crash or give wrong results if characters with values
755 greater than 0xc0 were present in the subject string. (Detail: it assumed
756 UTF-8 input when processing these items.)
758 16. Added a lot of (int) casts to avoid compiler warnings in systems where
759 size_t is 64-bit (#991).
761 17. Added a check for running out of memory when PCRE is compiled with
762 --disable-stack-for-recursion (#990).
764 18. If the last data line in a file for pcretest does not have a newline on
765 the end, a newline was missing in the output.
767 19. The default pcre_chartables.c file recognizes only ASCII characters (values
768 less than 128) in its various bitmaps. However, there is a facility for
769 generating tables according to the current locale when PCRE is compiled. It
770 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
771 characters, are recognized by isspace() and therefore were getting set in
772 these tables, and indeed these tables seem to approximate to ISO 8859. This
773 caused a problem in UTF-8 mode when pcre_study() was used to create a list
774 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
775 which of course cannot start UTF-8 characters. I have changed the code so
776 that only real ASCII characters (less than 128) and the correct starting
777 bytes for UTF-8 encodings are set for characters greater than 127 when in
778 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
779 altogether.)
781 20. Added the /T option to pcretest so as to be able to run tests with non-
782 standard character tables, thus making it possible to include the tests
783 used for 19 above in the standard set of tests.
785 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
786 reference to a subpattern the other side of a comment that contains an
787 opening parenthesis caused either an internal compiling error, or a
788 reference to the wrong subpattern.
791 Version 8.02 19-Mar-2010
792 ------------------------
794 1. The Unicode data tables have been updated to Unicode 5.2.0.
796 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
797 configured.
799 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
800 original author of that file, following a query about its status.
802 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
803 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
805 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
806 quantifier applied to a forward-referencing subroutine call, could compile
807 incorrect code or give the error "internal error: previously-checked
808 referenced subpattern not found".
810 6. Both MS Visual Studio and Symbian OS have problems with initializing
811 variables to point to external functions. For these systems, therefore,
812 pcre_malloc etc. are now initialized to local functions that call the
813 relevant global functions.
815 7. There were two entries missing in the vectors called coptable and poptable
816 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
817 I've fixed the data, and added a kludgy way of testing at compile time that
818 the lengths are correct (equal to the number of opcodes).
820 8. Following on from 7, I added a similar kludge to check the length of the
821 eint vector in pcreposix.c.
823 9. Error texts for pcre_compile() are held as one long string to avoid too
824 much relocation at load time. To find a text, the string is searched,
825 counting zeros. There was no check for running off the end of the string,
826 which could happen if a new error number was added without updating the
827 string.
829 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
831 11. \K was not working if it appeared in an atomic group or in a group that
832 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
833 \K is "not well defined" if used in an assertion. PCRE now accepts it if
834 the assertion is positive, but not if it is negative.
836 12. Change 11 fortuitously reduced the size of the stack frame used in the
837 "match()" function of pcre_exec.c by one pointer. Forthcoming
838 implementation of support for (*MARK) will need an extra pointer on the
839 stack; I have reserved it now, so that the stack frame size does not
840 decrease.
842 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
843 item in branch that calls a recursion is a subroutine call - as in the
844 second branch in the above example - was incorrectly given the compile-
845 time error "recursive call could loop indefinitely" because pcre_compile()
846 was not correctly checking the subroutine for matching a non-empty string.
848 14. The checks for overrunning compiling workspace could trigger after an
849 overrun had occurred. This is a "should never occur" error, but it can be
850 triggered by pathological patterns such as hundreds of nested parentheses.
851 The checks now trigger 100 bytes before the end of the workspace.
853 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
856 Version 8.01 19-Jan-2010
857 ------------------------
859 1. If a pattern contained a conditional subpattern with only one branch (in
860 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
861 computed the wrong minimum data length (which is of course zero for such
862 subpatterns). This could cause incorrect "no match" results.
864 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
865 the pattern is reset in the first branch, pcre_compile() failed with
866 "internal error: code overflow at offset...". This happened only when
867 the reset was to the original external option setting. (An optimization
868 abstracts leading options settings into an external setting, which was the
869 cause of this.)
871 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
872 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
873 assertion pattern did not match (meaning that the assertion was true), it
874 was incorrectly treated as false if the SKIP had been reached during the
875 matching. This also applied to assertions used as conditions.
877 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
878 assertion subpattern, including such a pattern used as a condition,
879 unpredictable results occurred, instead of the error return
882 5. The C++ GlobalReplace function was not working like Perl for the special
883 situation when an empty string is matched. It now does the fancy magic
884 stuff that is necessary.
886 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
887 removed. (These were left over from very, very early versions of PCRE.)
889 7. Some cosmetic changes to the code to make life easier when compiling it
890 as part of something else:
892 (a) Change DEBUG to PCRE_DEBUG.
894 (b) In pcre_compile(), rename the member of the "branch_chain" structure
895 called "current" as "current_branch", to prevent a collision with the
896 Linux macro when compiled as a kernel module.
898 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
899 prevent a collision with the Linux macro when compiled as a kernel
900 module.
902 8. In pcre_compile() there are some checks for integer overflows that used to
903 cast potentially large values to (double). This has been changed to that
904 when building, a check for int64_t is made, and if it is found, it is used
905 instead, thus avoiding the use of floating point arithmetic. (There is no
906 other use of FP in PCRE.) If int64_t is not found, the fallback is to
907 double.
909 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
910 2005 (difference between two addresses compared to an unsigned value).
912 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
913 custom one, because of the following reported problem in Windows:
915 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
916 under Win32.
917 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
918 therefore missing the function definition.
919 - The compiler thus generates a "C" signature for the test function.
920 - The linker fails to find the "C" function.
921 - PCRE fails to configure if asked to do so against libbz2.
923 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
924 messages were output:
926 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
927 rerunning libtoolize, to keep the correct libtool macros in-tree.
928 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
930 I have done both of these things.
932 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
933 most of the time, it *can* run out if it is given a pattern that contains a
934 runaway infinite recursion. I updated the discussion in the pcrestack man
935 page.
937 13. Now that we have gone to the x.xx style of version numbers, the minor
938 version may start with zero. Using 08 or 09 is a bad idea because users
939 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
940 interpreted as invalid octal numbers. I've updated the previous comment in
941 configure.ac, and also added a check that gives an error if 08 or 09 are
942 used.
944 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
945 causing partial matching to fail when the end of the subject matched \W
946 in a UTF-8 pattern where \W was quantified with a minimum of 3.
948 15. There were some discrepancies between the declarations in pcre_internal.h
949 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
950 their definitions. The declarations used "const uschar *" and the
951 definitions used USPTR. Even though USPTR is normally defined as "const
952 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
953 reported that: "This difference in casting confuses some C++ compilers, for
954 example, SunCC recognizes above declarations as different functions and
955 generates broken code for hbpcre." I have changed the declarations to use
956 USPTR.
958 16. GNU libtool is named differently on some systems. The autogen.sh script now
959 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
960 (FreeBSD).
962 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
963 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
964 comment: "Figure out how to create a longlong from a string: strtoll and
965 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
966 instance, but it only takes 2 args instead of 3!"
968 18. A subtle bug concerned with back references has been fixed by a change of
969 specification, with a corresponding code fix. A pattern such as
970 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
971 refers, was giving matches when it shouldn't. For example, xa=xaaa would
972 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
973 same bug. Such groups have to be quantified to be useful, or contained
974 inside another quantified group. (If there's no repetition, the reference
975 can never match.) The problem arises because, having left the group and
976 moved on to the rest of the pattern, a later failure that backtracks into
977 the group uses the captured value from the final iteration of the group
978 rather than the correct earlier one. I have fixed this in PCRE by forcing
979 any group that contains a reference to itself to be an atomic group; that
980 is, there cannot be any backtracking into it once it has completed. This is
981 similar to recursive and subroutine calls.
984 Version 8.00 19-Oct-09
985 ----------------------
987 1. The table for translating pcre_compile() error codes into POSIX error codes
988 was out-of-date, and there was no check on the pcre_compile() error code
989 being within the table. This could lead to an OK return being given in
990 error.
992 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
993 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
994 in a Windows environment.
996 3. The pcregrep --count option prints the count for each file even when it is
997 zero, as does GNU grep. However, pcregrep was also printing all files when
998 --files-with-matches was added. Now, when both options are given, it prints
999 counts only for those files that have at least one match. (GNU grep just
1000 prints the file name in this circumstance, but including the count seems
1001 more useful - otherwise, why use --count?) Also ensured that the
1002 combination -clh just lists non-zero counts, with no names.
1004 4. The long form of the pcregrep -F option was incorrectly implemented as
1005 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1006 but it seems right to fix it, and I didn't think it was worth preserving
1007 the old behaviour.
1009 5. The command line items --regex=pattern and --regexp=pattern were not
1010 recognized by pcregrep, which required --regex pattern or --regexp pattern
1011 (with a space rather than an '='). The man page documented the '=' forms,
1012 which are compatible with GNU grep; these now work.
1014 6. No libpcreposix.pc file was created for pkg-config; there was just
1015 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1017 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1018 when UCP support is not needed, by modifying the Python script that
1019 generates it from Unicode data files. This should not matter if the module
1020 is correctly used as a library, but I received one complaint about 50K of
1021 unwanted data. My guess is that the person linked everything into his
1022 program rather than using a library. Anyway, it does no harm.
1024 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1025 was a minimum greater than 1 for a wide character in a possessive
1026 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1027 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1028 character. Chaos in the form of incorrect output or a compiling loop could
1029 result.
1031 9. The restrictions on what a pattern can contain when partial matching is
1032 requested for pcre_exec() have been removed. All patterns can now be
1033 partially matched by this function. In addition, if there are at least two
1034 slots in the offset vector, the offset of the earliest inspected character
1035 for the match and the offset of the end of the subject are set in them when
1036 PCRE_ERROR_PARTIAL is returned.
1038 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1039 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1040 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1041 and may be more useful for multi-segment matching.
1043 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1044 used to be given if ever the end of the subject was reached; now it is
1045 given only if matching could not proceed because another character was
1046 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1047 string "Z", which now yields "no match" instead of "partial match". In the
1048 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1049 final character ended with (*FAIL).
1051 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1052 if the pattern had a "must contain" character that was already found in the
1053 earlier partial match, unless partial matching was again requested. For
1054 example, with the pattern /dog.(body)?/, the "must contain" character is
1055 "g". If the first part-match was for the string "dog", restarting with
1056 "sbody" failed. This bug has been fixed.
1058 13. The string returned by pcre_dfa_exec() after a partial match has been
1059 changed so that it starts at the first inspected character rather than the
1060 first character of the match. This makes a difference only if the pattern
1061 starts with a lookbehind assertion or \b or \B (\K is not supported by
1062 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1063 matching functions compatible, and I think it's the right thing to do.
1065 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1066 so that the demonstration program is easily available in environments where
1067 PCRE has not been installed from source.
1069 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1070 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1071 library.
1073 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1074 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1075 is not the first non-POSIX option to be added. Clearly some people find
1076 these options useful.
1078 17. If a caller to the POSIX matching function regexec() passes a non-zero
1079 value for nmatch with a NULL value for pmatch, the value of
1080 nmatch is forced to zero.
1082 18. RunGrepTest did not have a test for the availability of the -u option of
1083 the diff command, as RunTest does. It now checks in the same way as
1084 RunTest, and also checks for the -b option.
1086 19. If an odd number of negated classes containing just a single character
1087 interposed, within parentheses, between a forward reference to a named
1088 subpattern and the definition of the subpattern, compilation crashed with
1089 an internal error, complaining that it could not find the referenced
1090 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1091 [The bug was that it was starting one character too far in when skipping
1092 over the character class, thus treating the ] as data rather than
1093 terminating the class. This meant it could skip too much.]
1095 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1096 /g option in pcretest when the pattern contains \K, which makes it possible
1097 to have an empty string match not at the start, even when the pattern is
1098 anchored. Updated pcretest and pcredemo to use this option.
1100 21. If the maximum number of capturing subpatterns in a recursion was greater
1101 than the maximum at the outer level, the higher number was returned, but
1102 with unset values at the outer level. The correct (outer level) value is
1103 now given.
1105 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1106 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1107 make it do so. The string so far is captured, making this feature
1108 compatible with Perl.
1110 23. The tests have been re-organized, adding tests 11 and 12, to make it
1111 possible to check the Perl 5.10 features against Perl 5.10.
1113 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1114 pattern matches a fixed length string. PCRE did not allow this; now it
1115 does. Neither allows recursion.
1117 25. I finally figured out how to implement a request to provide the minimum
1118 length of subject string that was needed in order to match a given pattern.
1119 (It was back references and recursion that I had previously got hung up
1120 on.) This code has now been added to pcre_study(); it finds a lower bound
1121 to the length of subject needed. It is not necessarily the greatest lower
1122 bound, but using it to avoid searching strings that are too short does give
1123 some useful speed-ups. The value is available to calling programs via
1124 pcre_fullinfo().
1126 26. While implementing 25, I discovered to my embarrassment that pcretest had
1127 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1128 study optimizations had never been tested with that matching function.
1129 Oops. What is worse, even when it was passed study data, there was a bug in
1130 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1131 were also very few tests of studied patterns with pcre_dfa_exec().
1133 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1134 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1135 on the other side of the coin, they are no longer allowed to have different
1136 names, because these cannot be distinguished in PCRE, and this has caused
1137 confusion. (This is a difference from Perl.)
1139 28. When duplicate subpattern names are present (necessarily with different
1140 numbers, as required by 27 above), and a test is made by name in a
1141 conditional pattern, either for a subpattern having been matched, or for
1142 recursion in such a pattern, all the associated numbered subpatterns are
1143 tested, and the overall condition is true if the condition is true for any
1144 one of them. This is the way Perl works, and is also more like the way
1145 testing by number works.
1148 Version 7.9 11-Apr-09
1149 ---------------------
1151 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1152 (pcretest), all targets were linked against these libraries. This included
1153 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1154 libraries. This caused unwanted dependencies to be created. This problem
1155 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1156 pcretest is linked with readline.
1158 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1159 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1160 moved outside it again, because FALSE and TRUE are already defined in AIX,
1161 but BOOL is not.
1163 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1164 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1166 4. The pcregrep documentation said spaces were inserted as well as colons (or
1167 hyphens) following file names and line numbers when outputting matching
1168 lines. This is not true; no spaces are inserted. I have also clarified the
1169 wording for the --colour (or --color) option.
1171 5. In pcregrep, when --colour was used with -o, the list of matching strings
1172 was not coloured; this is different to GNU grep, so I have changed it to be
1173 the same.
1175 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1176 each matching line was coloured. Now it goes on to look for further matches
1177 of any of the test patterns, which is the same behaviour as GNU grep.
1179 7. A pattern that could match an empty string could cause pcregrep to loop; it
1180 doesn't make sense to accept an empty string match in pcregrep, so I have
1181 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1182 seems to be how GNU grep behaves.
1184 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1185 start or after a newline", because the conditional assertion was not being
1186 correctly handled. The rule now is that both the assertion and what follows
1187 in the first alternative must satisfy the test.
1189 9. If auto-callout was enabled in a pattern with a conditional group whose
1190 condition was an assertion, PCRE could crash during matching, both with
1191 pcre_exec() and pcre_dfa_exec().
1193 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1194 used for matching.
1196 11. Unicode property support in character classes was not working for
1197 characters (bytes) greater than 127 when not in UTF-8 mode.
1199 12. Added the -M command line option to pcretest.
1201 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1203 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1205 16. Added comments and documentation about mis-use of no_arg in the C++
1206 wrapper.
1208 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1209 from Martin Jerabek that uses macro names for all relevant character and
1210 string constants.
1212 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1213 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1214 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1215 these, but not everybody uses configure.
1217 19. A conditional group that had only one branch was not being correctly
1218 recognized as an item that could match an empty string. This meant that an
1219 enclosing group might also not be so recognized, causing infinite looping
1220 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1221 with the subject "ab", where knowledge that the repeated group can match
1222 nothing is needed in order to break the loop.
1224 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1225 exec(), but without supplying a callout function, matching went wrong.
1227 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1228 leak if the size of the offset vector was greater than 30. When the vector
1229 is smaller, the saved offsets during recursion go onto a local stack
1230 vector, but for larger vectors malloc() is used. It was failing to free
1231 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1232 error, in fact).
1234 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1235 heapframe that is used only when UTF-8 support is enabled. This caused no
1236 problem, but was untidy.
1238 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1239 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1240 included within another project.
1242 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1243 slightly modified by me:
1245 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1246 not building pcregrep.
1248 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1249 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1251 25. Forward references, both numeric and by name, in patterns that made use of
1252 duplicate group numbers, could behave incorrectly or give incorrect errors,
1253 because when scanning forward to find the reference group, PCRE was not
1254 taking into account the duplicate group numbers. A pattern such as
1255 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1257 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1258 the feature of a custom pointer more persuasive (as requested by a user).
1260 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1261 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1262 pre-defined.
1264 28. Added support for (*UTF8) at the start of a pattern.
1266 29. Arrange for flags added by the "release type" setting in CMake to be shown
1267 in the configuration summary.
1270 Version 7.8 05-Sep-08
1271 ---------------------
1273 1. Replaced UCP searching code with optimized version as implemented for Ad
1274 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1275 stage table and inline lookup instead of a function, giving speed ups of 2
1276 to 5 times on some simple patterns that I tested. Permission was given to
1277 distribute the MultiStage2.py script that generates the tables (it's not in
1278 the tarball, but is in the Subversion repository).
1280 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1281 scripts.
1283 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1284 a group with a zero qualifier. The result of the study could be incorrect,
1285 or the function might crash, depending on the pattern.
1287 4. Caseless matching was not working for non-ASCII characters in back
1288 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1289 It now works when Unicode Property Support is available.
1291 5. In pcretest, an escape such as \x{de} in the data was always generating
1292 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1293 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1294 truncation.
1296 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1298 7. Added two (int) casts to pcregrep when printing the difference of two
1299 pointers, in case they are 64-bit values.
1301 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1302 test 2 if it fails.
1304 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1305 and a #define of that name to empty if it is not externally set. This is to
1306 allow users of MSVC to set it if necessary.
1308 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1309 the convenience functions in the pcre_get.c source file.
1311 11. An option change at the start of a pattern that had top-level alternatives
1312 could cause overwriting and/or a crash. This command provoked a crash in
1313 some environments:
1315 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1317 This potential security problem was recorded as CVE-2008-2371.
1319 12. For a pattern where the match had to start at the beginning or immediately
1320 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1321 pcre_dfa_exec() could read past the end of the passed subject if there was
1322 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1323 pcretest so that it places the subject at the end of its malloc-ed buffer.
1325 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1326 exec() might read past the end of the data buffer in UTF-8 mode.
1328 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1329 the data contained the byte 0x85 as part of a UTF-8 character within its
1330 first line. This applied both to normal and DFA matching.
1332 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1333 /^[^d]*?$/8 failed to match "abc".
1335 16. Added a missing copyright notice to pcrecpp_internal.h.
1337 17. Make it more clear in the documentation that values returned from
1338 pcre_exec() in ovector are byte offsets, not character counts.
1340 18. Tidied a few places to stop certain compilers from issuing warnings.
1342 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1343 supplied by Stefan Weber. I made a further small update for 7.8 because
1344 there is a change of source arrangements: the pcre_searchfuncs.c module is
1345 replaced by pcre_ucd.c.
1348 Version 7.7 07-May-08
1349 ---------------------
1351 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1352 a string to a long long, pretend we don't even have a long long." This is
1353 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1355 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1356 pre-7.6 versions, which defined a global no_arg variable instead of putting
1357 it in the RE class. (See also #8 below.)
1359 3. Remove a line of dead code, identified by coverity and reported by Nuno
1360 Lopes.
1362 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1364 (1) The include/exclude patterns were being applied to the whole pathnames
1365 of files, instead of just to the final components.
1367 (2) If there was more than one level of directory, the subdirectories were
1368 skipped unless they satisfied the include/exclude conditions. This is
1369 inconsistent with GNU grep (and could even be seen as contrary to the
1370 pcregrep specification - which I improved to make it absolutely clear).
1371 The action now is always to scan all levels of directory, and just
1372 apply the include/exclude patterns to regular files.
1374 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1375 --exclude_dir in the tests to avoid scanning .svn directories.
1377 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1378 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1379 doesn't support NULs in patterns.
1381 7. Added some missing "const"s to declarations of static tables in
1382 pcre_compile.c and pcre_dfa_exec.c.
1384 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1385 caused by fix #2 above. (Subsequently also a second patch to fix the
1386 first patch. And a third patch - this was a messy problem.)
1388 9. Applied Craig's patch to remove the use of push_back().
1390 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1391 matching function regexec().
1393 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1394 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1395 references. PCRE supports relative numbers with this syntax (I don't think
1396 Oniguruma does).
1398 12. Previously, a group with a zero repeat such as (...){0} was completely
1399 omitted from the compiled regex. However, this means that if the group
1400 was called as a subroutine from elsewhere in the pattern, things went wrong
1401 (an internal error was given). Such groups are now left in the compiled
1402 pattern, with a new opcode that causes them to be skipped at execution
1403 time.
1405 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1406 to the way PCRE behaves:
1408 (a) A lone ] character is dis-allowed (Perl treats it as data).
1410 (b) A back reference to an unmatched subpattern matches an empty string
1411 (Perl fails the current match path).
1413 (c) A data ] in a character class must be notated as \] because if the
1414 first data character in a class is ], it defines an empty class. (In
1415 Perl it is not possible to have an empty class.) The empty class []
1416 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1417 The negative empty class [^] matches any one character, independently
1418 of the DOTALL setting.
1420 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1421 non-existent subpattern following a character class starting with ']' and
1422 containing () gave an internal compiling error instead of "reference to
1423 non-existent subpattern". Fortunately, when the pattern did exist, the
1424 compiled code was correct. (When scanning forwards to check for the
1425 existencd of the subpattern, it was treating the data ']' as terminating
1426 the class, so got the count wrong. When actually compiling, the reference
1427 was subsequently set up correctly.)
1429 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1430 it was being rejected as not supported by pcre_dfa_exec(), even though
1431 other assertions are supported. I have made pcre_dfa_exec() support
1432 (*FAIL).
1434 16. The implementation of 13c above involved the invention of a new opcode,
1435 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1436 cannot be changed at match time, I realized I could make a small
1437 improvement to matching performance by compiling OP_ALLANY instead of
1438 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1439 on the OP_ANY path.
1441 17. Compiling pcretest on Windows with readline support failed without the
1442 following two fixes: (1) Make the unistd.h include conditional on
1443 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1445 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1446 ncurses library to be included for pcretest when ReadLine support is
1447 requested, but also to allow for it to be overridden. This patch came from
1448 Daniel Bergström.
1450 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1451 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1452 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1453 spotting this.
1456 Version 7.6 28-Jan-08
1457 ---------------------
1459 1. A character class containing a very large number of characters with
1460 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1461 overflow.
1463 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1464 HAVE_LONG_LONG is not defined.
1466 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1467 bring them up to date and include new features. This patch includes:
1469 - Fixed PH's badly added libz and libbz2 support.
1470 - Fixed a problem with static linking.
1471 - Added pcredemo. [But later removed - see 7 below.]
1472 - Fixed dftables problem and added an option.
1473 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1475 - Added readline support for pcretest.
1476 - Added an listing of the option settings after cmake has run.
1478 4. A user submitted a patch to Makefile that makes it easy to create
1479 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1480 Makefile.am that cause it to include this special target, without
1481 affecting anything else. Note that the same mingw target plus all
1482 the other distribution libraries and programs are now supported
1483 when configuring with CMake (see 6 below) instead of with
1484 Configure/Make.
1486 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1487 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1488 exported in the Windows port". It has not yet been confirmed that the patch
1489 solves the problem, but it does no harm.
1491 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1492 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1493 with CMake, and also correct the comment about stack recursion.
1495 7. Remove the automatic building of pcredemo from the ./configure system and
1496 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1497 of a program that users should build themselves after PCRE is installed, so
1498 building it automatically is not really right. What is more, it gave
1499 trouble in some build environments.
1501 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1504 Version 7.5 10-Jan-08
1505 ---------------------
1507 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1508 values in parens when parsing an RE using the C++ wrapper."
1510 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1511 Characters greater than 255 were excluded from the class instead of being
1512 included.
1514 3. The same bug as (2) above applied to negated POSIX classes such as
1515 [:^space:].
1517 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1518 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1519 I have changed it.
1521 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1522 first named subpattern) and a construct such as (?&a) would reference the
1523 first named subpattern whose name started with "a" (in other words, the
1524 length check was missing). Both these problems are fixed. "Subpattern name
1525 expected" is now given for (?&) (a zero-length name), and this patch also
1526 makes it give the same error for \k'' (previously it complained that that
1527 was a reference to a non-existent subpattern).
1529 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1530 this is right because (?- can be followed by option settings as well as by
1531 digits. I have, however, made the messages clearer.
1533 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1534 than the number used in the conditional) now cause a compile-time error.
1535 This is actually not compatible with Perl, which accepts such patterns, but
1536 treats the conditional as always being FALSE (as PCRE used to), but it
1537 seems to me that giving a diagnostic is better.
1539 8. Change "alphameric" to the more common word "alphanumeric" in comments
1540 and messages.
1542 9. Fix two occurrences of "backslash" in comments that should have been
1543 "backspace".
1545 10. Remove two redundant lines of code that can never be obeyed (their function
1546 was moved elsewhere).
1548 11. The program that makes PCRE's Unicode character property table had a bug
1549 which caused it to generate incorrect table entries for sequences of
1550 characters that have the same character type, but are in different scripts.
1551 It amalgamated them into a single range, with the script of the first of
1552 them. In other words, some characters were in the wrong script. There were
1553 thirteen such cases, affecting characters in the following ranges:
1555 U+002b0 - U+002c1
1556 U+0060c - U+0060d
1557 U+0061e - U+00612
1558 U+0064b - U+0065e
1559 U+0074d - U+0076d
1560 U+01800 - U+01805
1561 U+01d00 - U+01d77
1562 U+01d9b - U+01dbf
1563 U+0200b - U+0200f
1564 U+030fc - U+030fe
1565 U+03260 - U+0327f
1566 U+0fb46 - U+0fbb1
1567 U+10450 - U+1049d
1569 12. The -o option (show only the matching part of a line) for pcregrep was not
1570 compatible with GNU grep in that, if there was more than one match in a
1571 line, it showed only the first of them. It now behaves in the same way as
1572 GNU grep.
1574 13. If the -o and -v options were combined for pcregrep, it printed a blank
1575 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1576 does the same. The return code can be used to tell if there were any
1577 non-matching lines.
1579 14. Added --file-offsets and --line-offsets to pcregrep.
1581 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1582 infinitely looping recursion. The bug was that positive lookaheads were not
1583 being skipped when checking for a possible empty match (negative lookaheads
1584 and both kinds of lookbehind were skipped).
1586 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1587 inclusion of <windows.h> to before rather than after the definition of
1588 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1590 17. Specifying a possessive quantifier with a specific limit for a Unicode
1591 character property caused pcre_compile() to compile bad code, which led at
1592 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1593 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1594 caused the error; without that there was no problem.
1596 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1598 19. Added --enable-pcretest-libreadline.
1600 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1601 RE::GlobalReplace(). As a result, the number of replacements returned was
1602 double what it should be. I removed one of the increments, but Craig sent a
1603 later patch that removed the other one (the right fix) and added unit tests
1604 that check the return values (which was not done before).
1606 21. Several CMake things:
1608 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1609 the names libpcre and libpcreposix, not just pcre and pcreposix.
1611 (2) The above change means that pcretest and pcregrep are now correctly
1612 linked with the newly-built libraries, not previously installed ones.
1616 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1617 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1618 UTF-8 newline character). The key issue is that the pattern starts .*;
1619 this means that the match must be either at the beginning, or after a
1620 newline. The bug was in the code for advancing after a failed match and
1621 checking that the new position followed a newline. It was not taking
1622 account of UTF-8 characters correctly.
1624 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1625 character classes. PCRE was not treating the sequence [:...:] as a
1626 character class unless the ... were all letters. Perl, however, seems to
1627 allow any characters between [: and :], though of course it rejects as
1628 unknown any "names" that contain non-letters, because all the known class
1629 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1630 for example, whereas PCRE did not - it did not recognize a POSIX character
1631 class. This seemed a bit dangerous, so the code has been changed to be
1632 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1633 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1634 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1635 Perl does, and where it didn't before.
1637 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1638 Windows environments %n is disabled by default.
1641 Version 7.4 21-Sep-07
1642 ---------------------
1644 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1645 means that a class such as [\s] counted as "explicit reference to CR or
1646 LF". That isn't really right - the whole point of the change was to try to
1647 help when there was an actual mention of one of the two characters. So now
1648 the change happens only if \r or \n (or a literal CR or LF) character is
1649 encountered.
1651 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1652 of both had grown to the point where there were only 3 bits left.
1653 Fortunately, there was spare space in the data structure, and so I have
1654 moved the internal flags into a new 16-bit field to free up more option
1655 bits.
1657 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1658 but did not set the internal JCHANGED flag - either of these is enough to
1659 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1660 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1661 start sets both bits.
1663 4. Added options (at build time, compile time, exec time) to change \R from
1664 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1666 5. doc/pcresyntax.html was missing from the distribution.
1668 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1669 compatibility, even though it is no longer used.
1671 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1672 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1673 windows.h file is present (where different names are used). [This was
1674 reversed later after testing - see 16 below.]
1676 8. Changed all #include <config.h> to #include "config.h". There were also
1677 some further <pcre.h> cases that I changed to "pcre.h".
1679 9. When pcregrep was used with the --colour option, it missed the line ending
1680 sequence off the lines that it output.
1682 10. It was pointed out to me that arrays of string pointers cause lots of
1683 relocations when a shared library is dynamically loaded. A technique of
1684 using a single long string with a table of offsets can drastically reduce
1685 these. I have refactored PCRE in four places to do this. The result is
1686 dramatic:
1688 Originally: 290
1689 After changing UCP table: 187
1690 After changing error message table: 43
1691 After changing table of "verbs" 36
1692 After changing table of Posix names 22
1694 Thanks to the folks working on Gregex for glib for this insight.
1696 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1697 unicode-properties was also set.
1699 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1701 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1702 checked only for CRLF.
1704 14. Added casts to pcretest.c to avoid compiler warnings.
1706 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1708 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1709 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1710 entirely. This removes changes made in 7 above.
1712 17. The CMake files have been updated, and there is now more information about
1713 building with CMake in the NON-UNIX-USE document.
1716 Version 7.3 28-Aug-07
1717 ---------------------
1719 1. In the rejigging of the build system that eventually resulted in 7.1, the
1720 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1721 brackets there is not right, since it causes compilers to look for an
1722 installed pcre.h, not the version that is in the source that is being
1723 compiled (which of course may be different). I have changed it back to:
1725 #include "pcre.h"
1727 I have a vague recollection that the change was concerned with compiling in
1728 different directories, but in the new build system, that is taken care of
1729 by the VPATH setting the Makefile.
1731 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1732 when the subject happened to end in the byte 0x85 (e.g. if the last
1733 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1734 characters but of course it shouldn't be taken as a newline when it is part
1735 of another character. The bug was that, for an unlimited repeat of . in
1736 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1737 characters when looking for a newline.
1739 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1741 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1742 in debug output.
1744 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1745 long printing in the pcrecpp unittest when running under MinGW.
1747 6. ESC_K was left out of the EBCDIC table.
1749 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1750 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1751 limit also applies to "virtual nesting" when a pattern is recursive, and in
1752 this case 1000 isn't so big. I have been able to remove this limit at the
1753 expense of backing off one optimization in certain circumstances. Normally,
1754 when pcre_exec() would call its internal match() function recursively and
1755 immediately return the result unconditionally, it uses a "tail recursion"
1756 feature to save stack. However, when a subpattern that can match an empty
1757 string has an unlimited repetition quantifier, it no longer makes this
1758 optimization. That gives it a stack frame in which to save the data for
1759 checking that an empty string has been matched. Previously this was taken
1760 from the 1000-entry workspace that had been reserved. So now there is no
1761 explicit limit, but more stack is used.
1763 8. Applied Daniel's patches to solve problems with the import/export magic
1764 syntax that is required for Windows, and which was going wrong for the
1765 pcreposix and pcrecpp parts of the library. These were overlooked when this
1766 problem was solved for the main library.
1768 9. There were some crude static tests to avoid integer overflow when computing
1769 the size of patterns that contain repeated groups with explicit upper
1770 limits. As the maximum quantifier is 65535, the maximum group length was
1771 set at 30,000 so that the product of these two numbers did not overflow a
1772 32-bit integer. However, it turns out that people want to use groups that
1773 are longer than 30,000 bytes (though not repeat them that many times).
1774 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1775 made it possible to implement the integer overflow checks in a much more
1776 dynamic way, which I have now done. The artificial limitation on group
1777 length has been removed - we now have only the limit on the total length of
1778 the compiled pattern, which depends on the LINK_SIZE setting.
1780 10. Fixed a bug in the documentation for get/copy named substring when
1781 duplicate names are permitted. If none of the named substrings are set, the
1782 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1783 empty string.
1785 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1786 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1787 because the ] is interpreted as the first data character and the
1788 terminating ] is not found. PCRE has been made compatible with Perl in this
1789 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1790 cause memory overwriting.
1792 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1793 string has been matched (to stop an infinite loop). It was not recognizing
1794 a conditional subpattern that could match an empty string if that
1795 subpattern was within another subpattern. For example, it looped when
1796 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1797 condition was not nested. This bug has been fixed.
1799 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1800 past the start of the subject in the presence of bytes with the top bit
1801 set, for example "\x8aBCD".
1803 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1804 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1806 14. Optimized (?!) to (*FAIL).
1808 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1809 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1810 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1811 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1812 does: it's just the validity check that is more restrictive.
1814 16. Inserted checks for integer overflows during escape sequence (backslash)
1815 processing, and also fixed erroneous offset values for syntax errors during
1816 backslash processing.
1818 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1819 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1821 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1822 caused an overrun.
1824 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1825 something other than just ASCII characters) inside a group that had an
1826 unlimited repeat caused a loop at compile time (while checking to see
1827 whether the group could match an empty string).
1829 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1830 [\P{Any}] did so. (Error in the code for printing property names.)
1832 21. An orphan \E inside a character class could cause a crash.
1834 22. A repeated capturing bracket such as (A)? could cause a wild memory
1835 reference during compilation.
1837 23. There are several functions in pcre_compile() that scan along a compiled
1838 expression for various reasons (e.g. to see if it's fixed length for look
1839 behind). There were bugs in these functions when a repeated \p or \P was
1840 present in the pattern. These operators have additional parameters compared
1841 with \d, etc, and these were not being taken into account when moving along
1842 the compiled data. Specifically:
1844 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1845 length.
1847 (b) An item such as \pL+ within a repeated group could cause crashes or
1848 loops.
1850 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1851 "reference to non-existent subpattern" error.
1853 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1855 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1856 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1858 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1860 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1861 character were causing crashes (broken optimization).
1863 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1864 \p or \P) caused a compile-time loop.
1866 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1867 break. For example, the unstudied pattern [\r\n]A does not match the string
1868 "\r\nA" because change 7.0/46 below moves the current point on by two
1869 characters after failing to match at the start. However, the pattern \nA
1870 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1871 the same is true. There doesn't seem any very clean way out of this, but
1872 what I have chosen to do makes the common cases work: PCRE now takes note
1873 of whether there can be an explicit match for \r or \n anywhere in the
1874 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1875 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1876 pattern has explicit CR or LF references.
1878 29. Added (*CR) etc for changing newline setting at start of pattern.
1881 Version 7.2 19-Jun-07
1882 ---------------------
1884 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1885 which is apparently normally available under Windows.
1887 2. Re-jig the pcregrep tests with different newline settings in an attempt
1888 to make them independent of the local environment's newline setting.
1890 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1892 4. Some of the "internals" tests were previously cut out when the link size
1893 was not 2, because the output contained actual offsets. The recent new
1894 "Z" feature of pcretest means that these can be cut out, making the tests
1895 usable with all link sizes.
1897 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1898 stack recursion. This gives a massive performance boost under BSD, but just
1899 a small improvement under Linux. However, it saves one field in the frame
1900 in all cases.
1902 6. Added more features from the forthcoming Perl 5.10:
1904 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1905 recursion call. It refers to the nth most recently opened parentheses.
1907 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1908 to be opened parentheses.
1910 (c) Conditions that refer to capturing parentheses can be specified
1911 relatively, for example, (?(-2)... or (?(+3)...
1913 (d) \K resets the start of the current match so that everything before
1914 is not part of it.
1916 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1918 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1919 reference syntax.
1921 (g) (?| introduces a group in which the numbering of parentheses in each
1922 alternative starts with the same number.
1924 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1926 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1929 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1930 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1931 for detecting groups that can match an empty string.
1933 9. A pattern with a very large number of alternatives (more than several
1934 hundred) was running out of internal workspace during the pre-compile
1935 phase, where pcre_compile() figures out how much memory will be needed. A
1936 bit of new cunning has reduced the workspace needed for groups with
1937 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1938 workspace instead of running out of the 4096 that are available.
1940 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1942 11. Applied patch from Google to remove an optimization that didn't quite work.
1943 The report of the bug said:
1945 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1946 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1947 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1949 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1950 it matched the wrong number of bytes.
1953 Version 7.1 24-Apr-07
1954 ---------------------
1956 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1957 that is more "standard", making use of automake and other Autotools. There
1958 is some re-arrangement of the files and adjustment of comments consequent
1959 on this.
1961 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1962 for recursive directory scanning broke on some systems because the files
1963 are not scanned in any specific order and on different systems the order
1964 was different. A call to "sort" has been inserted into RunGrepTest for the
1965 approprate test as a short-term fix. In the longer term there may be an
1966 alternative.
1968 3. I had an email from Eric Raymond about problems translating some of PCRE's
1969 man pages to HTML (despite the fact that I distribute HTML pages, some
1970 people do their own conversions for various reasons). The problems
1971 concerned the use of low-level troff macros .br and .in. I have therefore
1972 removed all such uses from the man pages (some were redundant, some could
1973 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1974 HTML has been updated to handle .nf/.fi and to complain if it encounters
1975 .br or .in.
1977 4. Updated comments in configure.ac that get placed in config.h.in and also
1978 arranged for config.h to be included in the distribution, with the name
1979 config.h.generic, for the benefit of those who have to compile without
1980 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1982 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1983 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1984 makevp.bat for the current PCRE, using the additional files
1985 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1987 6. A Windows user reported a minor discrepancy with test 2, which turned out
1988 to be caused by a trailing space on an input line that had got lost in his
1989 copy. The trailing space was an accident, so I've just removed it.
1991 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1992 that is needed.
1994 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1995 as "const" (a) because they are and (b) because it helps the PHP
1996 maintainers who have recently made a script to detect big data structures
1997 in the php code that should be moved to the .rodata section. I remembered
1998 to update Builducptable as well, so it won't revert if ucptable.h is ever
1999 re-created.
2001 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2002 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2003 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2004 support is not required. This saves 1.5-2K of code, which is important in
2005 some applications.
2007 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2008 so as not to refer to the tables, even though these functions will never be
2009 called when UTF-8 support is disabled. Otherwise there are problems with a
2010 shared library.
2012 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2014 (a) It was defining its arguments as char * instead of void *.
2016 (b) It was assuming that all moves were upwards in memory; this was true
2017 a long time ago when I wrote it, but is no longer the case.
2019 The emulated memove() is provided for those environments that have neither
2020 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2021 is clearly not the case, as these two bugs were recently reported.
2023 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2024 and Detrail to create the HTML documentation, the .txt form of the man
2025 pages, and it removes trailing spaces from listed files. It also creates
2026 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2027 case, it wraps all the #defines with #ifndefs. This script should be run
2028 before "make dist".
2030 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2031 with Unicode property support.
2033 (a) For a maximizing quantifier, if the two different cases of the
2034 character were of different lengths in their UTF-8 codings (there are
2035 some cases like this - I found 11), and the matching function had to
2036 back up over a mixture of the two cases, it incorrectly assumed they
2037 were both the same length.
2039 (b) When PCRE was configured to use the heap rather than the stack for
2040 recursion during matching, it was not correctly preserving the data for
2041 the other case of a UTF-8 character when checking ahead for a match
2042 while processing a minimizing repeat. If the check also involved
2043 matching a wide character, but failed, corruption could cause an
2044 erroneous result when trying to check for a repeat of the original
2045 character.
2047 13. Some tidying changes to the testing mechanism:
2049 (a) The RunTest script now detects the internal link size and whether there
2050 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2051 values substituted by "configure". (The RunGrepTest script already did
2052 this for UTF-8.) The configure.ac script no longer substitutes the
2053 relevant variables.
2055 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2056 with length and offset values. This means that the output is different
2057 for different internal link sizes. Test 2 is skipped for link sizes
2058 other than 2 because of this, bypassing the problem. Unfortunately,
2059 there was also a test in test 3 (the locale tests) that used /B and
2060 failed for link sizes other than 2. Rather than cut the whole test out,
2061 I have added a new /Z option to pcretest that replaces the length and
2062 offset values with spaces. This is now used to make test 3 independent
2063 of link size. (Test 2 will be tidied up later.)
2065 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2066 segmentation fault instead of returning the appropriate error message.
2068 15. In multiline mode when the newline sequence was set to "any", the pattern
2069 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2070 This doesn't seem right; it now treats the CRLF combination as the line
2071 ending, and so does not match in that case. It's only a pattern such as ^$
2072 that would hit this one: something like ^ABC$ would have failed after \r
2073 and then tried again after \r\n.
2075 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2076 in an attempt to make files that differ only in their line terminators
2077 compare equal. This works on Linux.
2079 17. Under certain error circumstances pcregrep might try to free random memory
2080 as it exited. This is now fixed, thanks to valgrind.
2082 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2083 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2084 was because its rules for how to advance for /g after matching an empty
2085 string at the end of a line did not allow for this case. They now check for
2086 it specially.
2088 20. pcretest is supposed to handle patterns and data of any length, by
2089 extending its buffers when necessary. It was getting this wrong when the
2090 buffer for a data line had to be extended.
2092 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2093 CRLF as a newline sequence.
2095 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2096 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2097 I have nevertheless tidied it up.
2099 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2101 24. Added a man page for pcre-config.
2104 Version 7.0 19-Dec-06
2105 ---------------------
2107 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2108 moving to gcc 4.1.1.
2110 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2111 sys/time.h, which is documented as needed for this function. It doesn't
2112 seem to matter on Linux, but it showed up on some releases of OS X.
2114 3. It seems that there are systems where bytes whose values are greater than
2115 127 match isprint() in the "C" locale. The "C" locale should be the
2116 default when a C program starts up. In most systems, only ASCII printing
2117 characters match isprint(). This difference caused the output from pcretest
2118 to vary, making some of the tests fail. I have changed pcretest so that:
2120 (a) When it is outputting text in the compiled version of a pattern, bytes
2121 other than 32-126 are always shown as hex escapes.
2123 (b) When it is outputting text that is a matched part of a subject string,
2124 it does the same, unless a different locale has been set for the match
2125 (using the /L modifier). In this case, it uses isprint() to decide.
2127 4. Fixed a major bug that caused incorrect computation of the amount of memory
2128 required for a compiled pattern when options that changed within the
2129 pattern affected the logic of the preliminary scan that determines the
2130 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2131 that the computed length was too small. The symptoms of this bug were
2132 either the PCRE error "internal error: code overflow" from pcre_compile(),
2133 or a glibc crash with a message such as "pcretest: free(): invalid next
2134 size (fast)". Examples of patterns that provoked this bug (shown in
2135 pcretest format) are:
2137 /(?-x: )/x
2138 /(?x)(?-x: \s*#\s*)/
2139 /((?i)[\x{c0}])/8
2140 /(?i:[\x{c0}])/8
2142 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2143 is now done differently.
2145 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2146 wrapper classes; (b) implement a new function in the C++ scanner that is
2147 more efficient than the old way of doing things because it avoids levels of
2148 recursion in the regex matching; (c) add a paragraph to the documentation
2149 for the FullMatch() function.
2151 6. The escape sequence \n was being treated as whatever was defined as
2152 "newline". Not only was this contrary to the documentation, which states
2153 that \n is character 10 (hex 0A), but it also went horribly wrong when
2154 "newline" was defined as CRLF. This has been fixed.
2156 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2157 was being set to -1 for the "end of line" case (supposedly a value that no
2158 character can have). Though this value is never used (the check for end of
2159 line is "zero bytes in current character"), it caused compiler complaints.
2160 I've changed it to 0xffffffff.
2162 8. In pcre_version.c, the version string was being built by a sequence of
2163 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2164 string (as it is for production releases) called a macro with an empty
2165 argument. The C standard says the result of this is undefined. The gcc
2166 compiler treats it as an empty string (which was what was wanted) but it is
2167 reported that Visual C gives an error. The source has been hacked around to
2168 avoid this problem.
2170 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2171 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2172 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2173 of them did).
2175 10. Originally, pcretest opened its input and output without "b"; then I was
2176 told that "b" was needed in some environments, so it was added for release
2177 5.0 to both the input and output. (It makes no difference on Unix-like
2178 systems.) Later I was told that it is wrong for the input on Windows. I've
2179 now abstracted the modes into two macros, to make it easier to fiddle with
2180 them, and removed "b" from the input mode under Windows.
2182 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2184 12. Added -help and --help to pcretest as an official way of being reminded
2185 of the options.
2187 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2188 and pcrecpp.cc because they annoy compilers at high warning levels.
2190 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2192 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2193 scripts are not C programs :-) and which was not noticed because it works
2194 on Linux.
2196 16. pcretest is supposed to handle any length of pattern and data line (as one
2197 line or as a continued sequence of lines) by extending its input buffer if
2198 necessary. This feature was broken for very long pattern lines, leading to
2199 a string of junk being passed to pcre_compile() if the pattern was longer
2200 than about 50K.
2202 17. I have done a major re-factoring of the way pcre_compile() computes the
2203 amount of memory needed for a compiled pattern. Previously, there was code
2204 that made a preliminary scan of the pattern in order to do this. That was
2205 OK when PCRE was new, but as the facilities have expanded, it has become
2206 harder and harder to keep it in step with the real compile phase, and there
2207 have been a number of bugs (see for example, 4 above). I have now found a
2208 cunning way of running the real compile function in a "fake" mode that
2209 enables it to compute how much memory it would need, while actually only
2210 ever using a few hundred bytes of working memory and without too many
2211 tests of the mode. This should make future maintenance and development
2212 easier. A side effect of this work is that the limit of 200 on the nesting
2213 depth of parentheses has been removed (though this was never a serious
2214 limitation, I suspect). However, there is a downside: pcre_compile() now
2215 runs more slowly than before (30% or more, depending on the pattern). I
2216 hope this isn't a big issue. There is no effect on runtime performance.
2218 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2219 newline (only possible for the last line of a file) and it was a
2220 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2222 19. Added additional timing features to pcretest. (1) The -tm option now times
2223 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2224 separate command line item, by a number that specifies the number of
2225 repeats to use when timing. The default is 50000; this gives better
2226 precision, but takes uncomfortably long for very large patterns.
2228 20. Extended pcre_study() to be more clever in cases where a branch of a
2229 subpattern has no definite first character. For example, (a*|b*)[cd] would
2230 previously give no result from pcre_study(). Now it recognizes that the
2231 first character must be a, b, c, or d.
2233 21. There was an incorrect error "recursive call could loop indefinitely" if
2234 a subpattern (or the entire pattern) that was being tested for matching an
2235 empty string contained only one non-empty item after a nested subpattern.
2236 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2237 incorrectly, because the \d was being skipped in the check.
2239 22. The pcretest program now has a new pattern option /B and a command line
2240 option -b, which is equivalent to adding /B to every pattern. This causes
2241 it to show the compiled bytecode, without the additional information that
2242 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2243 is the same as /B/I).
2245 23. A new optimization is now able automatically to treat some sequences such
2246 as a*b as a*+b. More specifically, if something simple (such as a character
2247 or a simple class like \d) has an unlimited quantifier, and is followed by
2248 something that cannot possibly match the quantified thing, the quantifier
2249 is automatically "possessified".
2251 24. A recursive reference to a subpattern whose number was greater than 39
2252 went wrong under certain circumstances in UTF-8 mode. This bug could also
2253 have affected the operation of pcre_study().
2255 25. Realized that a little bit of performance could be had by replacing
2256 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2258 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2260 27. Possessive quantifiers such as a++ were previously implemented by turning
2261 them into atomic groups such as ($>a+). Now they have their own opcodes,
2262 which improves performance. This includes the automatically created ones
2263 from 23 above.
2265 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2266 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2267 the first matched character to be a colon. This applied both to named and
2268 numbered groups.
2270 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2272 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2273 building PCRE on Windows easier, so I have included it in the distribution.
2275 31. There is now a check in pcretest against a ridiculously large number being
2276 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2277 loop, the loop is abandoned.
2279 32. Forward references to subpatterns in conditions such as (?(2)...) where
2280 subpattern 2 is defined later cause pcre_compile() to search forwards in
2281 the pattern for the relevant set of parentheses. This search went wrong
2282 when there were unescaped parentheses in a character class, parentheses
2283 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2285 33. "Subroutine" calls and backreferences were previously restricted to
2286 referencing subpatterns earlier in the regex. This restriction has now
2287 been removed.
2289 34. Added a number of extra features that are going to be in Perl 5.10. On the
2290 whole, these are just syntactic alternatives for features that PCRE had
2291 previously implemented using the Python syntax or my own invention. The
2292 other formats are all retained for compatibility.
2294 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2295 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2296 also .NET compatible.
2298 (b) A recursion or subroutine call to a named group can now be defined as
2299 (?&name) as well as (?P>name).
2301 (c) A backreference to a named group can now be defined as \k<name> or
2302 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2303 5.10, are also .NET compatible.
2305 (d) A conditional reference to a named group can now use the syntax
2306 (?(<name>) or (?('name') as well as (?(name).
2308 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2309 groups (named and numbered) that are never evaluated inline, but can be
2310 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2311 is always false. There may be only one alternative in such a group.
2313 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2314 as the simple (?(R). The condition is true only if the most recent
2315 recursion is that of the given number or name. It does not search out
2316 through the entire recursion stack.
2318 (g) The escape \gN or \g{N} has been added, where N is a positive or
2319 negative number, specifying an absolute or relative reference.
2321 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2322 some "unreachable code" warnings.
2324 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2325 things, this adds five new scripts.
2327 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2328 There were also incompatibilities regarding the handling of \Q..\E inside
2329 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2330 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2332 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2333 matches an empty string, and forcibly breaks the loop. There were bugs in
2334 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2335 against aaaa the result was just "a" rather than "aaaa", for example. Two
2336 separate and independent bugs (that affected different cases) have been
2337 fixed.
2339 39. Refactored the code to abolish the use of different opcodes for small
2340 capturing bracket numbers. This is a tidy that I avoided doing when I
2341 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2342 The new approach is not only tidier, it makes it possible to reduce the
2343 memory needed to fix the previous bug (38).
2345 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2346 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2347 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2348 mode.
2350 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2351 report.
2353 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2354 copy construction and assignment in the C++ wrapper.
2356 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2357 couple of bugs that could have given wrong results in the "--newline=crlf"
2358 case.
2360 44. Added a number of casts and did some reorganization of signed/unsigned int
2361 variables following suggestions from Dair Grant. Also renamed the variable
2362 "this" as "item" because it is a C++ keyword.
2364 45. Arranged for dftables to add
2366 #include "pcre_internal.h"
2368 to pcre_chartables.c because without it, gcc 4.x may remove the array
2369 definition from the final binary if PCRE is built into a static library and
2370 dead code stripping is activated.
2372 46. For an unanchored pattern, if a match attempt fails at the start of a
2373 newline sequence, and the newline setting is CRLF or ANY, and the next two
2374 characters are CRLF, advance by two characters instead of one.
2377 Version 6.7 04-Jul-06
2378 ---------------------
2380 1. In order to handle tests when input lines are enormously long, pcretest has
2381 been re-factored so that it automatically extends its buffers when
2382 necessary. The code is crude, but this _is_ just a test program. The
2383 default size has been increased from 32K to 50K.
2385 2. The code in pcre_study() was using the value of the re argument before
2386 testing it for NULL. (Of course, in any sensible call of the function, it
2387 won't be NULL.)
2389 3. The memmove() emulation function in pcre_internal.h, which is used on
2390 systems that lack both memmove() and bcopy() - that is, hardly ever -
2391 was missing a "static" storage class specifier.
2393 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2394 containing an extended class (one that cannot be represented by a bitmap
2395 because it contains high-valued characters or Unicode property items, e.g.
2396 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2397 pattern, but PCRE should not loop if you do not (it no longer does).
2398 [Detail: two cases were found: (a) a repeated subpattern containing an
2399 extended class; (b) a recursive reference to a subpattern that followed a
2400 previous extended class. It wasn't skipping over the extended class
2401 correctly when UTF-8 mode was not set.]
2403 5. A negated single-character class was not being recognized as fixed-length
2404 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2405 compile error "lookbehind assertion is not fixed length".
2407 6. The RunPerlTest auxiliary script was showing an unexpected difference
2408 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2409 write a Perl script that can interpret lines of an input file either as
2410 byte characters or as UTF-8, which is what "perltest" was being required to
2411 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2412 can't do is switch easily at run time between having the "use utf8;" pragma
2413 or not. In the end, I fudged it by using the RunPerlTest script to insert
2414 "use utf8;" explicitly for the UTF-8 tests.
2416 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2417 the end of the subject string, contrary to the documentation and to what
2418 Perl does. This was true of both matching functions. Now it matches only at
2419 the start of the subject and immediately after *internal* newlines.
2421 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2422 a pointer to an int instead of a pointer to an unsigned long int. This
2423 caused problems on 64-bit systems.
2425 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2426 instance of the 'standard' template library not being so standard".
2428 10. There was no check on the number of named subpatterns nor the maximum
2429 length of a subpattern name. The product of these values is used to compute
2430 the size of the memory block for a compiled pattern. By supplying a very
2431 long subpattern name and a large number of named subpatterns, the size
2432 computation could be caused to overflow. This is now prevented by limiting
2433 the length of names to 32 characters, and the number of named subpatterns
2434 to 10,000.
2436 11. Subpatterns that are repeated with specific counts have to be replicated in
2437 the compiled pattern. The size of memory for this was computed from the
2438 length of the subpattern and the repeat count. The latter is limited to
2439 65535, but there was no limit on the former, meaning that integer overflow
2440 could in principle occur. The compiled length of a repeated subpattern is
2441 now limited to 30,000 bytes in order to prevent this.
2443 12. Added the optional facility to have named substrings with the same name.
2445 13. Added the ability to use a named substring as a condition, using the
2446 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2447 are numbers (not recommended). Forward references are permitted.
2449 14. Added forward references in named backreferences (if you see what I mean).
2451 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2452 pattern could run off the end of the subject. For example, the pattern
2453 "(?s)(.{1,5})"8 did this with the subject "ab".
2455 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2456 PCRE_CASELESS was set when matching characters that were quantified with ?
2457 or *.
2459 17. A character class other than a single negated character that had a minimum
2460 but no maximum quantifier - for example [ab]{6,} - was not handled
2461 correctly by pce_dfa_exec(). It would match only one character.
2463 18. A valid (though odd) pattern that looked like a POSIX character
2464 class but used an invalid character after [ (for example [[,abc,]]) caused
2465 pcre_compile() to give the error "Failed: internal error: code overflow" or
2466 in some cases to crash with a glibc free() error. This could even happen if
2467 the pattern terminated after [[ but there just happened to be a sequence of
2468 letters, a binary zero, and a closing ] in the memory that followed.
2470 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2471 over the years. Originally (before any Unicode support), just the bottom 8
2472 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2473 output from "man perlunicode" includes this:
2475 The regular expression compiler produces polymorphic opcodes. That
2476 is, the pattern adapts to the data and automatically switches to
2477 the Unicode character scheme when presented with Unicode data--or
2478 instead uses a traditional byte scheme when presented with byte
2479 data.
2481 Sadly, a wide octal escape does not cause a switch, and in a string with
2482 no other multibyte characters, these octal escapes are treated as before.
2483 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2484 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2485 Unicode string.
2487 I have not perpetrated such confusion in PCRE. Up till now, it took just
2488 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2489 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2490 translate to the appropriate multibyte character.
2492 29. Applied some refactoring to reduce the number of warnings from Microsoft
2493 and Borland compilers. This has included removing the fudge introduced
2494 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2495 a warning about an unused variable.
2497 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2498 characters since release 4.0, because Perl (from release 5.004) does not.
2499 [Or at least, is documented not to: some releases seem to be in conflict
2500 with the documentation.] However, when a pattern was studied with
2501 pcre_study() and all its branches started with \s, PCRE still included VT
2502 as a possible starting character. Of course, this did no harm; it just
2503 caused an unnecessary match attempt.
2505 22. Removed a now-redundant internal flag bit that recorded the fact that case
2506 dependency changed within the pattern. This was once needed for "required
2507 byte" processing, but is no longer used. This recovers a now-scarce options
2508 bit. Also moved the least significant internal flag bit to the most-
2509 significant bit of the word, which was not previously used (hangover from
2510 the days when it was an int rather than a uint) to free up another bit for
2511 the future.
2513 23. Added support for CRLF line endings as well as CR and LF. As well as the
2514 default being selectable at build time, it can now be changed at runtime
2515 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2516 specify that it is scanning data with non-default line endings.
2518 24. Changed the definition of CXXLINK to make it agree with the definition of
2519 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2521 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2522 recursions. This makes a big different to stack usage for some patterns.
2524 26. If a subpattern containing a named recursion or subroutine reference such
2525 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2526 the space required for the compiled pattern went wrong and gave too small a
2527 value. Depending on the environment, this could lead to "Failed: internal
2528 error: code overflow at offset 49" or "glibc detected double free or
2529 corruption" errors.
2531 27. Applied patches from Google (a) to support the new newline modes and (b) to
2532 advance over multibyte UTF-8 characters in GlobalReplace.
2534 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2535 difference for some implementation of PCRE in some Windows version.
2537 29. Added some extra testing facilities to pcretest:
2539 \q<number> in a data line sets the "match limit" value
2540 \Q<number> in a data line sets the "match recursion limt" value
2541 -S <number> sets the stack size, where <number> is in megabytes
2543 The -S option isn't available for Windows.
2546 Version 6.6 06-Feb-06
2547 ---------------------
2549 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2550 in pcreposix.h. I have copied the definition from pcre.h.
2552 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2553 because pcre.h is no longer a built file.
2555 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2556 not normally included in the compiled code.
2559 Version 6.5 01-Feb-06
2560 ---------------------
2562 1. When using the partial match feature with pcre_dfa_exec(), it was not
2563 anchoring the second and subsequent partial matches at the new starting
2564 point. This could lead to incorrect results. For example, with the pattern
2565 /1234/, partially matching against "123" and then "a4" gave a match.
2567 2. Changes to pcregrep:
2569 (a) All non-match returns from pcre_exec() were being treated as failures
2570 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2571 error message is output. Some extra information is given for the
2573 probably the only errors that are likely to be caused by users (by
2574 specifying a regex that has nested indefinite repeats, for instance).
2575 If there are more than 20 of these errors, pcregrep is abandoned.
2577 (b) A binary zero was treated as data while matching, but terminated the
2578 output line if it was written out. This has been fixed: binary zeroes
2579 are now no different to any other data bytes.
2581 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2582 used to set a locale for matching. The --locale=xxxx long option has
2583 been added (no short equivalent) to specify a locale explicitly on the
2584 pcregrep command, overriding the environment variables.
2586 (d) When -B was used with -n, some line numbers in the output were one less
2587 than they should have been.
2589 (e) Added the -o (--only-matching) option.
2591 (f) If -A or -C was used with -c (count only), some lines of context were
2592 accidentally printed for the final match.
2594 (g) Added the -H (--with-filename) option.
2596 (h) The combination of options -rh failed to suppress file names for files
2597 that were found from directory arguments.
2599 (i) Added the -D (--devices) and -d (--directories) options.
2601 (j) Added the -F (--fixed-strings) option.
2603 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2605 (l) Added the --colo(u)r option.
2607 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2608 is not present by default.
2610 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2611 items such as (?R) or (?1), when the recursion could match a number of
2612 alternatives. If it matched one of the alternatives, but subsequently,
2613 outside the recursion, there was a failure, the code tried to back up into
2614 the recursion. However, because of the way PCRE is implemented, this is not
2615 possible, and the result was an incorrect result from the match.
2617 In order to prevent this happening, the specification of recursion has
2618 been changed so that all such subpatterns are automatically treated as
2619 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2621 4. I had overlooked the fact that, in some locales, there are characters for
2622 which isalpha() is true but neither isupper() nor islower() are true. In
2623 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2624 and ordfeminine) are like this. This affected the treatment of \w and \W
2625 when they appeared in character classes, but not when they appeared outside
2626 a character class. The bit map for "word" characters is now created
2627 separately from the results of isalnum() instead of just taking it from the
2628 upper, lower, and digit maps. (Plus the underscore character, of course.)
2630 5. The above bug also affected the handling of POSIX character classes such as
2631 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2632 permanent tables. Instead, the bit maps for such a class were previously
2633 created as the appropriate unions of the upper, lower, and digit bitmaps.
2634 Now they are created by subtraction from the [[:word:]] class, which has
2635 its own bitmap.
2637 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2638 It is created by subtracting the vertical space characters (\x09, \x0a,
2639 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2640 subtraction was done in the overall bitmap for a character class, meaning
2641 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2642 be recognized. This bug has been fixed.
2644 7. Patches from the folks at Google:
2646 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2647 real life, but is still worth protecting against".
2649 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2650 regular expressions".
2652 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2653 have it.
2655 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2656 "configure" and the latter not, in order to fix a problem somebody had
2657 with compiling the Arg class on HP-UX.
2659 (e) Improve the error-handling of the C++ wrapper a little bit.
2661 (f) New tests for checking recursion limiting.
2663 8. The pcre_memmove() function, which is used only if the environment does not
2664 have a standard memmove() function (and is therefore rarely compiled),
2665 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2666 returning a result (though PCRE never actually uses the result).
2668 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2669 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2670 returned instead of calling malloc() with an overflowing number that would
2671 most likely cause subsequent chaos.
2673 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2675 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2676 with this option is matched, the nmatch and pmatch options of regexec() are
2677 ignored.
2679 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2680 provided in case anyone wants to the the POSIX interface with UTF-8
2681 strings.
2683 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2684 C++ linking (needed for some HP-UX environments).
2686 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2687 (unused parameter) and in the pcre_printint() function (omitted "default"
2688 switch label when the default is to do nothing).
2690 15. Added some code to make it possible, when PCRE is compiled as a C++
2691 library, to replace subject pointers for pcre_exec() with a smart pointer
2692 class, thus making it possible to process discontinuous strings.
2694 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2695 much the same function. They were added by different people who were trying
2696 to make PCRE easy to compile on non-Unix systems. It has been suggested
2697 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2698 for compiling on Windows systems. I have therefore replaced it with
2699 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2700 defaults to "extern" for C or "extern C" for C++, which works fine on
2701 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2702 SCOPE with something explicit in config.h. In addition:
2704 (a) pcreposix.h still had just "extern" instead of either of these macros;
2705 I have replaced it with PCRE_DATA_SCOPE.
2707 (b) Functions such as _pcre_xclass(), which are internal to the library,
2708 but external in the C sense, all had PCRE_EXPORT in their definitions.
2709 This is apparently wrong for the Windows case, so I have removed it.
2710 (It makes no difference on Unix-like systems.)
2712 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2713 of recursive calls to match(). This is different to MATCH_LIMIT because
2714 that limits the total number of calls to match(), not all of which increase
2715 the depth of recursion. Limiting the recursion depth limits the amount of
2716 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2717 when PCRE is compiled, and changed at run time. A patch from Google adds
2718 this functionality to the C++ interface.
2720 18. Changes to the handling of Unicode character properties:
2722 (a) Updated the table to Unicode 4.1.0.
2724 (b) Recognize characters that are not in the table as "Cn" (undefined).
2726 (c) I revised the way the table is implemented to a much improved format
2727 which includes recognition of ranges. It now supports the ranges that
2728 are defined in UnicodeData.txt, and it also amalgamates other
2729 characters into ranges. This has reduced the number of entries in the
2730 table from around 16,000 to around 3,000, thus reducing its size
2731 considerably. I realized I did not need to use a tree structure after
2732 all - a binary chop search is just as efficient. Having reduced the
2733 number of entries, I extended their size from 6 bytes to 8 bytes to
2734 allow for more data.
2736 (d) Added support for Unicode script names via properties such as \p{Han}.
2738 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2739 matching that character.
2741 20. When matching a repeated Unicode property with a minimum greater than zero,
2742 (for example \pL{2,}), PCRE could look past the end of the subject if it
2743 reached it while seeking the minimum number of characters. This could
2744 happen only if some of the characters were more than one byte long, because
2745 there is a check for at least the minimum number of bytes.
2747 21. Refactored the implementation of \p and \P so as to be more general, to
2748 allow for more different types of property in future. This has changed the
2749 compiled form incompatibly. Anybody with saved compiled patterns that use
2750 \p or \P will have to recompile them.
2752 22. Added "Any" and "L&" to the supported property types.
2754 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2755 but give a compile time error if the value is greater than 0xff.
2757 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2758 accidentally not being installed or uninstalled.
2760 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2761 made were to insert the current release number. This seemed silly, because
2762 it made things harder for people building PCRE on systems that don't run
2763 "configure". I have turned pcre.h into a distributed file, no longer built
2764 by "configure", with the version identification directly included. There is
2765 no longer a pcre.h.in file.
2767 However, this change necessitated a change to the pcre-config script as
2768 well. It is built from pcre-config.in, and one of the substitutions was the
2769 release number. I have updated configure.ac so that ./configure now finds
2770 the release number by grepping pcre.h.
2772 26. Added the ability to run the tests under valgrind.
2775 Version 6.4 05-Sep-05
2776 ---------------------
2778 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2779 "--" to be printed when multiple files were scanned, even when none of the
2780 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2781 consider it to be a bug, and have restored the previous behaviour.
2783 2. A couple of code tidies to get rid of compiler warnings.
2785 3. The pcretest program used to cheat by referring to symbols in the library
2786 whose names begin with _pcre_. These are internal symbols that are not
2787 really supposed to be visible externally, and in some environments it is
2788 possible to suppress them. The cheating is now confined to including
2789 certain files from the library's source, which is a bit cleaner.
2791 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2792 file's purpose clearer.
2794 5. Reorganized pcre_ucp_findchar().
2797 Version 6.3 15-Aug-05
2798 ---------------------
2800 1. The file libpcre.pc.in did not have general read permission in the tarball.
2802 2. There were some problems when building without C++ support:
2804 (a) If C++ support was not built, "make install" and "make test" still
2805 tried to test it.
2807 (b) There were problems when the value of CXX was explicitly set. Some
2808 changes have been made to try to fix these, and ...
2810 (c) --disable-cpp can now be used to explicitly disable C++ support.
2812 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2813 backslash in a target when C++ was disabled. This confuses some
2814 versions of "make", apparently. Using an intermediate variable solves
2815 this. (Same for CPP_LOBJ.)
2817 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2818 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2819 necessary on certain architectures.
2821 4. Added a setting of -export-symbols-regex to the link command to remove
2822 those symbols that are exported in the C sense, but actually are local
2823 within the library, and not documented. Their names all begin with
2824 "_pcre_". This is not a perfect job, because (a) we have to except some
2825 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2826 available (and never for static libraries). I have made a note to try to
2827 find a way round (a) in the future.
2830 Version 6.2 01-Aug-05
2831 ---------------------
2833 1. There was no test for integer overflow of quantifier values. A construction
2834 such as {1111111111111111} would give undefined results. What is worse, if
2835 a minimum quantifier for a parenthesized subpattern overflowed and became
2836 negative, the calculation of the memory size went wrong. This could have
2837 led to memory overwriting.
2839 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2841 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2842 operating environments where this matters.
2844 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2845 PCRE options from within the C++ wrapper.
2847 5. Named capturing subpatterns were not being correctly counted when a pattern
2848 was compiled. This caused two problems: (a) If there were more than 100
2849 such subpatterns, the calculation of the memory needed for the whole
2850 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2851 back references of the form \12, where the number was greater than 9, were
2852 not recognized as back references, even though there were sufficient
2853 previous subpatterns.
2855 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2856 versions of gcc, e.g. 2.95.4.
2859 Version 6.1 21-Jun-05
2860 ---------------------
2862 1. There was one reference to the variable "posix" in pcretest.c that was not
2863 surrounded by "#if !defined NOPOSIX".
2865 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2866 the cross-check on the old pcre_info() function, for the benefit of the
2867 cut-down version of PCRE that is currently imported into Exim.
2869 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2870 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2871 patterns that start (?i) but I don't think that matters. The use of (?i) is
2872 just an example; this all applies to the other options as well.
2874 4. Since libtool seems to echo the compile commands it is issuing, the output
2875 from "make" can be reduced a bit by putting "@" in front of each libtool
2876 compile command.
2878 5. Patch from the folks at Google for configure.in to be a bit more thorough
2879 in checking for a suitable C++ installation before trying to compile the
2880 C++ stuff. This should fix a reported problem when a compiler was present,
2881 but no suitable headers.
2883 6. The man pages all had just "PCRE" as their title. I have changed them to
2884 be the relevant file name. I have also arranged that these names are
2885 retained in the file doc/pcre.txt, which is a concatenation in text format
2886 of all the man pages except the little individual ones for each function.
2888 7. The NON-UNIX-USE file had not been updated for the different set of source
2889 files that come with release 6. I also added a few comments about the C++
2890 wrapper.
2893 Version 6.0 07-Jun-05
2894 ---------------------
2896 1. Some minor internal re-organization to help with my DFA experiments.
2898 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2899 didn't matter for the library itself when fully configured, but did matter
2900 when compiling without UCP support, or within Exim, where the ucp files are
2901 not imported.
2903 3. Refactoring of the library code to split up the various functions into
2904 different source modules. The addition of the new DFA matching code (see
2905 below) to a single monolithic source would have made it really too
2906 unwieldy, quite apart from causing all the code to be include in a
2907 statically linked application, when only some functions are used. This is
2908 relevant even without the DFA addition now that patterns can be compiled in
2909 one application and matched in another.
2911 The downside of splitting up is that there have to be some external
2912 functions and data tables that are used internally in different modules of
2913 the library but which are not part of the API. These have all had their
2914 names changed to start with "_pcre_" so that they are unlikely to clash
2915 with other external names.
2917 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2918 a different (DFA) algorithm. Although it is slower than the original
2919 function, it does have some advantages for certain types of matching
2920 problem.
2922 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2923 including restarting after a partial match.
2925 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2926 defined when compiling for Windows was sent to me. I have put it into the
2927 code, though I have no means of testing or verifying it.
2929 7. Added the pcre_refcount() auxiliary function.
2931 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2932 match before or at the first newline in the subject string. In pcretest,
2933 the /f option on a pattern can be used to set this.
2935 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2936 would behave wrongly. This has been present in PCRE since release 4.0.
2938 10. A number of changes to the pcregrep command:
2940 (a) Refactored how -x works; insert ^(...)$ instead of setting
2941 PCRE_ANCHORED and checking the length, in preparation for adding
2942 something similar for -w.
2944 (b) Added the -w (match as a word) option.
2946 (c) Refactored the way lines are read and buffered so as to have more
2947 than one at a time available.
2949 (d) Implemented a pcregrep test script.
2951 (e) Added the -M (multiline match) option. This allows patterns to match
2952 over several lines of the subject. The buffering ensures that at least
2953 8K, or the rest of the document (whichever is the shorter) is available
2954 for matching (and similarly the previous 8K for lookbehind assertions).
2956 (f) Changed the --help output so that it now says
2958 -w, --word-regex(p)
2960 instead of two lines, one with "regex" and the other with "regexp"
2961 because that confused at least one person since the short forms are the
2962 same. (This required a bit of code, as the output is generated
2963 automatically from a table. It wasn't just a text change.)
2965 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2966 option but starts with a hyphen. Could be a pattern or a path name
2967 starting with a hyphen, for instance.
2969 (h) "-" can be given as a file name to represent stdin.
2971 (i) When file names are being printed, "(standard input)" is used for
2972 the standard input, for compatibility with GNU grep. Previously
2973 "<stdin>" was used.
2975 (j) The option --label=xxx can be used to supply a name to be used for
2976 stdin when file names are being printed. There is no short form.
2978 (k) Re-factored the options decoding logic because we are going to add
2979 two more options that take data. Such options can now be given in four
2980 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2982 (l) Added the -A, -B, and -C options for requesting that lines of context
2983 around matches be printed.
2985 (m) Added the -L option to print the names of files that do not contain
2986 any matching lines, that is, the complement of -l.
2988 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2989 continue to scan other files.
2991 (o) The -s option was incorrectly implemented. For compatibility with other
2992 greps, it now suppresses the error message for a non-existent or non-
2993 accessible file (but not the return code). There is a new option called
2994 -q that suppresses the output of matching lines, which was what -s was
2995 previously doing.
2997 (p) Added --include and --exclude options to specify files for inclusion
2998 and exclusion when recursing.
3000 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3001 Hopefully, it now does.
3003 12. Missing cast in pcre_study().
3005 13. Added an "uninstall" target to the makefile.
3007 14. Replaced "extern" in the function prototypes in Makefile.in with
3008 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3009 world, but is set differently for Windows.
3011 15. Added a second compiling function called pcre_compile2(). The only
3012 difference is that it has an extra argument, which is a pointer to an
3013 integer error code. When there is a compile-time failure, this is set
3014 non-zero, in addition to the error test pointer being set to point to an
3015 error message. The new argument may be NULL if no error number is required
3016 (but then you may as well call pcre_compile(), which is now just a
3017 wrapper). This facility is provided because some applications need a
3018 numeric error indication, but it has also enabled me to tidy up the way
3019 compile-time errors are handled in the POSIX wrapper.
3021 16. Added VPATH=.libs to the makefile; this should help when building with one
3022 prefix path and installing with another. (Or so I'm told by someone who
3023 knows more about this stuff than I do.)
3025 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3026 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3027 match everything, including newlines. This is not POSIX-compatible, but
3028 somebody wanted the feature. From pcretest it can be activated by using
3029 both the P and the s flags.
3031 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3033 19. libpcre.pc was being incorrectly installed as executable.
3035 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3036 it now also looks for '\r' so that it will work unmodified on Windows.
3038 21. Added Google's contributed C++ wrapper to the distribution.
3040 22. Added some untidy missing memory free() calls in pcretest, to keep
3041 Electric Fence happy when testing.
3045 Version 5.0 13-Sep-04
3046 ---------------------
3048 1. Internal change: literal characters are no longer packed up into items
3049 containing multiple characters in a single byte-string. Each character
3050 is now matched using a separate opcode. However, there may be more than one
3051 byte in the character in UTF-8 mode.
3053 2. The pcre_callout_block structure has two new fields: pattern_position and
3054 next_item_length. These contain the offset in the pattern to the next match
3055 item, and its length, respectively.
3057 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3058 insertion of callouts before each pattern item. Added the /C option to
3059 pcretest to make use of this.
3061 4. On the advice of a Windows user, the lines
3063 #if defined(_WIN32) || defined(WIN32)
3064 _setmode( _fileno( stdout ), 0x8000 );
3065 #endif /* defined(_WIN32) || defined(WIN32) */
3067 have been added to the source of pcretest. This apparently does useful
3068 magic in relation to line terminators.
3070 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3071 for the benefit of those environments where the "b" makes a difference.
3073 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3074 to know about it. I have put a hack into configure.in that adds in code
3075 to set GCC=yes if CC=icc. This seems to end up at a point in the
3076 generated configure script that is early enough to affect the setting of
3077 compiler options, which is what is needed, but I have no means of testing
3078 whether it really works. (The user who reported this had patched the
3079 generated configure script, which of course I cannot do.)
3081 LATER: After change 22 below (new libtool files), the configure script
3082 seems to know about icc (and also ecc). Therefore, I have commented out
3083 this hack in configure.in.
3085 7. Added support for pkg-config (2 patches were sent in).
3087 8. Negated POSIX character classes that used a combination of internal tables
3088 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3089 [[:^ascii]]. Typically, they would match almost any characters. The other
3090 POSIX classes were not broken in this way.
3092 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3093 to find the match, as PCRE was deluded into thinking that the match had to
3094 start at the start point or following a newline. The same bug applied to
3095 patterns with negative forward assertions or any backward assertions
3096 preceding ".*" at the start, unless the pattern required a fixed first
3097 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3099 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3100 starting at the last subject character, bytes beyond the end of the subject
3101 string were read.
3103 11. Renamed the variable "class" as "classbits" to make life easier for C++
3104 users. (Previously there was a macro definition, but it apparently wasn't
3105 enough.)
3107 12. Added the new field "tables" to the extra data so that tables can be passed
3108 in at exec time, or the internal tables can be re-selected. This allows
3109 a compiled regex to be saved and re-used at a later time by a different
3110 program that might have everything at different addresses.
3112 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3113 -R library as well as a -L library.
3115 14. The debugging options of pcretest (-d on the command line or D on a
3116 pattern) showed incorrect output for anything following an extended class
3117 that contained multibyte characters and which was followed by a quantifier.
3119 15. Added optional support for general category Unicode character properties
3120 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3121 support. It adds about 90K to the size of the library. The meanings of the
3122 inbuilt class escapes such as \d and \s have NOT been changed.
3124 16. Updated pcredemo.c to include calls to free() to release the memory for the
3125 compiled pattern.
3127 17. The generated file chartables.c was being created in the source directory
3128 instead of in the building directory. This caused the build to fail if the
3129 source directory was different from the building directory, and was
3130 read-only.
3132 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3133 file. No doubt somebody will tell me if they don't make sense... Also added
3134 Dan Mooney's comments about building on OpenVMS.
3136 19. Added support for partial matching via the PCRE_PARTIAL option for
3137 pcre_exec() and the \P data escape in pcretest.
3139 20. Extended pcretest with 3 new pattern features:
3141 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3142 write the compiled pattern to the file whose name is "rest-of-line".
3143 This is a straight binary dump of the data, with the saved pointer to
3144 the character tables forced to be NULL. The study data, if any, is
3145 written too. After writing, pcretest reads a new pattern.
3147 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3148 compiled pattern from the given file. There must not be any
3149 occurrences of "<" in the file name (pretty unlikely); if there are,
3150 pcretest will instead treat the initial "<" as a pattern delimiter.
3151 After reading in the pattern, pcretest goes on to read data lines as
3152 usual.
3154 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3155 and 16-bit fields in a compiled pattern, to simulate a pattern that
3156 was compiled on a host of opposite endianness.
3158 21. The pcre-exec() function can now cope with patterns that were compiled on
3159 hosts of opposite endianness, with this restriction:
3161 As for any compiled expression that is saved and used later, the tables
3162 pointer field cannot be preserved; the extra_data field in the arguments
3163 to pcre_exec() should be used to pass in a tables address if a value
3164 other than the default internal tables were used at compile time.
3166 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3167 now diagnosed as an error. Previously, most of the time, a negative number
3168 would have been treated as zero, but if in addition "ovector" was passed as
3169 NULL, a crash could occur.
3171 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3172 new versions from the libtool 1.5 distribution (the last one is a copy of
3173 a file called libtool.m4). This seems to have fixed the need to patch
3174 "configure" to support Darwin 1.3 (which I used to do). However, I still
3175 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3176 workstation).
3178 24. Changed the PCRE licence to be the more standard "BSD" licence.
3181 Version 4.5 01-Dec-03
3182 ---------------------
3184 1. There has been some re-arrangement of the code for the match() function so
3185 that it can be compiled in a version that does not call itself recursively.
3186 Instead, it keeps those local variables that need separate instances for
3187 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3188 needs to "recurse". Keeping track of where control must go is done by means
3189 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3190 hide most of the details from the main code, and operates only if
3191 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3192 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3193 operating.
3195 To make it easier for callers to provide specially tailored get/free
3196 functions for this usage, two new functions, pcre_stack_malloc, and
3197 pcre_stack_free, are used. They are always called in strict stacking order,
3198 and the size of block requested is always the same.
3200 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3201 PCRE has been compiled to use the stack or the heap for recursion. The
3202 -C option of pcretest uses this to show which version is compiled.
3204 A new data escape \S, is added to pcretest; it causes the amounts of store
3205 obtained and freed by both kinds of malloc/free at match time to be added
3206 to the output.
3208 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3209 what's available on my current Linux desktop machine.
3211 3. When matching a UTF-8 string, the test for a valid string at the start has
3212 been extended. If start_offset is not zero, PCRE now checks that it points
3213 to a byte that is the start of a UTF-8 character. If not, it returns
3214 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3215 this is necessary because there may be backward assertions in the pattern.
3216 When matching the same subject several times, it may save resources to use
3217 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3219 4. The code for checking the validity of UTF-8 strings has been tightened so
3220 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3221 containing "overlong sequences".
3223 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3224 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3225 should have been "&", but it just so happened that all the cases this let
3226 through by mistake were picked up later in the function.
3228 6. I had used a variable called "isblank" - this is a C99 function, causing
3229 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3231 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3232 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3233 so the version is shown; (c) stop "make test" echoing "./RunTest".
3235 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3237 9. The prototype for memmove() for systems that don't have it was using
3238 size_t, but the inclusion of the header that defines size_t was later. I've
3239 moved the #includes for the C headers earlier to avoid this.
3241 10. Added some adjustments to the code to make it easier to compiler on certain
3242 special systems:
3244 (a) Some "const" qualifiers were missing.
3245 (b) Added the macro EXPORT before all exported functions; by default this
3246 is defined to be empty.
3247 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3248 that it reads its output file name as an argument instead of writing
3249 to the standard output and assuming this can be redirected.
3251 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3252 class containing characters with values greater than 255, PCRE compilation
3253 went into a loop.
3255 12. A recursive reference to a subpattern that was within another subpattern
3256 that had a minimum quantifier of zero caused PCRE to crash. For example,
3257 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3258 recursion. If the recursively-called subpattern itself had a zero repeat,
3259 that was OK.
3261 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3262 buffer into which it was copied (for escape processing) was still set at
3263 1024, so long lines caused crashes.
3265 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3266 "internal error: code overflow...". This applied to any character class
3267 that was followed by a possessive quantifier.
3269 15. Modified the Makefile to add libpcre.la as a prerequisite for
3270 libpcreposix.la because I was told this is needed for a parallel build to
3271 work.
3273 16. If a pattern that contained .* following optional items at the start was
3274 studied, the wrong optimizing data was generated, leading to matching
3275 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3276 matching string must start with a or b or c. The correct conclusion for
3277 this pattern is that a match can start with any character.
3280 Version 4.4 13-Aug-03
3281 ---------------------
3283 1. In UTF-8 mode, a character class containing characters with values between
3284 127 and 255 was not handled correctly if the compiled pattern was studied.
3285 In fixing this, I have also improved the studying algorithm for such
3286 classes (slightly).
3288 2. Three internal functions had redundant arguments passed to them. Removal
3289 might give a very teeny performance improvement.
3291 3. Documentation bug: the value of the capture_top field in a callout is *one
3292 more than* the number of the hightest numbered captured substring.
3294 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3295 in incorrectly linking with a previously installed version. They now link
3296 explicitly with libpcre.la.
3298 5. configure.in no longer needs to recognize Cygwin specially.
3300 6. A problem in pcre.in for Windows platforms is fixed.
3302 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3303 pcretest, it used to include the size of the study block as part of its
3304 output. Unfortunately, the structure contains a field that has a different
3305 size on different hardware architectures. This meant that the tests that
3306 showed this size failed. As the block is currently always of a fixed size,
3307 this information isn't actually particularly useful in pcretest output, so
3308 I have just removed it.
3310 8. Three pre-processor statements accidentally did not start in column 1.
3311 Sadly, there are *still* compilers around that complain, even though
3312 standard C has not required this for well over a decade. Sigh.
3314 9. In pcretest, the code for checking callouts passed small integers in the
3315 callout_data field, which is a void * field. However, some picky compilers
3316 complained about the casts involved for this on 64-bit systems. Now
3317 pcretest passes the address of the small integer instead, which should get
3318 rid of the warnings.
3320 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3321 both compile and run time, and gives an error if an invalid UTF-8 sequence
3322 is found. There is a option for disabling this check in cases where the
3323 string is known to be correct and/or the maximum performance is wanted.
3325 11. In response to a bug report, I changed one line in Makefile.in from
3327 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3328 to
3329 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3331 to look similar to other lines, but I have no way of telling whether this
3332 is the right thing to do, as I do not use Windows. No doubt I'll get told
3333 if it's wrong...
3336 Version 4.3 21-May-03
3337 ---------------------
3339 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3340 Makefile.
3342 2. Some refactoring to improve the quality of the code:
3344 (i) The utf8_table... variables are now declared "const".
3346 (ii) The code for \cx, which used the "case flipping" table to upper case
3347 lower case letters, now just substracts 32. This is ASCII-specific,
3348 but the whole concept of \cx is ASCII-specific, so it seems
3349 reasonable.
3351 (iii) PCRE was using its character types table to recognize decimal and
3352 hexadecimal digits in the pattern. This is silly, because it handles
3353 only 0-9, a-f, and A-F, but the character types table is locale-
3354 specific, which means strange things might happen. A private
3355 table is now used for this - though it costs 256 bytes, a table is
3356 much faster than multiple explicit tests. Of course, the standard
3357 character types table is still used for matching digits in subject
3358 strings against \d.
3360 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3361 ending in _t are). So I've renamed it as ESC_tee.
3363 3. The first argument for regexec() in the POSIX wrapper should have been
3364 defined as "const".
3366 4. Changed pcretest to use malloc() for its buffers so that they can be
3367 Electric Fenced for debugging.
3369 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3370 to read one or more bytes before the start of the subject string. Often this
3371 had no effect on PCRE's behaviour, but in some circumstances it could
3372 provoke a segmentation fault.
3374 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3375 to try to read one or more bytes before the start of the subject string.
3377 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3378 UTF-8 support could misbehave in various ways if the subject string
3379 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3380 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3381 back over UTF-8 characters.)
3384 Version 4.2 14-Apr-03
3385 ---------------------
3387 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3389 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3390 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3391 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3392 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3393 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3395 Note: automatic setting of the BUILD variables is not yet working
3396 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3397 compile-time but not at link-time
3398 [LINK]: use for linking executables only
3399 make different versions for Windows and non-Windows
3400 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3401 libraries
3402 [LINK_FOR_BUILD]: new variable
3403 [OBJEXT]: use throughout
3404 [EXEEXT]: use throughout
3405 <winshared>: new target
3406 <wininstall>: new target
3407 <dftables.o>: use native compiler
3408 <dftables>: use native linker
3409 <install>: handle Windows platform correctly
3410 <clean>: ditto
3411 <check>: ditto
3412 copy DLL to top builddir before testing
3414 As part of these changes, -no-undefined was removed again. This was reported
3415 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3416 in any case.
3418 3. Some tidies to get rid of compiler warnings:
3420 . In the match_data structure, match_limit was an unsigned long int, whereas
3421 match_call_count was an int. I've made them both unsigned long ints.
3423 . In pcretest the fact that a const uschar * doesn't automatically cast to
3424 a void * provoked a warning.
3426 . Turning on some more compiler warnings threw up some "shadow" variables
3427 and a few more missing casts.
3429 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3430 option, a class that contained a single character with a value between 128
3431 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3433 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3434 option, a class that contained several characters, but with at least one
3435 whose value was between 128 and 255 caused PCRE to crash.
3438 Version 4.1 12-Mar-03
3439 ---------------------
3441 1. Compiling with gcc -pedantic found a couple of places where casts were
3442 needed, and a string in dftables.c that was longer than standard compilers are
3443 required to support.
3445 2. Compiling with Sun's compiler found a few more places where the code could
3446 be tidied up in order to avoid warnings.
3448 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3449 first of these names is deprecated in the latest Autoconf in favour of the name
3450 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3451 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3452 analogy I have changed it to CFLAGS_FOR_BUILD.
3454 4. Added -no-undefined to the linking command in the Makefile, because this is
3455 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3456 linking step for the pcreposix library.
3458 5. PCRE was failing to diagnose the case of two named groups with the same
3459 name.
3461 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3462 literal character that is needed in the subject for a match, and scans along to
3463 ensure that it is present before embarking on the full matching process. This
3464 saves time in cases of nested unlimited repeats that are never going to match.
3465 Problem: the scan can take a lot of time if the subject is very long (e.g.
3466 megabytes), thus penalizing straightforward matches. It is now done only if the
3467 amount of subject to be scanned is less than 1000 bytes.
3469 7. A lesser problem with the same optimization is that it was recording the
3470 first character of an anchored pattern as "needed", thus provoking a search
3471 right along the subject, even when the first match of the pattern was going to
3472 fail. The "needed" character is now not set for anchored patterns, unless it
3473 follows something in the pattern that is of non-fixed length. Thus, it still
3474 fulfils its original purpose of finding quick non-matches in cases of nested
3475 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3478 Version 4.0 17-Feb-03
3479 ---------------------
3481 1. If a comment in an extended regex that started immediately after a meta-item
3482 extended to the end of string, PCRE compiled incorrect data. This could lead to
3483 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3485 2. Moved to autoconf 2.53 and libtool 1.4.2.
3487 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3488 the special perltest8 script is no longer needed - all the tests can be run
3489 from a single perltest script.
3491 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3492 by \s. It has now been removed in PCRE. This means it isn't recognized as
3493 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3494 class [:space:] *does* include VT, thereby creating a mess.
3496 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3497 space and tab.
3499 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3500 its new features into the main test script, reducing the number of scripts.
3502 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3503 were backward compatible, and made the (?i) apply to the whole pattern, as if
3504 /i were given. Now it behaves more logically, and applies the option setting
3505 only to what follows. PCRE has been changed to follow suit. However, if it
3506 finds options settings right at the start of the pattern, it extracts them into
3507 the global options, as before. Thus, they show up in the info data.
3509 8. Added support for the \Q...\E escape sequence. Characters in between are
3510 treated as literals. This is slightly different from Perl in that $ and @ are
3511 also handled as literals inside the quotes. In Perl, they will cause variable
3512 interpolation. Note the following examples:
3514 Pattern PCRE matches Perl matches
3516 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3517 \Qabc\$xyz\E abc\$xyz abc\$xyz
3518 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3520 For compatibility with Perl, \Q...\E sequences are recognized inside character
3521 classes as well as outside them.
3523 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3524 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3525 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3526 signed/unsigned warnings.
3528 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3529 option for pcretest, so I've replaced it by a simple function that does just
3530 that job.
3532 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3533 "pcregrep -".
3535 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3536 Java package. This provides some syntactic sugar for simple cases of what my
3537 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3538 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3539 item, you can use this simplified notation. Note that only makes sense with
3540 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3541 greediness, whatever the setting of the PCRE_UNGREEDY option.
3543 13. A change of greediness default within a pattern was not taking effect at
3544 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3545 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3546 was abstracted outside.
3548 14. PCRE now supports the \G assertion. It is true when the current matching
3549 position is at the start point of the match. This differs from \A when the
3550 starting offset is non-zero. Used with the /g option of pcretest (or similar
3551 code), it works in the same way as it does for Perl's /g option. If all
3552 alternatives of a regex begin with \G, the expression is anchored to the start
3553 match position, and the "anchored" flag is set in the compiled expression.
3555 15. Some bugs concerning the handling of certain option changes within patterns
3556 have been fixed. These applied to options other than (?ims). For example,
3557 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3558 been the other way round. Some of this was related to change 7 above.
3560 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3561 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3562 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3563 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3565 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3566 mode. Unlike ".", it always matches newline, whatever the setting of
3567 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3568 assertions. Perl allows it, but it doesn't (in general) work because it can't
3569 calculate the length of the lookbehind. At least, that's the case for Perl
3570 5.8.0 - I've been told they are going to document that it doesn't work in
3571 future.
3573 18. Added an error diagnosis for escapes that PCRE does not support: these are
3574 \L, \l, \N, \P, \p, \U, \u, and \X.
3576 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3577 reading past the end of the pattern in cases such as /[abcd/.
3579 20. PCRE was getting more memory than necessary for patterns with classes that
3580 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3582 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3583 compiling PCRE for use with Virtual Pascal.
3585 22. Small fix to the Makefile to make it work properly if the build is done
3586 outside the source tree.
3588 23. Added a new extension: a condition to go with recursion. If a conditional
3589 subpattern starts with (?(R) the "true" branch is used if recursion has
3590 happened, whereas the "false" branch is used only at the top level.
3592 24. When there was a very long string of literal characters (over 255 bytes
3593 without UTF support, over 250 bytes with UTF support), the computation of how
3594 much memory was required could be incorrect, leading to segfaults or other
3595 strange effects.
3597 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3598 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3599 there was a subsequent back reference to those brackets. This meant that, for
3600 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3601 possible to check for precisely this case. All we can do is abandon the
3602 optimization if .* occurs inside capturing brackets when there are any back
3603 references whatsoever. (See below for a better fix that came later.)
3605 26. The handling of the optimization for finding the first character of a
3606 non-anchored pattern, and for finding a character that is required later in the
3607 match were failing in some cases. This didn't break the matching; it just
3608 failed to optimize when it could. The way this is done has been re-implemented.
3610 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3612 28. Added a new feature that provides some of the functionality that Perl
3613 provides with (?{...}). The facility is termed a "callout". The way it is done
3614 in PCRE is for the caller to provide an optional function, by setting
3615 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3616 global variable. By default it is unset, which disables all calling out. To get
3617 the function called, the regex must include (?C) at appropriate points. This
3618 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3619 This provides a means of identifying different callout points. When PCRE
3620 reaches such a point in the regex, if pcre_callout has been set, the external
3621 function is called. It is provided with data in a structure called
3622 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3623 matching continues; if it returns a non-zero value, the match at the current
3624 point fails. However, backtracking will occur if possible. [This was changed
3625 later and other features added - see item 49 below.]
3627 29. pcretest is upgraded to test the callout functionality. It provides a
3628 callout function that displays information. By default, it shows the start of
3629 the match and the current position in the text. There are some new data escapes
3630 to vary what happens:
3632 \C+ in addition, show current contents of captured substrings
3633 \C- do not supply a callout function
3634 \C!n return 1 when callout number n is reached
3635 \C!n!m return 1 when callout number n is reached for the mth time
3637 30. If pcregrep was called with the -l option and just a single file name, it
3638 output "<stdin>" if a match was found, instead of the file name.
3640 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3641 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3642 pcre_exec(). This saves a malloc/free per call. The default value of
3643 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3644 when configuring.
3646 32. The default maximum size of a compiled pattern is 64K. There have been a
3647 few cases of people hitting this limit. The code now uses macros to handle the
3648 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3649 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3650 configuring. Tests 2 and 5 work only with 2-byte links because they output
3651 debugging information about compiled patterns.
3653 33. Internal code re-arrangements:
3655 (a) Moved the debugging function for printing out a compiled regex into
3656 its own source file (printint.c) and used #include to pull it into
3657 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3658 separate copies.
3660 (b) Defined the list of op-code names for debugging as a macro in
3661 internal.h so that it is next to the definition of the opcodes.
3663 (c) Defined a table of op-code lengths for simpler skipping along compiled
3664 code. This is again a macro in internal.h so that it is next to the
3665 definition of the opcodes.
3667 34. Added support for recursive calls to individual subpatterns, along the
3668 lines of Robin Houston's patch (but implemented somewhat differently).
3670 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3671 allow it to read and process whole directories in Win32. This code was
3672 contributed by Lionel Fourquaux; it has not been tested by me.
3674 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3675 used to name a group. Names consist of alphanumerics and underscores, and must
3676 be unique. Back references use the syntax (?P=name) and recursive calls use
3677 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3678 numbers. The function pcre_fullinfo() can be used after compilation to extract
3679 a name/number map. There are three relevant calls:
3681 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3682 PCRE_INFO_NAMECOUNT yields the number of entries
3683 PCRE_INFO_NAMETABLE yields a pointer to the map.
3685 The map is a vector of fixed-size entries. The size of each entry depends on
3686 the length of the longest name used. The first two bytes of each entry are the
3687 group number, most significant byte first. There follows the corresponding
3688 name, zero terminated. The names are in alphabetical order.
3690 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3691 case instead of 255. Making it the same both with and without UTF-8 support
3692 means that the same test output works with both.
3694 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3695 calling malloc() with a zero argument.
3697 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3698 optimization. I've improved things by keeping a bitmap of backreferences with
3699 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3700 fact referenced, the optimization can be applied. It is unlikely that a
3701 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3702 the match to follow \n) will appear inside brackets with a number greater than
3703 31, but if it does, any back reference > 31 suppresses the optimization.
3705 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3706 of disabling numbered capturing parentheses. Any opening parenthesis that is
3707 not followed by ? behaves as if it were followed by ?: but named parentheses
3708 can still be used for capturing (and they will acquire numbers in the usual
3709 way).
3711 41. Redesigned the return codes from the match() function into yes/no/error so
3712 that errors can be passed back from deep inside the nested calls. A malloc
3713 failure while inside a recursive subpattern call now causes the
3714 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3716 42. It is now possible to set a limit on the number of times the match()
3717 function is called in a call to pcre_exec(). This facility makes it possible to
3718 limit the amount of recursion and backtracking, though not in a directly
3719 obvious way, because the match() function is used in a number of different
3720 circumstances. The count starts from zero for each position in the subject
3721 string (for non-anchored patterns). The default limit is, for compatibility, a
3722 large number, namely 10 000 000. You can change this in two ways:
3724 (a) When configuring PCRE before making, you can use --with-match-limit=n
3725 to set a default value for the compiled library.
3727 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3728 a different value is set. See 45 below.
3730 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3732 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3733 of things that can be changed at compile time. The first argument specifies
3734 what is wanted and the second points to where the information is to be placed.
3735 The current list of available information is:
3739 The output is an integer that is set to one if UTF-8 support is available;
3740 otherwise it is set to zero.
3744 The output is an integer that it set to the value of the code that is used for
3745 newline. It is either LF (10) or CR (13).
3749 The output is an integer that contains the number of bytes used for internal
3750 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3754 The output is an integer that contains the threshold above which the POSIX
3755 interface uses malloc() for output vectors. See item 31 above.
3759 The output is an unsigned integer that contains the default limit of the number
3760 of match() calls in a pcre_exec() execution. See 42 above.
3762 44. pcretest has been upgraded by the addition of the -C option. This causes it
3763 to extract all the available output from the new pcre_config() function, and to
3764 output it. The program then exits immediately.
3766 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3767 order to support additional features. One way would have been to define
3768 pcre_exec2() (for example) with extra arguments, but this would not have been
3769 extensible, and would also have required all calls to the original function to
3770 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3771 is used for passing in "extra" data from pcre_study().
3773 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3774 contains the following fields:
3776 flags a bitmap indicating which of the following fields are set
3777 study_data opaque data from pcre_study()
3778 match_limit a way of specifying a limit on match() calls for a specific
3779 call to pcre_exec()
3780 callout_data data for callouts (see 49 below)
3782 The flag bits are also defined in pcre.h, and are
3788 The pcre_study() function now returns one of these new pcre_extra blocks, with
3789 the actual study data pointed to by the study_data field, and the
3790 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3791 before. That is, this change is entirely upwards-compatible and requires no
3792 change to existing code.
3794 If you want to pass in additional data to pcre_exec(), you can either place it
3795 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3796 block.
3798 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3799 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3800 times with different match limits, until it finds the minimum value needed for
3801 pcre_exec() to complete. The value is then output. This can be instructive; for
3802 most simple matches the number is quite small, but for pathological cases it
3803 gets very large very quickly.
3805 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3806 returns the size of the data block pointed to by the study_data field in a
3807 pcre_extra block, that is, the value that was passed as the argument to
3808 pcre_malloc() when PCRE was getting memory in which to place the information
3809 created by pcre_study(). The fourth argument should point to a size_t variable.
3810 pcretest has been extended so that this information is shown after a successful
3811 pcre_study() call when information about the compiled regex is being displayed.
3813 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3814 because what follows is always an absolute path. (Later: it turns out that this
3815 is more than cosmetic for MinGW, because it doesn't like empty path
3816 components.)
3818 49. Some changes have been made to the callout feature (see 28 above):
3820 (i) A callout function now has three choices for what it returns:
3822 0 => success, carry on matching
3823 > 0 => failure at this point, but backtrack if possible
3824 < 0 => serious error, return this value from pcre_exec()
3826 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3827 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3828 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3829 use by callout functions. It will never be used by PCRE itself.
3831 (ii) The pcre_extra structure (see 45 above) has a void * field called
3832 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3833 pcre_callout_block structure has a field of the same name. The contents of
3834 the field passed in the pcre_extra structure are passed to the callout
3835 function in the corresponding field in the callout block. This makes it
3836 easier to use the same callout-containing regex from multiple threads. For
3837 testing, the pcretest program has a new data escape
3839 \C*n pass the number n (may be negative) as callout_data
3841 If the callout function in pcretest receives a non-zero value as
3842 callout_data, it returns that value.
3844 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3845 there were some redundant $(CFLAGS) in commands that are now specified as
3846 $(LINK), which already includes $(CFLAGS).
3848 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3849 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3850 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3851 one-byte characters throughout. Note that case-insensitive matching applies
3852 only to characters whose values are less than 256. PCRE doesn't support the
3853 notion of cases for higher-valued characters.
3855 (i) A character class whose characters are all within 0-255 is handled as
3856 a bit map, and the map is inverted for negative classes. Previously, a
3857 character > 255 always failed to match such a class; however it should
3858 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3860 (ii) A negated character class with a single character < 255 is coded as
3861 "not this character" (OP_NOT). This wasn't working properly when the test
3862 character was multibyte, either singly or repeated.
3864 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3865 mode, for example: \x{100}{2,3}.
3867 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3868 singly or repeated) now correctly test multibyte characters. However,
3869 PCRE doesn't recognize any characters with values greater than 255 as
3870 digits, spaces, or word characters. Such characters always match \D, \S,
3871 and \W, and never match \d, \s, or \w.
3873 (v) Classes may now contain characters and character ranges with values
3874 greater than 255. For example: [ab\x{100}-\x{400}].
3876 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3877 PCRE in UTF-8 mode.
3879 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3880 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3881 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3882 value.)
3884 53. The single man page has become too large. I have therefore split it up into
3885 a number of separate man pages. These also give rise to individual HTML pages;
3886 these are now put in a separate directory, and there is an index.html page that
3887 lists them all. Some hyperlinking between the pages has been installed.
3889 54. Added convenience functions for handling named capturing parentheses.
3891 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3892 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3893 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3894 are faulted.
3896 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3897 calling configure. These values are used when compiling the dftables.c program
3898 which is run to generate the source of the default character tables. They
3899 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3900 you will need to set these values.
3902 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3905 Version 3.9 02-Jan-02
3906 ---------------------
3908 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3910 2. If --disable-static was given, the building process failed when trying to
3911 build pcretest and pcregrep. (For some reason it was using libtool to compile
3912 them, which is not right, as they aren't part of the library.)
3915 Version 3.8 18-Dec-01
3916 ---------------------
3918 1. The experimental UTF-8 code was completely screwed up. It was packing the
3919 bytes in the wrong order. How dumb can you get?
3922 Version 3.7 29-Oct-01
3923 ---------------------
3925 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3926 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3927 this didn't happen under Solaris 8, where I normally test things.
3929 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3930 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3931 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3932 this fix an hour or so after the initial 3.7 release.)
3935 Version 3.6 23-Oct-01
3936 ---------------------
3938 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3939 offsets passed as NULL with zero offset count.
3941 2. The config.guess and config.sub files had not been updated when I moved to
3942 the latest autoconf.
3945 Version 3.5 15-Aug-01
3946 ---------------------
3948 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3949 had been forgotten.
3951 2. By using declared but undefined structures, we can avoid using "void"
3952 definitions in pcre.h while keeping the internal definitions of the structures
3953 private.
3955 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3956 user point of view, this means that both static and shared libraries are built
3957 by default, but this can be individually controlled. More of the work of
3958 handling this static/shared cases is now inside libtool instead of PCRE's make
3959 file.
3961 4. The pcretest utility is now installed along with pcregrep because it is
3962 useful for users (to test regexs) and by doing this, it automatically gets
3963 relinked by libtool. The documentation has been turned into a man page, so
3964 there are now .1, .txt, and .html versions in /doc.
3966 5. Upgrades to pcregrep:
3967 (i) Added long-form option names like gnu grep.
3968 (ii) Added --help to list all options with an explanatory phrase.
3969 (iii) Added -r, --recursive to recurse into sub-directories.
3970 (iv) Added -f, --file to read patterns from a file.
3972 6. pcre_exec() was referring to its "code" argument before testing that
3973 argument for NULL (and giving an error if it was NULL).
3975 7. Upgraded Makefile.in to allow for compiling in a different directory from
3976 the source directory.
3978 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3979 options bits, the pointer it was passed was to an int instead of to an unsigned
3980 long int. This mattered only on 64-bit systems.
3982 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3983 generated) instead of pcre.in, which it its source. Also made the same change
3984 in several of the .c files.
3986 10. A new release of gcc defines printf() as a macro, which broke pcretest
3987 because it had an ifdef in the middle of a string argument for printf(). Fixed
3988 by using separate calls to printf().
3990 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3991 script, to force use of CR or LF instead of \n in the source. On non-Unix
3992 systems, the value can be set in config.h.
3994 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3995 absolute limit. Changed the text of the error message to make this clear, and
3996 likewise updated the man page.
3998 13. The limit of 99 on the number of capturing subpatterns has been removed.
3999 The new limit is 65535, which I hope will not be a "real" limit.
4002 Version 3.4 22-Aug-00
4003 ---------------------
4005 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4007 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4010 Version 3.3 01-Aug-00
4011 ---------------------
4013 1. If an octal character was given, but the value was greater than \377, it
4014 was not getting masked to the least significant bits, as documented. This could
4015 lead to crashes in some systems.
4017 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4018 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4020 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4021 These just pass their arguments on to (pcre_free)(), but they are provided
4022 because some uses of PCRE bind it to non-C systems that can call its functions,
4023 but cannot call free() or pcre_free() directly.
4025 4. Add "make test" as a synonym for "make check". Corrected some comments in
4026 the Makefile.
4028 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4029 Makefile.
4031 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4032 command called pgrep for grepping around the active processes.
4034 7. Added the beginnings of support for UTF-8 character strings.
4036 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4037 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4038 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4039 out for the ar command.)
4042 Version 3.2 12-May-00
4043 ---------------------
4045 This is purely a bug fixing release.
4047 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4048 of ZA. This was just one example of several cases that could provoke this bug,
4049 which was introduced by change 9 of version 2.00. The code for breaking
4050 infinite loops after an iteration that matches an empty string was't working
4051 correctly.
4053 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4054 when matched against abbab (for example). After matching an empty string, it
4055 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4056 caused it to match further down the string than it should.
4058 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4059 was there because it doesn't seem to be needed, and it causes trouble on some
4060 systems, as it is not a Standard C header. It has been removed.
4062 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4063 were reported on the Macintosh. The changes were from
4065 while ((c = *(++ptr)) != 0 && c != '\n');
4066 to
4067 while ((c = *(++ptr)) != 0 && c != '\n') ;
4069 Totally extraordinary, but if that's what it takes...
4071 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4072 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4073 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4074 assumes the way PCRE uses memmove() (always moving upwards).
4076 6. PCRE is being used in one environment where strchr() is not available. There
4077 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4078 faster code anyway.
4081 Version 3.1 09-Feb-00
4082 ---------------------
4084 The only change in this release is the fixing of some bugs in Makefile.in for
4085 the "install" target:
4087 (1) It was failing to install pcreposix.h.
4089 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4092 Version 3.0 01-Feb-00
4093 ---------------------
4095 1. Add support for the /+ modifier to perltest (to output $` like it does in
4096 pcretest).
4098 2. Add support for the /g modifier to perltest.
4100 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4101 matches null strings.
4103 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4104 pattern. Perl treats empty patterns specially - it reuses the most recent
4105 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4106 effect.
4108 5. The POSIX interface was broken in that it was just handing over the POSIX
4109 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4110 required a bigger vector, with some working space on the end. This means that
4111 the POSIX wrapper now has to get and free some memory, and copy the results.
4113 6. Added some simple autoconf support, placing the test data and the
4114 documentation in separate directories, re-organizing some of the
4115 information files, and making it build pcre-config (a GNU standard). Also added
4116 libtool support for building PCRE as a shared library, which is now the
4117 default.
4119 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4120 09 are not valid octal constants. Single digits will be used for minor values
4121 less than 10.
4123 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4124 existing programs that set these in the POSIX interface can use PCRE without
4125 modification.
4127 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4128 return all that pcre_info() returns, plus additional data. The pcre_info()
4129 function is retained for compatibility, but is considered to be obsolete.
4131 10. Added experimental recursion feature (?R) to handle one common case that
4132 Perl 5.6 will be able to do with (?p{...}).
4134 11. Added support for POSIX character classes like [:alpha:], which Perl is
4135 adopting.
4138 Version 2.08 31-Aug-99
4139 ----------------------
4141 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4142 trying to match at the startoffset position, but instead was moving forward to
4143 the next newline as if a previous match had failed.
4145 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4146 and could get into a loop if a null string was matched other than at the start
4147 of the subject.
4149 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4150 be distinguished at compile time, and for completeness also added PCRE_DATE.
4152 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4153 in GnuWin32 environments.
4156 Version 2.07 29-Jul-99
4157 ----------------------
4159 1. The documentation is now supplied in plain text form and HTML as well as in
4160 the form of man page sources.
4162 2. C++ compilers don't like assigning (void *) values to other pointer types.
4163 In particular this affects malloc(). Although there is no problem in Standard
4164 C, I've put in casts to keep C++ compilers happy.
4166 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4167 should be (const char *).
4169 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4170 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4171 However, I haven't made this a standard facility. The documentation doesn't
4172 mention it, and the Makefile doesn't support it.
4174 5. The Makefile now contains an "install" target, with editable destinations at
4175 the top of the file. The pcretest program is not installed.
4177 6. pgrep -V now gives the PCRE version number and date.
4179 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4180 causing the entire string to be ignored, instead of just the last character.
4182 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4183 non-matching string, it can take a very, very long time, even for strings of
4184 quite modest length, because of the nested recursion. PCRE now does better in
4185 some of these cases. It does this by remembering the last required literal
4186 character in the pattern, and pre-searching the subject to ensure it is present
4187 before running the real match. In other words, it applies a heuristic to detect
4188 some types of certain failure quickly, and in the above example, if presented
4189 with a string that has no trailing " it gives "no match" very quickly.
4191 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4192 other alternatives are tried instead.
4195 Version 2.06 09-Jun-99
4196 ----------------------
4198 1. Change pcretest's output for amount of store used to show just the code
4199 space, because the remainder (the data block) varies in size between 32-bit and
4200 64-bit systems.
4202 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4203 start matching at. This allows lookbehinds to work when searching for multiple
4204 occurrences in a string.
4206 3. Added additional options to pcretest for testing multiple occurrences:
4208 /+ outputs the rest of the string that follows a match
4209 /g loops for multiple occurrences, using the new startoffset argument
4210 /G loops for multiple occurrences by passing an incremented pointer
4212 4. PCRE wasn't doing the "first character" optimization for patterns starting
4213 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4214 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4215 the letter 'x'. On long subject strings, this gives a significant speed-up.
4218 Version 2.05 21-Apr-99
4219 ----------------------
4221 1. Changed the type of magic_number from int to long int so that it works
4222 properly on 16-bit systems.
4224 2. Fixed a bug which caused patterns starting with .* not to work correctly
4225 when the subject string contained newline characters. PCRE was assuming
4226 anchoring for such patterns in all cases, which is not correct because .* will
4227 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4228 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4229 must be retried after every newline in the subject.
4232 Version 2.04 18-Feb-99
4233 ----------------------
4235 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4236 computation of the store needed to hold the pattern was incorrect (too large).
4237 If such patterns were nested a few deep, this could multiply and become a real
4238 problem.
4240 2. Added /M option to pcretest to show the memory requirement of a specific
4241 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4243 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4244 compiled in such a way that the backtracking after subsequent failure was
4245 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4246 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4249 Version 2.03 02-Feb-99
4250 ----------------------
4252 1. Fixed typo and small mistake in man page.
4254 2. Added 4th condition (GPL supersedes if conflict) and created separate
4255 LICENCE file containing the conditions.
4257 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4258 Perl, that is the internal \ allows the delimiter to be included in the
4259 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4260 the final delimiter, add \ to the end of the pattern (to test the error).
4262 4. Added the convenience functions for extracting substrings after a successful
4263 match. Updated pcretest to make it able to test these functions.
4266 Version 2.02 14-Jan-99
4267 ----------------------
4269 1. Initialized the working variables associated with each extraction so that
4270 their saving and restoring doesn't refer to uninitialized store.
4272 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4273 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4274 fix the problem.
4276 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4277 calls, and wasn't printing the correct value for compiling calls. Increased the
4278 default value of LOOPREPEAT, and the number of significant figures in the
4279 times.
4281 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4283 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4284 a building problem on Windows NT with a FAT file system.
4287 Version 2.01 21-Oct-98
4288 ----------------------
4290 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4291 to character tables built by pcre_maketables() in the current locale. If NULL
4292 is passed, the default tables are used.
4295 Version 2.00 24-Sep-98
4296 ----------------------
4298 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4299 it any more.
4301 2. Allow quantification of (?>) groups, and make it work correctly.
4303 3. The first character computation wasn't working for (?>) groups.
4305 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4306 end of the subject) and add 5.005's \z, which really does match only at the
4307 very end of the subject.
4309 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4311 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4312 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4313 localized options. All options to pcre_study() were also removed.
4315 7. Add other new features from 5.005:
4317 $(?<= positive lookbehind
4318 $(?<! negative lookbehind
4319 (?imsx-imsx) added the unsetting capability
4320 such a setting is global if at outer level; local otherwise
4321 (?imsx-imsx:) non-capturing groups with option setting
4322 (?(cond)re|re) conditional pattern matching
4324 A backreference to itself in a repeated group matches the previous
4325 captured string.
4327 8. General tidying up of studying (both automatic and via "study")
4328 consequential on the addition of new assertions.
4330 9. As in 5.005, unlimited repeated groups that could match an empty substring
4331 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4332 runtime if any iteration does actually match an empty substring.
4334 10. Include the RunTest script in the distribution.
4336 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4337 discrepancies, some of which were old and were also with respect to 5.004. They
4338 have now been fixed.
4341 Version 1.09 28-Apr-98
4342 ----------------------
4344 1. A negated single character class followed by a quantifier with a minimum
4345 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4346 program crashes, or just wrong answers. This did not apply to negated classes
4347 containing more than one character, or to minima other than one.
4350 Version 1.08 27-Mar-98
4351 ----------------------
4353 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4355 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4356 latter must appear before anything that relies on it in the pattern.
4359 Version 1.07 16-Feb-98
4360 ----------------------
4362 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4363 repeat of a potentially empty string).
4366 Version 1.06 23-Jan-98
4367 ----------------------
4369 1. Added Markus Oberhumer's little patches for C++.
4371 2. Literal strings longer than 255 characters were broken.
4374 Version 1.05 23-Dec-97
4375 ----------------------
4377 1. Negated character classes containing more than one character were failing if
4378 PCRE_CASELESS was set at run time.
4381 Version 1.04 19-Dec-97
4382 ----------------------
4384 1. Corrected the man page, where some "const" qualifiers had been omitted.
4386 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4387 input syntax.
4389 3. Fixed memory leak which occurred when a regex with back references was
4390 matched with an offsets vector that wasn't big enough. The temporary memory
4391 that is used in this case wasn't being freed if the match failed.
4393 4. Tidied pcretest to ensure it frees memory that it gets.
4395 5. Temporary memory was being obtained in the case where the passed offsets
4396 vector was exactly big enough.
4398 6. Corrected definition of offsetof() from change 5 below.
4400 7. I had screwed up change 6 below and broken the rules for the use of
4401 setjmp(). Now fixed.
4404 Version 1.03 18-Dec-97
4405 ----------------------
4407 1. A erroneous regex with a missing opening parenthesis was correctly
4408 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4409 on some systems.
4411 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4412 it was reported that one broken compiler failed on the former because "code" is
4413 also an independent variable.
4415 3. The erroneous regex a[]b caused an array overrun reference.
4417 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4418 fail on data ending with that character. (It was going on too far, and checking
4419 the next character, typically a binary zero.) This was specific to the
4420 optimized code for single-character negative classes.
4422 5. Added a contributed patch from the TIN world which does the following:
4424 + Add an undef for memmove, in case the the system defines a macro for it.
4426 + Add a definition of offsetof(), in case there isn't one. (I don't know
4427 the reason behind this - offsetof() is part of the ANSI standard - but
4428 it does no harm).
4430 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4431 most of the places where whitespace preceded '#'. I have given up and
4432 allowed the remaining 2 cases to be at the margin.
4434 + Rename some variables in pcre to eliminate shadowing. This seems very
4435 pedantic, but does no harm, of course.
4437 6. Moved the call to setjmp() into its own function, to get rid of warnings
4438 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4440 7. Constructs such as \d{8,} were compiling into the equivalent of
4441 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4442 outcome, but in this particular case used more store than had been allocated,
4443 which caused the bug to be discovered because it threw up an internal error.
4445 8. The debugging code in both pcre and pcretest for outputting the compiled
4446 form of a regex was going wrong in the case of back references followed by
4447 curly-bracketed repeats.
4450 Version 1.02 12-Dec-97
4451 ----------------------
4453 1. Typos in pcre.3 and comments in the source fixed.
4455 2. Applied a contributed patch to get rid of places where it used to remove
4456 'const' from variables, and fixed some signed/unsigned and uninitialized
4457 variable warnings.
4459 3. Added the "runtest" target to Makefile.
4461 4. Set default compiler flag to -O2 rather than just -O.
4464 Version 1.01 19-Nov-97
4465 ----------------------
4467 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4468 like /([ab]*)*/, that is, for classes with more than one character in them.
4470 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4471 as /((?>a*))*/ (a PCRE_EXTRA facility).
4474 Version 1.00 18-Nov-97
4475 ----------------------
4477 1. Added compile-time macros to support systems such as SunOS4 which don't have
4478 memmove() or strerror() but have other things that can be used instead.
4480 2. Arranged that "make clean" removes the executables.
4483 Version 0.99 27-Oct-97
4484 ----------------------
4486 1. Fixed bug in code for optimizing classes with only one character. It was
4487 initializing a 32-byte map regardless, which could cause it to run off the end
4488 of the memory it had got.
4490 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4493 Version 0.98 22-Oct-97
4494 ----------------------
4496 1. Fixed bug in code for handling temporary memory usage when there are more
4497 back references than supplied space in the ovector. This could cause segfaults.
4500 Version 0.97 21-Oct-97
4501 ----------------------
4503 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4505 2. Optimized negated single characters not to use a bit map.
4507 3. Brought error texts together as macro definitions; clarified some of them;
4508 fixed one that was wrong - it said "range out of order" when it meant "invalid
4509 escape sequence".
4511 4. Changed some char * arguments to const char *.
4513 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4515 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4516 pcretest.
4519 Version 0.96 16-Oct-97
4520 ----------------------
4522 1. Added a simple "pgrep" utility to the distribution.
4524 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4525 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4526 where "ddd" means "one or more decimal digits".
4528 3. Fixed serious bug. If a pattern had a back reference, but the call to
4529 pcre_exec() didn't supply a large enough ovector to record the related
4530 identifying subpattern, the match always failed. PCRE now remembers the number
4531 of the largest back reference, and gets some temporary memory in which to save
4532 the offsets during matching if necessary, in order to ensure that
4533 backreferences always work.
4535 4. Increased the compatibility with Perl in a number of ways:
4537 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4538 to request this handling. The option can be set at compile or exec time.
4540 (b) $ matches before a terminating newline by default; an option
4541 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4542 mode). The option can be set at compile or exec time.
4544 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4545 the same as Perl's. If the decimal number it represents is less than 10
4546 or there aren't that many previous left capturing parentheses, an octal
4547 escape is read. Inside a character class, it's always an octal escape,
4548 even if it is a single digit.
4550 (d) An escaped but undefined alphabetic character is taken as a literal,
4551 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4552 escapes.
4554 (e) {0} is now permitted. (The previous item is removed from the compiled
4555 pattern).
4557 5. Changed all the names of code files so that the basic parts are no longer
4558 than 10 characters, and abolished the teeny "globals.c" file.
4560 6. Changed the handling of character classes; they are now done with a 32-byte
4561 bit map always.
4563 7. Added the -d and /D options to pcretest to make it possible to look at the
4564 internals of compilation without having to recompile pcre.
4567 Version 0.95 23-Sep-97
4568 ----------------------
4570 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4571 \x20 at the start of a run of normal characters. These were being treated as
4572 real characters, instead of the source characters being re-checked.
4575 Version 0.94 18-Sep-97
4576 ----------------------
4578 1. The functions are now thread-safe, with the caveat that the global variables
4579 containing pointers to malloc() and free() or alternative functions are the
4580 same for all threads.
4582 2. Get pcre_study() to generate a bitmap of initial characters for non-
4583 anchored patterns when this is possible, and use it if passed to pcre_exec().
4586 Version 0.93 15-Sep-97
4587 ----------------------
4589 1. /(b)|(:+)/ was computing an incorrect first character.
4591 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4592 but not actually doing anything yet.
4594 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4595 as Perl does (e.g. [-az] or [az-]).
4597 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4598 all possible positions.
4600 5. Split up into different modules to avoid including unneeded functions in a
4601 compiled binary. However, compile and exec are still in one module. The "study"
4602 function is split off.
4604 6. The character tables are now in a separate module whose source is generated
4605 by an auxiliary program - but can then be edited by hand if required. There are
4606 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4607 toupper() in the code.
4609 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4610 make them global. Abolish the function for setting them, as the caller can now
4611 set them directly.
4614 Version 0.92 11-Sep-97
4615 ----------------------
4617 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4618 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4620 2. Caseless matching was not working in character classes if the characters in
4621 the pattern were in upper case.
4623 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4625 4. Make PCRE_ANCHORED public and accept as a compile option.
4627 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4628 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4629 pass them.
4631 6. Give an error if bad option bits passed at compile or run time.
4633 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4634 pcretest to cause it to pass that flag.
4636 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4637 options, and the first character, if set.
4639 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4642 Version 0.91 10-Sep-97
4643 ----------------------
4645 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4646 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4648 2. PCRE was looping on encountering an indefinitely repeated back reference to
4649 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4650 Perl does - treats the match as successful.
4652 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5