ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 751 - (show annotations)
Fri Nov 18 11:13:30 2011 UTC (8 years, 8 months ago) by ph10
File size: 219700 byte(s)
Code tidies.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.21
5 ------------
7 1. Updating the JIT compiler.
9 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
10 are added as well.
12 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
13 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
14 calling _pcre_jit_exec. Some extra comments are added.
16 4. Mark settings inside atomic groups that do not contain any capturing
17 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
18 was introduced by change 18 for 8.20.
20 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
21 ECMA-262 standard.
23 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
24 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
25 This bug was probably introduced by change 9 of 8.13.
27 7. While fixing 6 above, I noticed that a number of other items were being
28 incorrectly rejected as "not fixed length". This arose partly because newer
29 opcodes had not been added to the fixed-length checking code. I have (a)
30 corrected the bug and added tests for these items, and (b) arranged for an
31 error to occur if an unknown opcode is encountered while checking for fixed
32 length instead of just assuming "not fixed length". The items that were
33 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
34 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
35 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
37 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
38 being incorrectly compiled and would have given unpredicatble results.
40 9. A possessively repeated subpattern with minimum repeat count greater than
41 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
42 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
43 the first (A) could occur when it should not.
45 10. Add a cast and remove a redundant test from the code.
48 Version 8.20 21-Oct-2011
49 ------------------------
51 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
52 a POSIX class. After further experiments with Perl, which convinced me that
53 Perl has bugs and confusions, a closing square bracket is no longer allowed
54 in a POSIX name. This bug also affected patterns with classes that started
55 with full stops.
57 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
58 captured substring, but while checking the failing first alternative,
59 substring 1 is temporarily captured. If the output vector supplied to
60 pcre_exec() was not big enough for this capture, the yield of the function
61 was still zero ("insufficient space for captured substrings"). This cannot
62 be totally fixed without adding another stack variable, which seems a lot
63 of expense for a edge case. However, I have improved the situation in cases
64 such as /(a)(b)x|abc/ matched against "abc", where the return code
65 indicates that fewer than the maximum number of slots in the ovector have
66 been set.
68 3. Related to (2) above: when there are more back references in a pattern than
69 slots in the output vector, pcre_exec() uses temporary memory during
70 matching, and copies in the captures as far as possible afterwards. It was
71 using the entire output vector, but this conflicts with the specification
72 that only 2/3 is used for passing back captured substrings. Now it uses
73 only the first 2/3, for compatibility. This is, of course, another edge
74 case.
76 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
77 main code base, and can be used by building with --enable-jit. When this is
78 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
79 runtime --no-jit option is given.
81 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
82 ovector, the return from the function was zero, implying that there were
83 other matches that did not fit. The correct "exactly full" value is now
84 returned.
86 6. If a subpattern that was called recursively or as a subroutine contained
87 (*PRUNE) or any other control that caused it to give a non-standard return,
88 invalid errors such as "Error -26 (nested recursion at the same subject
89 position)" or even infinite loops could occur.
91 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
92 computing the minimum length on reaching *ACCEPT, and so ended up with the
93 wrong value of 1 rather than 0. Further investigation indicates that
94 computing a minimum subject length in the presence of *ACCEPT is difficult
95 (think back references, subroutine calls), and so I have changed the code
96 so that no minimum is registered for a pattern that contains *ACCEPT.
98 8. If (*THEN) was present in the first (true) branch of a conditional group,
99 it was not handled as intended. [But see 16 below.]
101 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
102 Sheri Pierce.
104 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
105 the first byte in a match must be "a".
107 11. Change 17 for 8.13 increased the recursion depth for patterns like
108 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
109 pattern contains any instances of (*THEN). If it does not, the old
110 optimizations are restored. It would be nice to do this on a per-group
111 basis, but at the moment that is not feasible.
113 12. In some environments, the output of pcretest -C is CRLF terminated. This
114 broke RunTest's code that checks for the link size. A single white space
115 character after the value is now allowed for.
117 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
118 For "fr", it uses the Windows-specific input and output files.
120 14. If (*THEN) appeared in a group that was called recursively or as a
121 subroutine, it did not work as intended. [But see next item.]
123 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
124 pattern fragments (but not containing any | characters). If A and B are
125 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
126 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
127 D. In other words, Perl considers parentheses that do not contain any |
128 characters to be part of a surrounding alternative, whereas PCRE was
129 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
130 differently. PCRE now behaves in the same way as Perl, except in the case
131 of subroutine/recursion calls such as (?1) which have in any case always
132 been different (but PCRE had them first :-).
134 16. Related to 15 above: Perl does not treat the | in a conditional group as
135 creating alternatives. Such a group is treated in the same way as an
136 ordinary group without any | characters when processing (*THEN). PCRE has
137 been changed to match Perl's behaviour.
139 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
140 RunGrepTest script failed.
142 18. Change 22 for version 13 caused atomic groups to use more stack. This is
143 inevitable for groups that contain captures, but it can lead to a lot of
144 stack use in large patterns. The old behaviour has been restored for atomic
145 groups that do not contain any capturing parentheses.
147 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
148 suppress the check for a minimum subject length at run time. (If it was
149 given to pcre_exec() or pcre_dfa_exec() it did work.)
151 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
152 fail to work when decoding hex characters in data strings in EBCDIC
153 environments.
155 21. It appears that in at least one Mac OS environment, the isxdigit() function
156 is implemented as a macro that evaluates to its argument more than once,
157 contravening the C 90 Standard (I haven't checked a later standard). There
158 was an instance in pcretest which caused it to go wrong when processing
159 \x{...} escapes in subject strings. The has been rewritten to avoid using
160 things like p++ in the argument of isxdigit().
163 Version 8.13 16-Aug-2011
164 ------------------------
166 1. The Unicode data tables have been updated to Unicode 6.0.0.
168 2. Two minor typos in pcre_internal.h have been fixed.
170 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
171 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
172 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
174 4. There were a number of related bugs in the code for matching backrefences
175 caselessly in UTF-8 mode when codes for the characters concerned were
176 different numbers of bytes. For example, U+023A and U+2C65 are an upper
177 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
178 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
179 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
180 2-byte code at the end of the subject (it thought there wasn't enough data
181 left).
183 5. Comprehensive information about what went wrong is now returned by
184 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
185 as the output vector has at least 2 elements. The offset of the start of
186 the failing character and a reason code are placed in the vector.
188 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
189 now returned is for the first byte of the failing character, instead of the
190 last byte inspected. This is an incompatible change, but I hope it is small
191 enough not to be a problem. It makes the returned offset consistent with
192 pcre_exec() and pcre_dfa_exec().
194 7. pcretest now gives a text phrase as well as the error number when
195 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
196 failure, the offset and reason code are output.
198 8. When \R was used with a maximizing quantifier it failed to skip backwards
199 over a \r\n pair if the subsequent match failed. Instead, it just skipped
200 back over a single character (\n). This seems wrong (because it treated the
201 two characters as a single entity when going forwards), conflicts with the
202 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
203 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
204 has been changed.
206 9. Some internal refactoring has changed the processing so that the handling
207 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
208 time (the PCRE_DOTALL option was changed this way some time ago: version
209 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
210 which was always a bit of a fudge. It also means that there is one less
211 argument for the match() function, which reduces its stack requirements
212 slightly. This change also fixes an incompatibility with Perl: the pattern
213 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
215 10. More internal refactoring has drastically reduced the number of recursive
216 calls to match() for possessively repeated groups such as (abc)++ when
217 using pcre_exec().
219 11. While implementing 10, a number of bugs in the handling of groups were
220 discovered and fixed:
222 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
223 (a|)*(?1) gave a compile-time internal error.
224 ((a|)+)+ did not notice that the outer group could match an empty string.
225 (^a|^)+ was not marked as anchored.
226 (.*a|.*)+ was not marked as matching at start or after a newline.
228 12. Yet more internal refactoring has removed another argument from the match()
229 function. Special calls to this function are now indicated by setting a
230 value in a variable in the "match data" data block.
232 13. Be more explicit in pcre_study() instead of relying on "default" for
233 opcodes that mean there is no starting character; this means that when new
234 ones are added and accidentally left out of pcre_study(), testing should
235 pick them up.
237 14. The -s option of pcretest has been documented for ages as being an old
238 synonym of -m (show memory usage). I have changed it to mean "force study
239 for every regex", that is, assume /S for every regex. This is similar to -i
240 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
241 using it. It makes it easier to run collections of tests with and without
242 study enabled, and thereby test pcre_study() more easily. All the standard
243 tests are now run with and without -s (but some patterns can be marked as
244 "never study" - see 20 below).
246 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
247 restoration of the capturing data to the outer values was not happening
248 correctly.
250 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
251 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
252 pattern had matched an empty string, and so incorrectly returned a no
253 match.
255 17. There was optimizing code for the last branch of non-capturing parentheses,
256 and also for the obeyed branch of a conditional subexpression, which used
257 tail recursion to cut down on stack usage. Unfortunately, now that there is
258 the possibility of (*THEN) occurring in these branches, tail recursion is
259 no longer possible because the return has to be checked for (*THEN). These
260 two optimizations have therefore been removed. [But see 8.20/11 above.]
262 18. If a pattern containing \R was studied, it was assumed that \R always
263 matched two bytes, thus causing the minimum subject length to be
264 incorrectly computed because \R can also match just one byte.
266 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
267 was incorrectly computed.
269 20. If /S is present twice on a test pattern in pcretest input, it now
270 *disables* studying, thereby overriding the use of -s on the command line
271 (see 14 above). This is necessary for one or two tests to keep the output
272 identical in both cases.
274 21. When (*ACCEPT) was used in an assertion that matched an empty string and
275 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
277 22. When an atomic group that contained a capturing parenthesis was
278 successfully matched, but the branch in which it appeared failed, the
279 capturing was not being forgotten if a higher numbered group was later
280 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
281 group 1 to "a", when in fact it should be unset. This applied to multi-
282 branched capturing and non-capturing groups, repeated or not, and also to
283 positive assertions (capturing in negative assertions does not happen
284 in PCRE) and also to nested atomic groups.
286 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
287 subject after a captured substring, to make it easier to tell which of a
288 number of identical substrings has been captured.
290 24. The way atomic groups are processed by pcre_exec() has been changed so that
291 if they are repeated, backtracking one repetition now resets captured
292 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
293 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
294 "aaa". Previously, it would have been "a". As part of this code
295 refactoring, the way recursive calls are handled has also been changed.
297 25. If an assertion condition captured any substrings, they were not passed
298 back unless some other capturing happened later. For example, if
299 (?(?=(a))a) was matched against "a", no capturing was returned.
301 26. When studying a pattern that contained subroutine calls or assertions,
302 the code for finding the minimum length of a possible match was handling
303 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
304 group 1 called group 2 while simultaneously a separate group 2 called group
305 1). A stack overflow occurred in this case. I have fixed this by limiting
306 the recursion depth to 10.
308 27. Updated RunTest.bat in the distribution to the version supplied by Tom
309 Fortmann. This supports explicit test numbers on the command line, and has
310 argument validation and error reporting.
312 28. An instance of \X with an unlimited repeat could fail if at any point the
313 first character it looked at was a mark character.
315 29. Some minor code refactoring concerning Unicode properties and scripts
316 should reduce the stack requirement of match() slightly.
318 30. Added the '=' option to pcretest to check the setting of unused capturing
319 slots at the end of the pattern, which are documented as being -1, but are
320 not included in the return count.
322 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
323 compiled something random. Now it gives a compile-time error (as does
324 Perl).
326 32. A *MARK encountered during the processing of a positive assertion is now
327 recorded and passed back (compatible with Perl).
329 33. If --only-matching or --colour was set on a pcregrep call whose pattern
330 had alternative anchored branches, the search for a second match in a line
331 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
332 matched the line "0102" twice. The same bug affected patterns that started
333 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
334 twice.
336 34. Previously, PCRE did not allow quantification of assertions. However, Perl
337 does, and because of capturing effects, quantifying parenthesized
338 assertions may at times be useful. Quantifiers are now allowed for
339 parenthesized assertions.
341 35. A minor code tidy in pcre_compile() when checking options for \R usage.
343 36. \g was being checked for fancy things in a character class, when it should
344 just be a literal "g".
346 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
347 appearance of a nested POSIX class supersedes an apparent external class.
348 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
349 unescaped square brackets may also appear as part of class names. For
350 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
351 more like Perl. (But see 8.20/1 above.)
353 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
354 was because it thought it was \N{name}, which is not supported).
356 39. Add minix to OS list not supporting the -S option in pcretest.
358 40. PCRE tries to detect cases of infinite recursion at compile time, but it
359 cannot analyze patterns in sufficient detail to catch mutual recursions
360 such as ((?1))((?2)). There is now a runtime test that gives an error if a
361 subgroup is called recursively as a subpattern for a second time at the
362 same position in the subject string. In previous releases this might have
363 been caught by the recursion limit, or it might have run out of stack.
365 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
366 happen only once. PCRE was, however incorrectly giving a compile time error
367 "recursive call could loop indefinitely" because it cannot analyze the
368 pattern in sufficient detail. The compile time test no longer happens when
369 PCRE is compiling a conditional subpattern, but actual runaway loops are
370 now caught at runtime (see 40 above).
372 42. It seems that Perl allows any characters other than a closing parenthesis
373 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
374 has been changed to be the same.
376 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
377 as not to get warnings when autogen.sh is called. Also changed
378 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
380 44. To help people who use pcregrep to scan files containing exceedingly long
381 lines, the following changes have been made:
383 (a) The default value of the buffer size parameter has been increased from
384 8K to 20K. (The actual buffer used is three times this size.)
386 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
387 PCRE is built.
389 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
390 to be set at run time.
392 (d) Numerical values in pcregrep options can be followed by K or M, for
393 example --buffer-size=50K.
395 (e) If a line being scanned overflows pcregrep's buffer, an error is now
396 given and the return code is set to 2.
398 45. Add a pointer to the latest mark to the callout data block.
400 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
401 partial match of an empty string instead of no match. This was specific to
402 the use of ".".
404 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
405 complete match instead of a partial match. This bug was dependent on both
406 the PCRE_UTF8 and PCRE_DOTALL options being set.
408 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
409 starting byte set, because \b was not being ignored.
412 Version 8.12 15-Jan-2011
413 ------------------------
415 1. Fixed some typos in the markup of the man pages, and wrote a script that
416 checks for such things as part of the documentation building process.
418 2. On a big-endian 64-bit system, pcregrep did not correctly process the
419 --match-limit and --recursion-limit options (added for 8.11). In
420 particular, this made one of the standard tests fail. (The integer value
421 went into the wrong half of a long int.)
423 3. If the --colour option was given to pcregrep with -v (invert match), it
424 did strange things, either producing crazy output, or crashing. It should,
425 of course, ignore a request for colour when reporting lines that do not
426 match.
428 4. Another pcregrep bug caused similar problems if --colour was specified with
429 -M (multiline) and the pattern match finished with a line ending.
431 5. In pcregrep, when a pattern that ended with a literal newline sequence was
432 matched in multiline mode, the following line was shown as part of the
433 match. This seems wrong, so I have changed it.
435 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
436 the check for further matches in the same line (so they could be coloured)
437 to overrun the end of the current line. If another match was found, it was
438 incorrectly shown (and then shown again when found in the next line).
440 7. If pcregrep was compiled under Windows, there was a reference to the
441 function pcregrep_exit() before it was defined. I am assuming this was
442 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
443 reported by a user. I've moved the definition above the reference.
446 Version 8.11 10-Dec-2010
447 ------------------------
449 1. (*THEN) was not working properly if there were untried alternatives prior
450 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
451 backtracked to try for "b" instead of moving to the next alternative branch
452 at the same level (in this case, to look for "c"). The Perl documentation
453 is clear that when (*THEN) is backtracked onto, it goes to the "next
454 alternative in the innermost enclosing group".
456 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
457 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
458 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
459 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
460 (*THEN).
462 3. If \s appeared in a character class, it removed the VT character from
463 the class, even if it had been included by some previous item, for example
464 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
465 of \s, but is part of the POSIX "space" class.)
467 4. A partial match never returns an empty string (because you can always
468 match an empty string at the end of the subject); however the checking for
469 an empty string was starting at the "start of match" point. This has been
470 changed to the "earliest inspected character" point, because the returned
471 data for a partial match starts at this character. This means that, for
472 example, /(?<=abc)def/ gives a partial match for the subject "abc"
473 (previously it gave "no match").
475 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
476 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
477 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
478 has an implication that the given string is incomplete (because a partial
479 match is preferred over a full match). For this reason, these items now
480 give a partial match in this situation. [Aside: previously, the one case
481 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
482 match rather than a full match, which was wrong by the old rules, but is
483 now correct.]
485 6. There was a bug in the handling of #-introduced comments, recognized when
486 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
487 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
488 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
489 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
490 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
491 places in pcre_compile().
493 7. Related to (6) above, when pcre_compile() was skipping #-introduced
494 comments when looking ahead for named forward references to subpatterns,
495 the only newline sequence it recognized was NL. It now handles newlines
496 according to the set newline convention.
498 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
499 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
500 cater for a lack of strerror(). These oversights have been fixed.
502 9. Added --match-limit and --recursion-limit to pcregrep.
504 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
506 11. When the -o option was used, pcregrep was setting a return code of 1, even
507 when matches were found, and --line-buffered was not being honoured.
509 12. Added an optional parentheses number to the -o and --only-matching options
510 of pcregrep.
512 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
513 can match an empty string. The code to do it in pcretest and pcredemo
514 needed fixing:
516 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
517 only one byte after an empty string match just before CRLF (this case
518 just got forgotten; "any" and "anycrlf" were OK).
520 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
521 mode when an empty string match preceded an ASCII character followed by
522 a non-ASCII character. (The code for advancing by one character rather
523 than one byte was nonsense.)
525 (c) The pcredemo.c sample program did not have any code at all to handle
526 the cases when CRLF is a valid newline sequence.
528 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
529 as a starting offset was within the subject string. There is now a new
530 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
531 negative or greater than the length of the string. In order to test this,
532 pcretest is extended to allow the setting of negative starting offsets.
534 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
535 starting offset points to the beginning of a UTF-8 character was
536 unnecessarily clumsy. I tidied it up.
538 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
539 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
541 17. Nobody had reported that the --include_dir option, which was added in
542 release 7.7 should have been called --include-dir (hyphen, not underscore)
543 for compatibility with GNU grep. I have changed it to --include-dir, but
544 left --include_dir as an undocumented synonym, and the same for
545 --exclude-dir, though that is not available in GNU grep, at least as of
546 release 2.5.4.
548 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
549 characters from a string of bytes) have been redefined so as not to use
550 loops, in order to improve performance in some environments. At the same
551 time, I abstracted some of the common code into auxiliary macros to save
552 repetition (this should not affect the compiled code).
554 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
555 compile-time error is now given if \c is not followed by an ASCII
556 character, that is, a byte less than 128. (In EBCDIC mode, the code is
557 different, and any byte value is allowed.)
559 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
560 START_OPTIMIZE option, which is now allowed at compile time - but just
561 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
562 to pcregrep and other applications that have no direct access to PCRE
563 options. The new /Y option in pcretest sets this option when calling
564 pcre_compile().
566 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
567 back references. Groups containing recursive back references were forced to
568 be atomic by that change, but in the case of named groups, the amount of
569 memory required was incorrectly computed, leading to "Failed: internal
570 error: code overflow". This has been fixed.
572 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
573 pcretest.c, to avoid build problems in some Borland environments.
576 Version 8.10 25-Jun-2010
577 ------------------------
579 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
580 THEN.
582 2. (*ACCEPT) was not working when inside an atomic group.
584 3. Inside a character class, \B is treated as a literal by default, but
585 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
586 causes the error). The code is unchanged, but I tidied the documentation.
588 4. Inside a character class, PCRE always treated \R and \X as literals,
589 whereas Perl faults them if its -w option is set. I have changed PCRE so
590 that it faults them when PCRE_EXTRA is set.
592 5. Added support for \N, which always matches any character other than
593 newline. (It is the same as "." when PCRE_DOTALL is not set.)
595 6. When compiling pcregrep with newer versions of gcc which may have
596 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
597 declared with attribute warn_unused_result" were given. Just casting the
598 result to (void) does not stop the warnings; a more elaborate fudge is
599 needed. I've used a macro to implement this.
601 7. Minor change to pcretest.c to avoid a compiler warning.
603 8. Added four artifical Unicode properties to help with an option to make
604 \s etc use properties (see next item). The new properties are: Xan
605 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
607 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
608 use Unicode properties. (*UCP) at the start of a pattern can be used to set
609 this option. Modified pcretest to add /W to test this facility. Added
610 REG_UCP to make it available via the POSIX interface.
612 10. Added --line-buffered to pcregrep.
614 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
615 studied, and the match started with a letter with a code point greater than
616 127 whose first byte was different to the first byte of the other case of
617 the letter, the other case of this starting letter was not recognized
618 (#976).
620 12. If a pattern that was studied started with a repeated Unicode property
621 test, for example, \p{Nd}+, there was the theoretical possibility of
622 setting up an incorrect bitmap of starting bytes, but fortunately it could
623 not have actually happened in practice until change 8 above was made (it
624 added property types that matched character-matching opcodes).
626 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
627 possible starting bytes for non-anchored patterns.
629 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
630 \R, and also a number of cases that involve Unicode properties, both
631 explicit and implicit when PCRE_UCP is set.
633 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
634 input, it could crash or give wrong results if characters with values
635 greater than 0xc0 were present in the subject string. (Detail: it assumed
636 UTF-8 input when processing these items.)
638 16. Added a lot of (int) casts to avoid compiler warnings in systems where
639 size_t is 64-bit (#991).
641 17. Added a check for running out of memory when PCRE is compiled with
642 --disable-stack-for-recursion (#990).
644 18. If the last data line in a file for pcretest does not have a newline on
645 the end, a newline was missing in the output.
647 19. The default pcre_chartables.c file recognizes only ASCII characters (values
648 less than 128) in its various bitmaps. However, there is a facility for
649 generating tables according to the current locale when PCRE is compiled. It
650 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
651 characters, are recognized by isspace() and therefore were getting set in
652 these tables, and indeed these tables seem to approximate to ISO 8859. This
653 caused a problem in UTF-8 mode when pcre_study() was used to create a list
654 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
655 which of course cannot start UTF-8 characters. I have changed the code so
656 that only real ASCII characters (less than 128) and the correct starting
657 bytes for UTF-8 encodings are set for characters greater than 127 when in
658 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
659 altogether.)
661 20. Added the /T option to pcretest so as to be able to run tests with non-
662 standard character tables, thus making it possible to include the tests
663 used for 19 above in the standard set of tests.
665 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
666 reference to a subpattern the other side of a comment that contains an
667 opening parenthesis caused either an internal compiling error, or a
668 reference to the wrong subpattern.
671 Version 8.02 19-Mar-2010
672 ------------------------
674 1. The Unicode data tables have been updated to Unicode 5.2.0.
676 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
677 configured.
679 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
680 original author of that file, following a query about its status.
682 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
683 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
685 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
686 quantifier applied to a forward-referencing subroutine call, could compile
687 incorrect code or give the error "internal error: previously-checked
688 referenced subpattern not found".
690 6. Both MS Visual Studio and Symbian OS have problems with initializing
691 variables to point to external functions. For these systems, therefore,
692 pcre_malloc etc. are now initialized to local functions that call the
693 relevant global functions.
695 7. There were two entries missing in the vectors called coptable and poptable
696 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
697 I've fixed the data, and added a kludgy way of testing at compile time that
698 the lengths are correct (equal to the number of opcodes).
700 8. Following on from 7, I added a similar kludge to check the length of the
701 eint vector in pcreposix.c.
703 9. Error texts for pcre_compile() are held as one long string to avoid too
704 much relocation at load time. To find a text, the string is searched,
705 counting zeros. There was no check for running off the end of the string,
706 which could happen if a new error number was added without updating the
707 string.
709 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
711 11. \K was not working if it appeared in an atomic group or in a group that
712 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
713 \K is "not well defined" if used in an assertion. PCRE now accepts it if
714 the assertion is positive, but not if it is negative.
716 12. Change 11 fortuitously reduced the size of the stack frame used in the
717 "match()" function of pcre_exec.c by one pointer. Forthcoming
718 implementation of support for (*MARK) will need an extra pointer on the
719 stack; I have reserved it now, so that the stack frame size does not
720 decrease.
722 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
723 item in branch that calls a recursion is a subroutine call - as in the
724 second branch in the above example - was incorrectly given the compile-
725 time error "recursive call could loop indefinitely" because pcre_compile()
726 was not correctly checking the subroutine for matching a non-empty string.
728 14. The checks for overrunning compiling workspace could trigger after an
729 overrun had occurred. This is a "should never occur" error, but it can be
730 triggered by pathological patterns such as hundreds of nested parentheses.
731 The checks now trigger 100 bytes before the end of the workspace.
733 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
736 Version 8.01 19-Jan-2010
737 ------------------------
739 1. If a pattern contained a conditional subpattern with only one branch (in
740 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
741 computed the wrong minimum data length (which is of course zero for such
742 subpatterns). This could cause incorrect "no match" results.
744 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
745 the pattern is reset in the first branch, pcre_compile() failed with
746 "internal error: code overflow at offset...". This happened only when
747 the reset was to the original external option setting. (An optimization
748 abstracts leading options settings into an external setting, which was the
749 cause of this.)
751 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
752 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
753 assertion pattern did not match (meaning that the assertion was true), it
754 was incorrectly treated as false if the SKIP had been reached during the
755 matching. This also applied to assertions used as conditions.
757 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
758 assertion subpattern, including such a pattern used as a condition,
759 unpredictable results occurred, instead of the error return
762 5. The C++ GlobalReplace function was not working like Perl for the special
763 situation when an empty string is matched. It now does the fancy magic
764 stuff that is necessary.
766 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
767 removed. (These were left over from very, very early versions of PCRE.)
769 7. Some cosmetic changes to the code to make life easier when compiling it
770 as part of something else:
772 (a) Change DEBUG to PCRE_DEBUG.
774 (b) In pcre_compile(), rename the member of the "branch_chain" structure
775 called "current" as "current_branch", to prevent a collision with the
776 Linux macro when compiled as a kernel module.
778 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
779 prevent a collision with the Linux macro when compiled as a kernel
780 module.
782 8. In pcre_compile() there are some checks for integer overflows that used to
783 cast potentially large values to (double). This has been changed to that
784 when building, a check for int64_t is made, and if it is found, it is used
785 instead, thus avoiding the use of floating point arithmetic. (There is no
786 other use of FP in PCRE.) If int64_t is not found, the fallback is to
787 double.
789 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
790 2005 (difference between two addresses compared to an unsigned value).
792 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
793 custom one, because of the following reported problem in Windows:
795 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
796 under Win32.
797 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
798 therefore missing the function definition.
799 - The compiler thus generates a "C" signature for the test function.
800 - The linker fails to find the "C" function.
801 - PCRE fails to configure if asked to do so against libbz2.
803 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
804 messages were output:
806 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
807 rerunning libtoolize, to keep the correct libtool macros in-tree.
808 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
810 I have done both of these things.
812 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
813 most of the time, it *can* run out if it is given a pattern that contains a
814 runaway infinite recursion. I updated the discussion in the pcrestack man
815 page.
817 13. Now that we have gone to the x.xx style of version numbers, the minor
818 version may start with zero. Using 08 or 09 is a bad idea because users
819 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
820 interpreted as invalid octal numbers. I've updated the previous comment in
821 configure.ac, and also added a check that gives an error if 08 or 09 are
822 used.
824 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
825 causing partial matching to fail when the end of the subject matched \W
826 in a UTF-8 pattern where \W was quantified with a minimum of 3.
828 15. There were some discrepancies between the declarations in pcre_internal.h
829 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
830 their definitions. The declarations used "const uschar *" and the
831 definitions used USPTR. Even though USPTR is normally defined as "const
832 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
833 reported that: "This difference in casting confuses some C++ compilers, for
834 example, SunCC recognizes above declarations as different functions and
835 generates broken code for hbpcre." I have changed the declarations to use
836 USPTR.
838 16. GNU libtool is named differently on some systems. The autogen.sh script now
839 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
840 (FreeBSD).
842 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
843 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
844 comment: "Figure out how to create a longlong from a string: strtoll and
845 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
846 instance, but it only takes 2 args instead of 3!"
848 18. A subtle bug concerned with back references has been fixed by a change of
849 specification, with a corresponding code fix. A pattern such as
850 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
851 refers, was giving matches when it shouldn't. For example, xa=xaaa would
852 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
853 same bug. Such groups have to be quantified to be useful, or contained
854 inside another quantified group. (If there's no repetition, the reference
855 can never match.) The problem arises because, having left the group and
856 moved on to the rest of the pattern, a later failure that backtracks into
857 the group uses the captured value from the final iteration of the group
858 rather than the correct earlier one. I have fixed this in PCRE by forcing
859 any group that contains a reference to itself to be an atomic group; that
860 is, there cannot be any backtracking into it once it has completed. This is
861 similar to recursive and subroutine calls.
864 Version 8.00 19-Oct-09
865 ----------------------
867 1. The table for translating pcre_compile() error codes into POSIX error codes
868 was out-of-date, and there was no check on the pcre_compile() error code
869 being within the table. This could lead to an OK return being given in
870 error.
872 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
873 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
874 in a Windows environment.
876 3. The pcregrep --count option prints the count for each file even when it is
877 zero, as does GNU grep. However, pcregrep was also printing all files when
878 --files-with-matches was added. Now, when both options are given, it prints
879 counts only for those files that have at least one match. (GNU grep just
880 prints the file name in this circumstance, but including the count seems
881 more useful - otherwise, why use --count?) Also ensured that the
882 combination -clh just lists non-zero counts, with no names.
884 4. The long form of the pcregrep -F option was incorrectly implemented as
885 --fixed_strings instead of --fixed-strings. This is an incompatible change,
886 but it seems right to fix it, and I didn't think it was worth preserving
887 the old behaviour.
889 5. The command line items --regex=pattern and --regexp=pattern were not
890 recognized by pcregrep, which required --regex pattern or --regexp pattern
891 (with a space rather than an '='). The man page documented the '=' forms,
892 which are compatible with GNU grep; these now work.
894 6. No libpcreposix.pc file was created for pkg-config; there was just
895 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
897 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
898 when UCP support is not needed, by modifying the Python script that
899 generates it from Unicode data files. This should not matter if the module
900 is correctly used as a library, but I received one complaint about 50K of
901 unwanted data. My guess is that the person linked everything into his
902 program rather than using a library. Anyway, it does no harm.
904 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
905 was a minimum greater than 1 for a wide character in a possessive
906 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
907 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
908 character. Chaos in the form of incorrect output or a compiling loop could
909 result.
911 9. The restrictions on what a pattern can contain when partial matching is
912 requested for pcre_exec() have been removed. All patterns can now be
913 partially matched by this function. In addition, if there are at least two
914 slots in the offset vector, the offset of the earliest inspected character
915 for the match and the offset of the end of the subject are set in them when
916 PCRE_ERROR_PARTIAL is returned.
918 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
919 synonymous with PCRE_PARTIAL, for backwards compatibility, and
920 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
921 and may be more useful for multi-segment matching.
923 11. Partial matching with pcre_exec() is now more intuitive. A partial match
924 used to be given if ever the end of the subject was reached; now it is
925 given only if matching could not proceed because another character was
926 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
927 string "Z", which now yields "no match" instead of "partial match". In the
928 case of pcre_dfa_exec(), "no match" is given if every matching path for the
929 final character ended with (*FAIL).
931 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
932 if the pattern had a "must contain" character that was already found in the
933 earlier partial match, unless partial matching was again requested. For
934 example, with the pattern /dog.(body)?/, the "must contain" character is
935 "g". If the first part-match was for the string "dog", restarting with
936 "sbody" failed. This bug has been fixed.
938 13. The string returned by pcre_dfa_exec() after a partial match has been
939 changed so that it starts at the first inspected character rather than the
940 first character of the match. This makes a difference only if the pattern
941 starts with a lookbehind assertion or \b or \B (\K is not supported by
942 pcre_dfa_exec()). It's an incompatible change, but it makes the two
943 matching functions compatible, and I think it's the right thing to do.
945 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
946 so that the demonstration program is easily available in environments where
947 PCRE has not been installed from source.
949 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
950 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
951 library.
953 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
954 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
955 is not the first non-POSIX option to be added. Clearly some people find
956 these options useful.
958 17. If a caller to the POSIX matching function regexec() passes a non-zero
959 value for nmatch with a NULL value for pmatch, the value of
960 nmatch is forced to zero.
962 18. RunGrepTest did not have a test for the availability of the -u option of
963 the diff command, as RunTest does. It now checks in the same way as
964 RunTest, and also checks for the -b option.
966 19. If an odd number of negated classes containing just a single character
967 interposed, within parentheses, between a forward reference to a named
968 subpattern and the definition of the subpattern, compilation crashed with
969 an internal error, complaining that it could not find the referenced
970 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
971 [The bug was that it was starting one character too far in when skipping
972 over the character class, thus treating the ] as data rather than
973 terminating the class. This meant it could skip too much.]
975 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
976 /g option in pcretest when the pattern contains \K, which makes it possible
977 to have an empty string match not at the start, even when the pattern is
978 anchored. Updated pcretest and pcredemo to use this option.
980 21. If the maximum number of capturing subpatterns in a recursion was greater
981 than the maximum at the outer level, the higher number was returned, but
982 with unset values at the outer level. The correct (outer level) value is
983 now given.
985 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
986 PCRE did not set those parentheses (unlike Perl). I have now found a way to
987 make it do so. The string so far is captured, making this feature
988 compatible with Perl.
990 23. The tests have been re-organized, adding tests 11 and 12, to make it
991 possible to check the Perl 5.10 features against Perl 5.10.
993 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
994 pattern matches a fixed length string. PCRE did not allow this; now it
995 does. Neither allows recursion.
997 25. I finally figured out how to implement a request to provide the minimum
998 length of subject string that was needed in order to match a given pattern.
999 (It was back references and recursion that I had previously got hung up
1000 on.) This code has now been added to pcre_study(); it finds a lower bound
1001 to the length of subject needed. It is not necessarily the greatest lower
1002 bound, but using it to avoid searching strings that are too short does give
1003 some useful speed-ups. The value is available to calling programs via
1004 pcre_fullinfo().
1006 26. While implementing 25, I discovered to my embarrassment that pcretest had
1007 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1008 study optimizations had never been tested with that matching function.
1009 Oops. What is worse, even when it was passed study data, there was a bug in
1010 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1011 were also very few tests of studied patterns with pcre_dfa_exec().
1013 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1014 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1015 on the other side of the coin, they are no longer allowed to have different
1016 names, because these cannot be distinguished in PCRE, and this has caused
1017 confusion. (This is a difference from Perl.)
1019 28. When duplicate subpattern names are present (necessarily with different
1020 numbers, as required by 27 above), and a test is made by name in a
1021 conditional pattern, either for a subpattern having been matched, or for
1022 recursion in such a pattern, all the associated numbered subpatterns are
1023 tested, and the overall condition is true if the condition is true for any
1024 one of them. This is the way Perl works, and is also more like the way
1025 testing by number works.
1028 Version 7.9 11-Apr-09
1029 ---------------------
1031 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1032 (pcretest), all targets were linked against these libraries. This included
1033 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1034 libraries. This caused unwanted dependencies to be created. This problem
1035 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1036 pcretest is linked with readline.
1038 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1039 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1040 moved outside it again, because FALSE and TRUE are already defined in AIX,
1041 but BOOL is not.
1043 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1044 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1046 4. The pcregrep documentation said spaces were inserted as well as colons (or
1047 hyphens) following file names and line numbers when outputting matching
1048 lines. This is not true; no spaces are inserted. I have also clarified the
1049 wording for the --colour (or --color) option.
1051 5. In pcregrep, when --colour was used with -o, the list of matching strings
1052 was not coloured; this is different to GNU grep, so I have changed it to be
1053 the same.
1055 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1056 each matching line was coloured. Now it goes on to look for further matches
1057 of any of the test patterns, which is the same behaviour as GNU grep.
1059 7. A pattern that could match an empty string could cause pcregrep to loop; it
1060 doesn't make sense to accept an empty string match in pcregrep, so I have
1061 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1062 seems to be how GNU grep behaves.
1064 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1065 start or after a newline", because the conditional assertion was not being
1066 correctly handled. The rule now is that both the assertion and what follows
1067 in the first alternative must satisfy the test.
1069 9. If auto-callout was enabled in a pattern with a conditional group whose
1070 condition was an assertion, PCRE could crash during matching, both with
1071 pcre_exec() and pcre_dfa_exec().
1073 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1074 used for matching.
1076 11. Unicode property support in character classes was not working for
1077 characters (bytes) greater than 127 when not in UTF-8 mode.
1079 12. Added the -M command line option to pcretest.
1081 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1083 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1085 16. Added comments and documentation about mis-use of no_arg in the C++
1086 wrapper.
1088 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1089 from Martin Jerabek that uses macro names for all relevant character and
1090 string constants.
1092 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1093 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1094 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1095 these, but not everybody uses configure.
1097 19. A conditional group that had only one branch was not being correctly
1098 recognized as an item that could match an empty string. This meant that an
1099 enclosing group might also not be so recognized, causing infinite looping
1100 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1101 with the subject "ab", where knowledge that the repeated group can match
1102 nothing is needed in order to break the loop.
1104 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1105 exec(), but without supplying a callout function, matching went wrong.
1107 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1108 leak if the size of the offset vector was greater than 30. When the vector
1109 is smaller, the saved offsets during recursion go onto a local stack
1110 vector, but for larger vectors malloc() is used. It was failing to free
1111 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1112 error, in fact).
1114 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1115 heapframe that is used only when UTF-8 support is enabled. This caused no
1116 problem, but was untidy.
1118 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1119 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1120 included within another project.
1122 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1123 slightly modified by me:
1125 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1126 not building pcregrep.
1128 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1129 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1131 25. Forward references, both numeric and by name, in patterns that made use of
1132 duplicate group numbers, could behave incorrectly or give incorrect errors,
1133 because when scanning forward to find the reference group, PCRE was not
1134 taking into account the duplicate group numbers. A pattern such as
1135 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1137 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1138 the feature of a custom pointer more persuasive (as requested by a user).
1140 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1141 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1142 pre-defined.
1144 28. Added support for (*UTF8) at the start of a pattern.
1146 29. Arrange for flags added by the "release type" setting in CMake to be shown
1147 in the configuration summary.
1150 Version 7.8 05-Sep-08
1151 ---------------------
1153 1. Replaced UCP searching code with optimized version as implemented for Ad
1154 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1155 stage table and inline lookup instead of a function, giving speed ups of 2
1156 to 5 times on some simple patterns that I tested. Permission was given to
1157 distribute the MultiStage2.py script that generates the tables (it's not in
1158 the tarball, but is in the Subversion repository).
1160 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1161 scripts.
1163 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1164 a group with a zero qualifier. The result of the study could be incorrect,
1165 or the function might crash, depending on the pattern.
1167 4. Caseless matching was not working for non-ASCII characters in back
1168 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1169 It now works when Unicode Property Support is available.
1171 5. In pcretest, an escape such as \x{de} in the data was always generating
1172 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1173 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1174 truncation.
1176 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1178 7. Added two (int) casts to pcregrep when printing the difference of two
1179 pointers, in case they are 64-bit values.
1181 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1182 test 2 if it fails.
1184 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1185 and a #define of that name to empty if it is not externally set. This is to
1186 allow users of MSVC to set it if necessary.
1188 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1189 the convenience functions in the pcre_get.c source file.
1191 11. An option change at the start of a pattern that had top-level alternatives
1192 could cause overwriting and/or a crash. This command provoked a crash in
1193 some environments:
1195 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1197 This potential security problem was recorded as CVE-2008-2371.
1199 12. For a pattern where the match had to start at the beginning or immediately
1200 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1201 pcre_dfa_exec() could read past the end of the passed subject if there was
1202 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1203 pcretest so that it places the subject at the end of its malloc-ed buffer.
1205 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1206 exec() might read past the end of the data buffer in UTF-8 mode.
1208 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1209 the data contained the byte 0x85 as part of a UTF-8 character within its
1210 first line. This applied both to normal and DFA matching.
1212 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1213 /^[^d]*?$/8 failed to match "abc".
1215 16. Added a missing copyright notice to pcrecpp_internal.h.
1217 17. Make it more clear in the documentation that values returned from
1218 pcre_exec() in ovector are byte offsets, not character counts.
1220 18. Tidied a few places to stop certain compilers from issuing warnings.
1222 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1223 supplied by Stefan Weber. I made a further small update for 7.8 because
1224 there is a change of source arrangements: the pcre_searchfuncs.c module is
1225 replaced by pcre_ucd.c.
1228 Version 7.7 07-May-08
1229 ---------------------
1231 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1232 a string to a long long, pretend we don't even have a long long." This is
1233 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1235 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1236 pre-7.6 versions, which defined a global no_arg variable instead of putting
1237 it in the RE class. (See also #8 below.)
1239 3. Remove a line of dead code, identified by coverity and reported by Nuno
1240 Lopes.
1242 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1244 (1) The include/exclude patterns were being applied to the whole pathnames
1245 of files, instead of just to the final components.
1247 (2) If there was more than one level of directory, the subdirectories were
1248 skipped unless they satisfied the include/exclude conditions. This is
1249 inconsistent with GNU grep (and could even be seen as contrary to the
1250 pcregrep specification - which I improved to make it absolutely clear).
1251 The action now is always to scan all levels of directory, and just
1252 apply the include/exclude patterns to regular files.
1254 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1255 --exclude_dir in the tests to avoid scanning .svn directories.
1257 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1258 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1259 doesn't support NULs in patterns.
1261 7. Added some missing "const"s to declarations of static tables in
1262 pcre_compile.c and pcre_dfa_exec.c.
1264 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1265 caused by fix #2 above. (Subsequently also a second patch to fix the
1266 first patch. And a third patch - this was a messy problem.)
1268 9. Applied Craig's patch to remove the use of push_back().
1270 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1271 matching function regexec().
1273 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1274 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1275 references. PCRE supports relative numbers with this syntax (I don't think
1276 Oniguruma does).
1278 12. Previously, a group with a zero repeat such as (...){0} was completely
1279 omitted from the compiled regex. However, this means that if the group
1280 was called as a subroutine from elsewhere in the pattern, things went wrong
1281 (an internal error was given). Such groups are now left in the compiled
1282 pattern, with a new opcode that causes them to be skipped at execution
1283 time.
1285 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1286 to the way PCRE behaves:
1288 (a) A lone ] character is dis-allowed (Perl treats it as data).
1290 (b) A back reference to an unmatched subpattern matches an empty string
1291 (Perl fails the current match path).
1293 (c) A data ] in a character class must be notated as \] because if the
1294 first data character in a class is ], it defines an empty class. (In
1295 Perl it is not possible to have an empty class.) The empty class []
1296 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1297 The negative empty class [^] matches any one character, independently
1298 of the DOTALL setting.
1300 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1301 non-existent subpattern following a character class starting with ']' and
1302 containing () gave an internal compiling error instead of "reference to
1303 non-existent subpattern". Fortunately, when the pattern did exist, the
1304 compiled code was correct. (When scanning forwards to check for the
1305 existencd of the subpattern, it was treating the data ']' as terminating
1306 the class, so got the count wrong. When actually compiling, the reference
1307 was subsequently set up correctly.)
1309 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1310 it was being rejected as not supported by pcre_dfa_exec(), even though
1311 other assertions are supported. I have made pcre_dfa_exec() support
1312 (*FAIL).
1314 16. The implementation of 13c above involved the invention of a new opcode,
1315 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1316 cannot be changed at match time, I realized I could make a small
1317 improvement to matching performance by compiling OP_ALLANY instead of
1318 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1319 on the OP_ANY path.
1321 17. Compiling pcretest on Windows with readline support failed without the
1322 following two fixes: (1) Make the unistd.h include conditional on
1323 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1325 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1326 ncurses library to be included for pcretest when ReadLine support is
1327 requested, but also to allow for it to be overridden. This patch came from
1328 Daniel Bergström.
1330 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1331 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1332 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1333 spotting this.
1336 Version 7.6 28-Jan-08
1337 ---------------------
1339 1. A character class containing a very large number of characters with
1340 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1341 overflow.
1343 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1344 HAVE_LONG_LONG is not defined.
1346 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1347 bring them up to date and include new features. This patch includes:
1349 - Fixed PH's badly added libz and libbz2 support.
1350 - Fixed a problem with static linking.
1351 - Added pcredemo. [But later removed - see 7 below.]
1352 - Fixed dftables problem and added an option.
1353 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1355 - Added readline support for pcretest.
1356 - Added an listing of the option settings after cmake has run.
1358 4. A user submitted a patch to Makefile that makes it easy to create
1359 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1360 Makefile.am that cause it to include this special target, without
1361 affecting anything else. Note that the same mingw target plus all
1362 the other distribution libraries and programs are now supported
1363 when configuring with CMake (see 6 below) instead of with
1364 Configure/Make.
1366 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1367 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1368 exported in the Windows port". It has not yet been confirmed that the patch
1369 solves the problem, but it does no harm.
1371 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1372 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1373 with CMake, and also correct the comment about stack recursion.
1375 7. Remove the automatic building of pcredemo from the ./configure system and
1376 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1377 of a program that users should build themselves after PCRE is installed, so
1378 building it automatically is not really right. What is more, it gave
1379 trouble in some build environments.
1381 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1384 Version 7.5 10-Jan-08
1385 ---------------------
1387 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1388 values in parens when parsing an RE using the C++ wrapper."
1390 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1391 Characters greater than 255 were excluded from the class instead of being
1392 included.
1394 3. The same bug as (2) above applied to negated POSIX classes such as
1395 [:^space:].
1397 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1398 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1399 I have changed it.
1401 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1402 first named subpattern) and a construct such as (?&a) would reference the
1403 first named subpattern whose name started with "a" (in other words, the
1404 length check was missing). Both these problems are fixed. "Subpattern name
1405 expected" is now given for (?&) (a zero-length name), and this patch also
1406 makes it give the same error for \k'' (previously it complained that that
1407 was a reference to a non-existent subpattern).
1409 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1410 this is right because (?- can be followed by option settings as well as by
1411 digits. I have, however, made the messages clearer.
1413 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1414 than the number used in the conditional) now cause a compile-time error.
1415 This is actually not compatible with Perl, which accepts such patterns, but
1416 treats the conditional as always being FALSE (as PCRE used to), but it
1417 seems to me that giving a diagnostic is better.
1419 8. Change "alphameric" to the more common word "alphanumeric" in comments
1420 and messages.
1422 9. Fix two occurrences of "backslash" in comments that should have been
1423 "backspace".
1425 10. Remove two redundant lines of code that can never be obeyed (their function
1426 was moved elsewhere).
1428 11. The program that makes PCRE's Unicode character property table had a bug
1429 which caused it to generate incorrect table entries for sequences of
1430 characters that have the same character type, but are in different scripts.
1431 It amalgamated them into a single range, with the script of the first of
1432 them. In other words, some characters were in the wrong script. There were
1433 thirteen such cases, affecting characters in the following ranges:
1435 U+002b0 - U+002c1
1436 U+0060c - U+0060d
1437 U+0061e - U+00612
1438 U+0064b - U+0065e
1439 U+0074d - U+0076d
1440 U+01800 - U+01805
1441 U+01d00 - U+01d77
1442 U+01d9b - U+01dbf
1443 U+0200b - U+0200f
1444 U+030fc - U+030fe
1445 U+03260 - U+0327f
1446 U+0fb46 - U+0fbb1
1447 U+10450 - U+1049d
1449 12. The -o option (show only the matching part of a line) for pcregrep was not
1450 compatible with GNU grep in that, if there was more than one match in a
1451 line, it showed only the first of them. It now behaves in the same way as
1452 GNU grep.
1454 13. If the -o and -v options were combined for pcregrep, it printed a blank
1455 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1456 does the same. The return code can be used to tell if there were any
1457 non-matching lines.
1459 14. Added --file-offsets and --line-offsets to pcregrep.
1461 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1462 infinitely looping recursion. The bug was that positive lookaheads were not
1463 being skipped when checking for a possible empty match (negative lookaheads
1464 and both kinds of lookbehind were skipped).
1466 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1467 inclusion of <windows.h> to before rather than after the definition of
1468 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1470 17. Specifying a possessive quantifier with a specific limit for a Unicode
1471 character property caused pcre_compile() to compile bad code, which led at
1472 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1473 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1474 caused the error; without that there was no problem.
1476 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1478 19. Added --enable-pcretest-libreadline.
1480 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1481 RE::GlobalReplace(). As a result, the number of replacements returned was
1482 double what it should be. I removed one of the increments, but Craig sent a
1483 later patch that removed the other one (the right fix) and added unit tests
1484 that check the return values (which was not done before).
1486 21. Several CMake things:
1488 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1489 the names libpcre and libpcreposix, not just pcre and pcreposix.
1491 (2) The above change means that pcretest and pcregrep are now correctly
1492 linked with the newly-built libraries, not previously installed ones.
1496 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1497 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1498 UTF-8 newline character). The key issue is that the pattern starts .*;
1499 this means that the match must be either at the beginning, or after a
1500 newline. The bug was in the code for advancing after a failed match and
1501 checking that the new position followed a newline. It was not taking
1502 account of UTF-8 characters correctly.
1504 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1505 character classes. PCRE was not treating the sequence [:...:] as a
1506 character class unless the ... were all letters. Perl, however, seems to
1507 allow any characters between [: and :], though of course it rejects as
1508 unknown any "names" that contain non-letters, because all the known class
1509 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1510 for example, whereas PCRE did not - it did not recognize a POSIX character
1511 class. This seemed a bit dangerous, so the code has been changed to be
1512 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1513 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1514 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1515 Perl does, and where it didn't before.
1517 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1518 Windows environments %n is disabled by default.
1521 Version 7.4 21-Sep-07
1522 ---------------------
1524 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1525 means that a class such as [\s] counted as "explicit reference to CR or
1526 LF". That isn't really right - the whole point of the change was to try to
1527 help when there was an actual mention of one of the two characters. So now
1528 the change happens only if \r or \n (or a literal CR or LF) character is
1529 encountered.
1531 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1532 of both had grown to the point where there were only 3 bits left.
1533 Fortunately, there was spare space in the data structure, and so I have
1534 moved the internal flags into a new 16-bit field to free up more option
1535 bits.
1537 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1538 but did not set the internal JCHANGED flag - either of these is enough to
1539 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1540 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1541 start sets both bits.
1543 4. Added options (at build time, compile time, exec time) to change \R from
1544 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1546 5. doc/pcresyntax.html was missing from the distribution.
1548 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1549 compatibility, even though it is no longer used.
1551 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1552 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1553 windows.h file is present (where different names are used). [This was
1554 reversed later after testing - see 16 below.]
1556 8. Changed all #include <config.h> to #include "config.h". There were also
1557 some further <pcre.h> cases that I changed to "pcre.h".
1559 9. When pcregrep was used with the --colour option, it missed the line ending
1560 sequence off the lines that it output.
1562 10. It was pointed out to me that arrays of string pointers cause lots of
1563 relocations when a shared library is dynamically loaded. A technique of
1564 using a single long string with a table of offsets can drastically reduce
1565 these. I have refactored PCRE in four places to do this. The result is
1566 dramatic:
1568 Originally: 290
1569 After changing UCP table: 187
1570 After changing error message table: 43
1571 After changing table of "verbs" 36
1572 After changing table of Posix names 22
1574 Thanks to the folks working on Gregex for glib for this insight.
1576 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1577 unicode-properties was also set.
1579 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1581 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1582 checked only for CRLF.
1584 14. Added casts to pcretest.c to avoid compiler warnings.
1586 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1588 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1589 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1590 entirely. This removes changes made in 7 above.
1592 17. The CMake files have been updated, and there is now more information about
1593 building with CMake in the NON-UNIX-USE document.
1596 Version 7.3 28-Aug-07
1597 ---------------------
1599 1. In the rejigging of the build system that eventually resulted in 7.1, the
1600 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1601 brackets there is not right, since it causes compilers to look for an
1602 installed pcre.h, not the version that is in the source that is being
1603 compiled (which of course may be different). I have changed it back to:
1605 #include "pcre.h"
1607 I have a vague recollection that the change was concerned with compiling in
1608 different directories, but in the new build system, that is taken care of
1609 by the VPATH setting the Makefile.
1611 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1612 when the subject happened to end in the byte 0x85 (e.g. if the last
1613 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1614 characters but of course it shouldn't be taken as a newline when it is part
1615 of another character. The bug was that, for an unlimited repeat of . in
1616 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1617 characters when looking for a newline.
1619 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1621 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1622 in debug output.
1624 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1625 long printing in the pcrecpp unittest when running under MinGW.
1627 6. ESC_K was left out of the EBCDIC table.
1629 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1630 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1631 limit also applies to "virtual nesting" when a pattern is recursive, and in
1632 this case 1000 isn't so big. I have been able to remove this limit at the
1633 expense of backing off one optimization in certain circumstances. Normally,
1634 when pcre_exec() would call its internal match() function recursively and
1635 immediately return the result unconditionally, it uses a "tail recursion"
1636 feature to save stack. However, when a subpattern that can match an empty
1637 string has an unlimited repetition quantifier, it no longer makes this
1638 optimization. That gives it a stack frame in which to save the data for
1639 checking that an empty string has been matched. Previously this was taken
1640 from the 1000-entry workspace that had been reserved. So now there is no
1641 explicit limit, but more stack is used.
1643 8. Applied Daniel's patches to solve problems with the import/export magic
1644 syntax that is required for Windows, and which was going wrong for the
1645 pcreposix and pcrecpp parts of the library. These were overlooked when this
1646 problem was solved for the main library.
1648 9. There were some crude static tests to avoid integer overflow when computing
1649 the size of patterns that contain repeated groups with explicit upper
1650 limits. As the maximum quantifier is 65535, the maximum group length was
1651 set at 30,000 so that the product of these two numbers did not overflow a
1652 32-bit integer. However, it turns out that people want to use groups that
1653 are longer than 30,000 bytes (though not repeat them that many times).
1654 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1655 made it possible to implement the integer overflow checks in a much more
1656 dynamic way, which I have now done. The artificial limitation on group
1657 length has been removed - we now have only the limit on the total length of
1658 the compiled pattern, which depends on the LINK_SIZE setting.
1660 10. Fixed a bug in the documentation for get/copy named substring when
1661 duplicate names are permitted. If none of the named substrings are set, the
1662 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1663 empty string.
1665 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1666 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1667 because the ] is interpreted as the first data character and the
1668 terminating ] is not found. PCRE has been made compatible with Perl in this
1669 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1670 cause memory overwriting.
1672 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1673 string has been matched (to stop an infinite loop). It was not recognizing
1674 a conditional subpattern that could match an empty string if that
1675 subpattern was within another subpattern. For example, it looped when
1676 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1677 condition was not nested. This bug has been fixed.
1679 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1680 past the start of the subject in the presence of bytes with the top bit
1681 set, for example "\x8aBCD".
1683 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1684 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1686 14. Optimized (?!) to (*FAIL).
1688 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1689 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1690 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1691 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1692 does: it's just the validity check that is more restrictive.
1694 16. Inserted checks for integer overflows during escape sequence (backslash)
1695 processing, and also fixed erroneous offset values for syntax errors during
1696 backslash processing.
1698 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1699 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1701 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1702 caused an overrun.
1704 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1705 something other than just ASCII characters) inside a group that had an
1706 unlimited repeat caused a loop at compile time (while checking to see
1707 whether the group could match an empty string).
1709 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1710 [\P{Any}] did so. (Error in the code for printing property names.)
1712 21. An orphan \E inside a character class could cause a crash.
1714 22. A repeated capturing bracket such as (A)? could cause a wild memory
1715 reference during compilation.
1717 23. There are several functions in pcre_compile() that scan along a compiled
1718 expression for various reasons (e.g. to see if it's fixed length for look
1719 behind). There were bugs in these functions when a repeated \p or \P was
1720 present in the pattern. These operators have additional parameters compared
1721 with \d, etc, and these were not being taken into account when moving along
1722 the compiled data. Specifically:
1724 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1725 length.
1727 (b) An item such as \pL+ within a repeated group could cause crashes or
1728 loops.
1730 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1731 "reference to non-existent subpattern" error.
1733 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1735 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1736 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1738 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1740 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1741 character were causing crashes (broken optimization).
1743 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1744 \p or \P) caused a compile-time loop.
1746 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1747 break. For example, the unstudied pattern [\r\n]A does not match the string
1748 "\r\nA" because change 7.0/46 below moves the current point on by two
1749 characters after failing to match at the start. However, the pattern \nA
1750 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1751 the same is true. There doesn't seem any very clean way out of this, but
1752 what I have chosen to do makes the common cases work: PCRE now takes note
1753 of whether there can be an explicit match for \r or \n anywhere in the
1754 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1755 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1756 pattern has explicit CR or LF references.
1758 29. Added (*CR) etc for changing newline setting at start of pattern.
1761 Version 7.2 19-Jun-07
1762 ---------------------
1764 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1765 which is apparently normally available under Windows.
1767 2. Re-jig the pcregrep tests with different newline settings in an attempt
1768 to make them independent of the local environment's newline setting.
1770 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1772 4. Some of the "internals" tests were previously cut out when the link size
1773 was not 2, because the output contained actual offsets. The recent new
1774 "Z" feature of pcretest means that these can be cut out, making the tests
1775 usable with all link sizes.
1777 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1778 stack recursion. This gives a massive performance boost under BSD, but just
1779 a small improvement under Linux. However, it saves one field in the frame
1780 in all cases.
1782 6. Added more features from the forthcoming Perl 5.10:
1784 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1785 recursion call. It refers to the nth most recently opened parentheses.
1787 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1788 to be opened parentheses.
1790 (c) Conditions that refer to capturing parentheses can be specified
1791 relatively, for example, (?(-2)... or (?(+3)...
1793 (d) \K resets the start of the current match so that everything before
1794 is not part of it.
1796 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1798 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1799 reference syntax.
1801 (g) (?| introduces a group in which the numbering of parentheses in each
1802 alternative starts with the same number.
1804 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1806 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1809 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1810 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1811 for detecting groups that can match an empty string.
1813 9. A pattern with a very large number of alternatives (more than several
1814 hundred) was running out of internal workspace during the pre-compile
1815 phase, where pcre_compile() figures out how much memory will be needed. A
1816 bit of new cunning has reduced the workspace needed for groups with
1817 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1818 workspace instead of running out of the 4096 that are available.
1820 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1822 11. Applied patch from Google to remove an optimization that didn't quite work.
1823 The report of the bug said:
1825 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1826 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1827 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1829 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1830 it matched the wrong number of bytes.
1833 Version 7.1 24-Apr-07
1834 ---------------------
1836 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1837 that is more "standard", making use of automake and other Autotools. There
1838 is some re-arrangement of the files and adjustment of comments consequent
1839 on this.
1841 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1842 for recursive directory scanning broke on some systems because the files
1843 are not scanned in any specific order and on different systems the order
1844 was different. A call to "sort" has been inserted into RunGrepTest for the
1845 approprate test as a short-term fix. In the longer term there may be an
1846 alternative.
1848 3. I had an email from Eric Raymond about problems translating some of PCRE's
1849 man pages to HTML (despite the fact that I distribute HTML pages, some
1850 people do their own conversions for various reasons). The problems
1851 concerned the use of low-level troff macros .br and .in. I have therefore
1852 removed all such uses from the man pages (some were redundant, some could
1853 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1854 HTML has been updated to handle .nf/.fi and to complain if it encounters
1855 .br or .in.
1857 4. Updated comments in configure.ac that get placed in config.h.in and also
1858 arranged for config.h to be included in the distribution, with the name
1859 config.h.generic, for the benefit of those who have to compile without
1860 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1862 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1863 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1864 makevp.bat for the current PCRE, using the additional files
1865 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1867 6. A Windows user reported a minor discrepancy with test 2, which turned out
1868 to be caused by a trailing space on an input line that had got lost in his
1869 copy. The trailing space was an accident, so I've just removed it.
1871 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1872 that is needed.
1874 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1875 as "const" (a) because they are and (b) because it helps the PHP
1876 maintainers who have recently made a script to detect big data structures
1877 in the php code that should be moved to the .rodata section. I remembered
1878 to update Builducptable as well, so it won't revert if ucptable.h is ever
1879 re-created.
1881 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1882 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1883 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1884 support is not required. This saves 1.5-2K of code, which is important in
1885 some applications.
1887 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1888 so as not to refer to the tables, even though these functions will never be
1889 called when UTF-8 support is disabled. Otherwise there are problems with a
1890 shared library.
1892 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1894 (a) It was defining its arguments as char * instead of void *.
1896 (b) It was assuming that all moves were upwards in memory; this was true
1897 a long time ago when I wrote it, but is no longer the case.
1899 The emulated memove() is provided for those environments that have neither
1900 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1901 is clearly not the case, as these two bugs were recently reported.
1903 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1904 and Detrail to create the HTML documentation, the .txt form of the man
1905 pages, and it removes trailing spaces from listed files. It also creates
1906 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1907 case, it wraps all the #defines with #ifndefs. This script should be run
1908 before "make dist".
1910 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1911 with Unicode property support.
1913 (a) For a maximizing quantifier, if the two different cases of the
1914 character were of different lengths in their UTF-8 codings (there are
1915 some cases like this - I found 11), and the matching function had to
1916 back up over a mixture of the two cases, it incorrectly assumed they
1917 were both the same length.
1919 (b) When PCRE was configured to use the heap rather than the stack for
1920 recursion during matching, it was not correctly preserving the data for
1921 the other case of a UTF-8 character when checking ahead for a match
1922 while processing a minimizing repeat. If the check also involved
1923 matching a wide character, but failed, corruption could cause an
1924 erroneous result when trying to check for a repeat of the original
1925 character.
1927 13. Some tidying changes to the testing mechanism:
1929 (a) The RunTest script now detects the internal link size and whether there
1930 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1931 values substituted by "configure". (The RunGrepTest script already did
1932 this for UTF-8.) The configure.ac script no longer substitutes the
1933 relevant variables.
1935 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1936 with length and offset values. This means that the output is different
1937 for different internal link sizes. Test 2 is skipped for link sizes
1938 other than 2 because of this, bypassing the problem. Unfortunately,
1939 there was also a test in test 3 (the locale tests) that used /B and
1940 failed for link sizes other than 2. Rather than cut the whole test out,
1941 I have added a new /Z option to pcretest that replaces the length and
1942 offset values with spaces. This is now used to make test 3 independent
1943 of link size. (Test 2 will be tidied up later.)
1945 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1946 segmentation fault instead of returning the appropriate error message.
1948 15. In multiline mode when the newline sequence was set to "any", the pattern
1949 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1950 This doesn't seem right; it now treats the CRLF combination as the line
1951 ending, and so does not match in that case. It's only a pattern such as ^$
1952 that would hit this one: something like ^ABC$ would have failed after \r
1953 and then tried again after \r\n.
1955 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1956 in an attempt to make files that differ only in their line terminators
1957 compare equal. This works on Linux.
1959 17. Under certain error circumstances pcregrep might try to free random memory
1960 as it exited. This is now fixed, thanks to valgrind.
1962 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1963 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1964 was because its rules for how to advance for /g after matching an empty
1965 string at the end of a line did not allow for this case. They now check for
1966 it specially.
1968 20. pcretest is supposed to handle patterns and data of any length, by
1969 extending its buffers when necessary. It was getting this wrong when the
1970 buffer for a data line had to be extended.
1972 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1973 CRLF as a newline sequence.
1975 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1976 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1977 I have nevertheless tidied it up.
1979 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1981 24. Added a man page for pcre-config.
1984 Version 7.0 19-Dec-06
1985 ---------------------
1987 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1988 moving to gcc 4.1.1.
1990 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1991 sys/time.h, which is documented as needed for this function. It doesn't
1992 seem to matter on Linux, but it showed up on some releases of OS X.
1994 3. It seems that there are systems where bytes whose values are greater than
1995 127 match isprint() in the "C" locale. The "C" locale should be the
1996 default when a C program starts up. In most systems, only ASCII printing
1997 characters match isprint(). This difference caused the output from pcretest
1998 to vary, making some of the tests fail. I have changed pcretest so that:
2000 (a) When it is outputting text in the compiled version of a pattern, bytes
2001 other than 32-126 are always shown as hex escapes.
2003 (b) When it is outputting text that is a matched part of a subject string,
2004 it does the same, unless a different locale has been set for the match
2005 (using the /L modifier). In this case, it uses isprint() to decide.
2007 4. Fixed a major bug that caused incorrect computation of the amount of memory
2008 required for a compiled pattern when options that changed within the
2009 pattern affected the logic of the preliminary scan that determines the
2010 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2011 that the computed length was too small. The symptoms of this bug were
2012 either the PCRE error "internal error: code overflow" from pcre_compile(),
2013 or a glibc crash with a message such as "pcretest: free(): invalid next
2014 size (fast)". Examples of patterns that provoked this bug (shown in
2015 pcretest format) are:
2017 /(?-x: )/x
2018 /(?x)(?-x: \s*#\s*)/
2019 /((?i)[\x{c0}])/8
2020 /(?i:[\x{c0}])/8
2022 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2023 is now done differently.
2025 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2026 wrapper classes; (b) implement a new function in the C++ scanner that is
2027 more efficient than the old way of doing things because it avoids levels of
2028 recursion in the regex matching; (c) add a paragraph to the documentation
2029 for the FullMatch() function.
2031 6. The escape sequence \n was being treated as whatever was defined as
2032 "newline". Not only was this contrary to the documentation, which states
2033 that \n is character 10 (hex 0A), but it also went horribly wrong when
2034 "newline" was defined as CRLF. This has been fixed.
2036 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2037 was being set to -1 for the "end of line" case (supposedly a value that no
2038 character can have). Though this value is never used (the check for end of
2039 line is "zero bytes in current character"), it caused compiler complaints.
2040 I've changed it to 0xffffffff.
2042 8. In pcre_version.c, the version string was being built by a sequence of
2043 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2044 string (as it is for production releases) called a macro with an empty
2045 argument. The C standard says the result of this is undefined. The gcc
2046 compiler treats it as an empty string (which was what was wanted) but it is
2047 reported that Visual C gives an error. The source has been hacked around to
2048 avoid this problem.
2050 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2051 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2052 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2053 of them did).
2055 10. Originally, pcretest opened its input and output without "b"; then I was
2056 told that "b" was needed in some environments, so it was added for release
2057 5.0 to both the input and output. (It makes no difference on Unix-like
2058 systems.) Later I was told that it is wrong for the input on Windows. I've
2059 now abstracted the modes into two macros, to make it easier to fiddle with
2060 them, and removed "b" from the input mode under Windows.
2062 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2064 12. Added -help and --help to pcretest as an official way of being reminded
2065 of the options.
2067 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2068 and pcrecpp.cc because they annoy compilers at high warning levels.
2070 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2072 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2073 scripts are not C programs :-) and which was not noticed because it works
2074 on Linux.
2076 16. pcretest is supposed to handle any length of pattern and data line (as one
2077 line or as a continued sequence of lines) by extending its input buffer if
2078 necessary. This feature was broken for very long pattern lines, leading to
2079 a string of junk being passed to pcre_compile() if the pattern was longer
2080 than about 50K.
2082 17. I have done a major re-factoring of the way pcre_compile() computes the
2083 amount of memory needed for a compiled pattern. Previously, there was code
2084 that made a preliminary scan of the pattern in order to do this. That was
2085 OK when PCRE was new, but as the facilities have expanded, it has become
2086 harder and harder to keep it in step with the real compile phase, and there
2087 have been a number of bugs (see for example, 4 above). I have now found a
2088 cunning way of running the real compile function in a "fake" mode that
2089 enables it to compute how much memory it would need, while actually only
2090 ever using a few hundred bytes of working memory and without too many
2091 tests of the mode. This should make future maintenance and development
2092 easier. A side effect of this work is that the limit of 200 on the nesting
2093 depth of parentheses has been removed (though this was never a serious
2094 limitation, I suspect). However, there is a downside: pcre_compile() now
2095 runs more slowly than before (30% or more, depending on the pattern). I
2096 hope this isn't a big issue. There is no effect on runtime performance.
2098 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2099 newline (only possible for the last line of a file) and it was a
2100 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2102 19. Added additional timing features to pcretest. (1) The -tm option now times
2103 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2104 separate command line item, by a number that specifies the number of
2105 repeats to use when timing. The default is 50000; this gives better
2106 precision, but takes uncomfortably long for very large patterns.
2108 20. Extended pcre_study() to be more clever in cases where a branch of a
2109 subpattern has no definite first character. For example, (a*|b*)[cd] would
2110 previously give no result from pcre_study(). Now it recognizes that the
2111 first character must be a, b, c, or d.
2113 21. There was an incorrect error "recursive call could loop indefinitely" if
2114 a subpattern (or the entire pattern) that was being tested for matching an
2115 empty string contained only one non-empty item after a nested subpattern.
2116 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2117 incorrectly, because the \d was being skipped in the check.
2119 22. The pcretest program now has a new pattern option /B and a command line
2120 option -b, which is equivalent to adding /B to every pattern. This causes
2121 it to show the compiled bytecode, without the additional information that
2122 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2123 is the same as /B/I).
2125 23. A new optimization is now able automatically to treat some sequences such
2126 as a*b as a*+b. More specifically, if something simple (such as a character
2127 or a simple class like \d) has an unlimited quantifier, and is followed by
2128 something that cannot possibly match the quantified thing, the quantifier
2129 is automatically "possessified".
2131 24. A recursive reference to a subpattern whose number was greater than 39
2132 went wrong under certain circumstances in UTF-8 mode. This bug could also
2133 have affected the operation of pcre_study().
2135 25. Realized that a little bit of performance could be had by replacing
2136 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2138 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2140 27. Possessive quantifiers such as a++ were previously implemented by turning
2141 them into atomic groups such as ($>a+). Now they have their own opcodes,
2142 which improves performance. This includes the automatically created ones
2143 from 23 above.
2145 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2146 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2147 the first matched character to be a colon. This applied both to named and
2148 numbered groups.
2150 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2152 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2153 building PCRE on Windows easier, so I have included it in the distribution.
2155 31. There is now a check in pcretest against a ridiculously large number being
2156 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2157 loop, the loop is abandoned.
2159 32. Forward references to subpatterns in conditions such as (?(2)...) where
2160 subpattern 2 is defined later cause pcre_compile() to search forwards in
2161 the pattern for the relevant set of parentheses. This search went wrong
2162 when there were unescaped parentheses in a character class, parentheses
2163 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2165 33. "Subroutine" calls and backreferences were previously restricted to
2166 referencing subpatterns earlier in the regex. This restriction has now
2167 been removed.
2169 34. Added a number of extra features that are going to be in Perl 5.10. On the
2170 whole, these are just syntactic alternatives for features that PCRE had
2171 previously implemented using the Python syntax or my own invention. The
2172 other formats are all retained for compatibility.
2174 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2175 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2176 also .NET compatible.
2178 (b) A recursion or subroutine call to a named group can now be defined as
2179 (?&name) as well as (?P>name).
2181 (c) A backreference to a named group can now be defined as \k<name> or
2182 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2183 5.10, are also .NET compatible.
2185 (d) A conditional reference to a named group can now use the syntax
2186 (?(<name>) or (?('name') as well as (?(name).
2188 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2189 groups (named and numbered) that are never evaluated inline, but can be
2190 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2191 is always false. There may be only one alternative in such a group.
2193 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2194 as the simple (?(R). The condition is true only if the most recent
2195 recursion is that of the given number or name. It does not search out
2196 through the entire recursion stack.
2198 (g) The escape \gN or \g{N} has been added, where N is a positive or
2199 negative number, specifying an absolute or relative reference.
2201 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2202 some "unreachable code" warnings.
2204 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2205 things, this adds five new scripts.
2207 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2208 There were also incompatibilities regarding the handling of \Q..\E inside
2209 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2210 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2212 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2213 matches an empty string, and forcibly breaks the loop. There were bugs in
2214 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2215 against aaaa the result was just "a" rather than "aaaa", for example. Two
2216 separate and independent bugs (that affected different cases) have been
2217 fixed.
2219 39. Refactored the code to abolish the use of different opcodes for small
2220 capturing bracket numbers. This is a tidy that I avoided doing when I
2221 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2222 The new approach is not only tidier, it makes it possible to reduce the
2223 memory needed to fix the previous bug (38).
2225 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2226 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2227 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2228 mode.
2230 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2231 report.
2233 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2234 copy construction and assignment in the C++ wrapper.
2236 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2237 couple of bugs that could have given wrong results in the "--newline=crlf"
2238 case.
2240 44. Added a number of casts and did some reorganization of signed/unsigned int
2241 variables following suggestions from Dair Grant. Also renamed the variable
2242 "this" as "item" because it is a C++ keyword.
2244 45. Arranged for dftables to add
2246 #include "pcre_internal.h"
2248 to pcre_chartables.c because without it, gcc 4.x may remove the array
2249 definition from the final binary if PCRE is built into a static library and
2250 dead code stripping is activated.
2252 46. For an unanchored pattern, if a match attempt fails at the start of a
2253 newline sequence, and the newline setting is CRLF or ANY, and the next two
2254 characters are CRLF, advance by two characters instead of one.
2257 Version 6.7 04-Jul-06
2258 ---------------------
2260 1. In order to handle tests when input lines are enormously long, pcretest has
2261 been re-factored so that it automatically extends its buffers when
2262 necessary. The code is crude, but this _is_ just a test program. The
2263 default size has been increased from 32K to 50K.
2265 2. The code in pcre_study() was using the value of the re argument before
2266 testing it for NULL. (Of course, in any sensible call of the function, it
2267 won't be NULL.)
2269 3. The memmove() emulation function in pcre_internal.h, which is used on
2270 systems that lack both memmove() and bcopy() - that is, hardly ever -
2271 was missing a "static" storage class specifier.
2273 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2274 containing an extended class (one that cannot be represented by a bitmap
2275 because it contains high-valued characters or Unicode property items, e.g.
2276 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2277 pattern, but PCRE should not loop if you do not (it no longer does).
2278 [Detail: two cases were found: (a) a repeated subpattern containing an
2279 extended class; (b) a recursive reference to a subpattern that followed a
2280 previous extended class. It wasn't skipping over the extended class
2281 correctly when UTF-8 mode was not set.]
2283 5. A negated single-character class was not being recognized as fixed-length
2284 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2285 compile error "lookbehind assertion is not fixed length".
2287 6. The RunPerlTest auxiliary script was showing an unexpected difference
2288 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2289 write a Perl script that can interpret lines of an input file either as
2290 byte characters or as UTF-8, which is what "perltest" was being required to
2291 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2292 can't do is switch easily at run time between having the "use utf8;" pragma
2293 or not. In the end, I fudged it by using the RunPerlTest script to insert
2294 "use utf8;" explicitly for the UTF-8 tests.
2296 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2297 the end of the subject string, contrary to the documentation and to what
2298 Perl does. This was true of both matching functions. Now it matches only at
2299 the start of the subject and immediately after *internal* newlines.
2301 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2302 a pointer to an int instead of a pointer to an unsigned long int. This
2303 caused problems on 64-bit systems.
2305 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2306 instance of the 'standard' template library not being so standard".
2308 10. There was no check on the number of named subpatterns nor the maximum
2309 length of a subpattern name. The product of these values is used to compute
2310 the size of the memory block for a compiled pattern. By supplying a very
2311 long subpattern name and a large number of named subpatterns, the size
2312 computation could be caused to overflow. This is now prevented by limiting
2313 the length of names to 32 characters, and the number of named subpatterns
2314 to 10,000.
2316 11. Subpatterns that are repeated with specific counts have to be replicated in
2317 the compiled pattern. The size of memory for this was computed from the
2318 length of the subpattern and the repeat count. The latter is limited to
2319 65535, but there was no limit on the former, meaning that integer overflow
2320 could in principle occur. The compiled length of a repeated subpattern is
2321 now limited to 30,000 bytes in order to prevent this.
2323 12. Added the optional facility to have named substrings with the same name.
2325 13. Added the ability to use a named substring as a condition, using the
2326 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2327 are numbers (not recommended). Forward references are permitted.
2329 14. Added forward references in named backreferences (if you see what I mean).
2331 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2332 pattern could run off the end of the subject. For example, the pattern
2333 "(?s)(.{1,5})"8 did this with the subject "ab".
2335 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2336 PCRE_CASELESS was set when matching characters that were quantified with ?
2337 or *.
2339 17. A character class other than a single negated character that had a minimum
2340 but no maximum quantifier - for example [ab]{6,} - was not handled
2341 correctly by pce_dfa_exec(). It would match only one character.
2343 18. A valid (though odd) pattern that looked like a POSIX character
2344 class but used an invalid character after [ (for example [[,abc,]]) caused
2345 pcre_compile() to give the error "Failed: internal error: code overflow" or
2346 in some cases to crash with a glibc free() error. This could even happen if
2347 the pattern terminated after [[ but there just happened to be a sequence of
2348 letters, a binary zero, and a closing ] in the memory that followed.
2350 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2351 over the years. Originally (before any Unicode support), just the bottom 8
2352 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2353 output from "man perlunicode" includes this:
2355 The regular expression compiler produces polymorphic opcodes. That
2356 is, the pattern adapts to the data and automatically switches to
2357 the Unicode character scheme when presented with Unicode data--or
2358 instead uses a traditional byte scheme when presented with byte
2359 data.
2361 Sadly, a wide octal escape does not cause a switch, and in a string with
2362 no other multibyte characters, these octal escapes are treated as before.
2363 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2364 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2365 Unicode string.
2367 I have not perpetrated such confusion in PCRE. Up till now, it took just
2368 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2369 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2370 translate to the appropriate multibyte character.
2372 29. Applied some refactoring to reduce the number of warnings from Microsoft
2373 and Borland compilers. This has included removing the fudge introduced
2374 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2375 a warning about an unused variable.
2377 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2378 characters since release 4.0, because Perl (from release 5.004) does not.
2379 [Or at least, is documented not to: some releases seem to be in conflict
2380 with the documentation.] However, when a pattern was studied with
2381 pcre_study() and all its branches started with \s, PCRE still included VT
2382 as a possible starting character. Of course, this did no harm; it just
2383 caused an unnecessary match attempt.
2385 22. Removed a now-redundant internal flag bit that recorded the fact that case
2386 dependency changed within the pattern. This was once needed for "required
2387 byte" processing, but is no longer used. This recovers a now-scarce options
2388 bit. Also moved the least significant internal flag bit to the most-
2389 significant bit of the word, which was not previously used (hangover from
2390 the days when it was an int rather than a uint) to free up another bit for
2391 the future.
2393 23. Added support for CRLF line endings as well as CR and LF. As well as the
2394 default being selectable at build time, it can now be changed at runtime
2395 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2396 specify that it is scanning data with non-default line endings.
2398 24. Changed the definition of CXXLINK to make it agree with the definition of
2399 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2401 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2402 recursions. This makes a big different to stack usage for some patterns.
2404 26. If a subpattern containing a named recursion or subroutine reference such
2405 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2406 the space required for the compiled pattern went wrong and gave too small a
2407 value. Depending on the environment, this could lead to "Failed: internal
2408 error: code overflow at offset 49" or "glibc detected double free or
2409 corruption" errors.
2411 27. Applied patches from Google (a) to support the new newline modes and (b) to
2412 advance over multibyte UTF-8 characters in GlobalReplace.
2414 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2415 difference for some implementation of PCRE in some Windows version.
2417 29. Added some extra testing facilities to pcretest:
2419 \q<number> in a data line sets the "match limit" value
2420 \Q<number> in a data line sets the "match recursion limt" value
2421 -S <number> sets the stack size, where <number> is in megabytes
2423 The -S option isn't available for Windows.
2426 Version 6.6 06-Feb-06
2427 ---------------------
2429 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2430 in pcreposix.h. I have copied the definition from pcre.h.
2432 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2433 because pcre.h is no longer a built file.
2435 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2436 not normally included in the compiled code.
2439 Version 6.5 01-Feb-06
2440 ---------------------
2442 1. When using the partial match feature with pcre_dfa_exec(), it was not
2443 anchoring the second and subsequent partial matches at the new starting
2444 point. This could lead to incorrect results. For example, with the pattern
2445 /1234/, partially matching against "123" and then "a4" gave a match.
2447 2. Changes to pcregrep:
2449 (a) All non-match returns from pcre_exec() were being treated as failures
2450 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2451 error message is output. Some extra information is given for the
2453 probably the only errors that are likely to be caused by users (by
2454 specifying a regex that has nested indefinite repeats, for instance).
2455 If there are more than 20 of these errors, pcregrep is abandoned.
2457 (b) A binary zero was treated as data while matching, but terminated the
2458 output line if it was written out. This has been fixed: binary zeroes
2459 are now no different to any other data bytes.
2461 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2462 used to set a locale for matching. The --locale=xxxx long option has
2463 been added (no short equivalent) to specify a locale explicitly on the
2464 pcregrep command, overriding the environment variables.
2466 (d) When -B was used with -n, some line numbers in the output were one less
2467 than they should have been.
2469 (e) Added the -o (--only-matching) option.
2471 (f) If -A or -C was used with -c (count only), some lines of context were
2472 accidentally printed for the final match.
2474 (g) Added the -H (--with-filename) option.
2476 (h) The combination of options -rh failed to suppress file names for files
2477 that were found from directory arguments.
2479 (i) Added the -D (--devices) and -d (--directories) options.
2481 (j) Added the -F (--fixed-strings) option.
2483 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2485 (l) Added the --colo(u)r option.
2487 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2488 is not present by default.
2490 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2491 items such as (?R) or (?1), when the recursion could match a number of
2492 alternatives. If it matched one of the alternatives, but subsequently,
2493 outside the recursion, there was a failure, the code tried to back up into
2494 the recursion. However, because of the way PCRE is implemented, this is not
2495 possible, and the result was an incorrect result from the match.
2497 In order to prevent this happening, the specification of recursion has
2498 been changed so that all such subpatterns are automatically treated as
2499 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2501 4. I had overlooked the fact that, in some locales, there are characters for
2502 which isalpha() is true but neither isupper() nor islower() are true. In
2503 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2504 and ordfeminine) are like this. This affected the treatment of \w and \W
2505 when they appeared in character classes, but not when they appeared outside
2506 a character class. The bit map for "word" characters is now created
2507 separately from the results of isalnum() instead of just taking it from the
2508 upper, lower, and digit maps. (Plus the underscore character, of course.)
2510 5. The above bug also affected the handling of POSIX character classes such as
2511 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2512 permanent tables. Instead, the bit maps for such a class were previously
2513 created as the appropriate unions of the upper, lower, and digit bitmaps.
2514 Now they are created by subtraction from the [[:word:]] class, which has
2515 its own bitmap.
2517 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2518 It is created by subtracting the vertical space characters (\x09, \x0a,
2519 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2520 subtraction was done in the overall bitmap for a character class, meaning
2521 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2522 be recognized. This bug has been fixed.
2524 7. Patches from the folks at Google:
2526 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2527 real life, but is still worth protecting against".
2529 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2530 regular expressions".
2532 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2533 have it.
2535 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2536 "configure" and the latter not, in order to fix a problem somebody had
2537 with compiling the Arg class on HP-UX.
2539 (e) Improve the error-handling of the C++ wrapper a little bit.
2541 (f) New tests for checking recursion limiting.
2543 8. The pcre_memmove() function, which is used only if the environment does not
2544 have a standard memmove() function (and is therefore rarely compiled),
2545 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2546 returning a result (though PCRE never actually uses the result).
2548 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2549 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2550 returned instead of calling malloc() with an overflowing number that would
2551 most likely cause subsequent chaos.
2553 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2555 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2556 with this option is matched, the nmatch and pmatch options of regexec() are
2557 ignored.
2559 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2560 provided in case anyone wants to the the POSIX interface with UTF-8
2561 strings.
2563 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2564 C++ linking (needed for some HP-UX environments).
2566 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2567 (unused parameter) and in the pcre_printint() function (omitted "default"
2568 switch label when the default is to do nothing).
2570 15. Added some code to make it possible, when PCRE is compiled as a C++
2571 library, to replace subject pointers for pcre_exec() with a smart pointer
2572 class, thus making it possible to process discontinuous strings.
2574 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2575 much the same function. They were added by different people who were trying
2576 to make PCRE easy to compile on non-Unix systems. It has been suggested
2577 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2578 for compiling on Windows systems. I have therefore replaced it with
2579 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2580 defaults to "extern" for C or "extern C" for C++, which works fine on
2581 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2582 SCOPE with something explicit in config.h. In addition:
2584 (a) pcreposix.h still had just "extern" instead of either of these macros;
2585 I have replaced it with PCRE_DATA_SCOPE.
2587 (b) Functions such as _pcre_xclass(), which are internal to the library,
2588 but external in the C sense, all had PCRE_EXPORT in their definitions.
2589 This is apparently wrong for the Windows case, so I have removed it.
2590 (It makes no difference on Unix-like systems.)
2592 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2593 of recursive calls to match(). This is different to MATCH_LIMIT because
2594 that limits the total number of calls to match(), not all of which increase
2595 the depth of recursion. Limiting the recursion depth limits the amount of
2596 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2597 when PCRE is compiled, and changed at run time. A patch from Google adds
2598 this functionality to the C++ interface.
2600 18. Changes to the handling of Unicode character properties:
2602 (a) Updated the table to Unicode 4.1.0.
2604 (b) Recognize characters that are not in the table as "Cn" (undefined).
2606 (c) I revised the way the table is implemented to a much improved format
2607 which includes recognition of ranges. It now supports the ranges that
2608 are defined in UnicodeData.txt, and it also amalgamates other
2609 characters into ranges. This has reduced the number of entries in the
2610 table from around 16,000 to around 3,000, thus reducing its size
2611 considerably. I realized I did not need to use a tree structure after
2612 all - a binary chop search is just as efficient. Having reduced the
2613 number of entries, I extended their size from 6 bytes to 8 bytes to
2614 allow for more data.
2616 (d) Added support for Unicode script names via properties such as \p{Han}.
2618 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2619 matching that character.
2621 20. When matching a repeated Unicode property with a minimum greater than zero,
2622 (for example \pL{2,}), PCRE could look past the end of the subject if it
2623 reached it while seeking the minimum number of characters. This could
2624 happen only if some of the characters were more than one byte long, because
2625 there is a check for at least the minimum number of bytes.
2627 21. Refactored the implementation of \p and \P so as to be more general, to
2628 allow for more different types of property in future. This has changed the
2629 compiled form incompatibly. Anybody with saved compiled patterns that use
2630 \p or \P will have to recompile them.
2632 22. Added "Any" and "L&" to the supported property types.
2634 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2635 but give a compile time error if the value is greater than 0xff.
2637 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2638 accidentally not being installed or uninstalled.
2640 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2641 made were to insert the current release number. This seemed silly, because
2642 it made things harder for people building PCRE on systems that don't run
2643 "configure". I have turned pcre.h into a distributed file, no longer built
2644 by "configure", with the version identification directly included. There is
2645 no longer a pcre.h.in file.
2647 However, this change necessitated a change to the pcre-config script as
2648 well. It is built from pcre-config.in, and one of the substitutions was the
2649 release number. I have updated configure.ac so that ./configure now finds
2650 the release number by grepping pcre.h.
2652 26. Added the ability to run the tests under valgrind.
2655 Version 6.4 05-Sep-05
2656 ---------------------
2658 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2659 "--" to be printed when multiple files were scanned, even when none of the
2660 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2661 consider it to be a bug, and have restored the previous behaviour.
2663 2. A couple of code tidies to get rid of compiler warnings.
2665 3. The pcretest program used to cheat by referring to symbols in the library
2666 whose names begin with _pcre_. These are internal symbols that are not
2667 really supposed to be visible externally, and in some environments it is
2668 possible to suppress them. The cheating is now confined to including
2669 certain files from the library's source, which is a bit cleaner.
2671 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2672 file's purpose clearer.
2674 5. Reorganized pcre_ucp_findchar().
2677 Version 6.3 15-Aug-05
2678 ---------------------
2680 1. The file libpcre.pc.in did not have general read permission in the tarball.
2682 2. There were some problems when building without C++ support:
2684 (a) If C++ support was not built, "make install" and "make test" still
2685 tried to test it.
2687 (b) There were problems when the value of CXX was explicitly set. Some
2688 changes have been made to try to fix these, and ...
2690 (c) --disable-cpp can now be used to explicitly disable C++ support.
2692 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2693 backslash in a target when C++ was disabled. This confuses some
2694 versions of "make", apparently. Using an intermediate variable solves
2695 this. (Same for CPP_LOBJ.)
2697 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2698 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2699 necessary on certain architectures.
2701 4. Added a setting of -export-symbols-regex to the link command to remove
2702 those symbols that are exported in the C sense, but actually are local
2703 within the library, and not documented. Their names all begin with
2704 "_pcre_". This is not a perfect job, because (a) we have to except some
2705 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2706 available (and never for static libraries). I have made a note to try to
2707 find a way round (a) in the future.
2710 Version 6.2 01-Aug-05
2711 ---------------------
2713 1. There was no test for integer overflow of quantifier values. A construction
2714 such as {1111111111111111} would give undefined results. What is worse, if
2715 a minimum quantifier for a parenthesized subpattern overflowed and became
2716 negative, the calculation of the memory size went wrong. This could have
2717 led to memory overwriting.
2719 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2721 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2722 operating environments where this matters.
2724 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2725 PCRE options from within the C++ wrapper.
2727 5. Named capturing subpatterns were not being correctly counted when a pattern
2728 was compiled. This caused two problems: (a) If there were more than 100
2729 such subpatterns, the calculation of the memory needed for the whole
2730 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2731 back references of the form \12, where the number was greater than 9, were
2732 not recognized as back references, even though there were sufficient
2733 previous subpatterns.
2735 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2736 versions of gcc, e.g. 2.95.4.
2739 Version 6.1 21-Jun-05
2740 ---------------------
2742 1. There was one reference to the variable "posix" in pcretest.c that was not
2743 surrounded by "#if !defined NOPOSIX".
2745 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2746 the cross-check on the old pcre_info() function, for the benefit of the
2747 cut-down version of PCRE that is currently imported into Exim.
2749 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2750 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2751 patterns that start (?i) but I don't think that matters. The use of (?i) is
2752 just an example; this all applies to the other options as well.
2754 4. Since libtool seems to echo the compile commands it is issuing, the output
2755 from "make" can be reduced a bit by putting "@" in front of each libtool
2756 compile command.
2758 5. Patch from the folks at Google for configure.in to be a bit more thorough
2759 in checking for a suitable C++ installation before trying to compile the
2760 C++ stuff. This should fix a reported problem when a compiler was present,
2761 but no suitable headers.
2763 6. The man pages all had just "PCRE" as their title. I have changed them to
2764 be the relevant file name. I have also arranged that these names are
2765 retained in the file doc/pcre.txt, which is a concatenation in text format
2766 of all the man pages except the little individual ones for each function.
2768 7. The NON-UNIX-USE file had not been updated for the different set of source
2769 files that come with release 6. I also added a few comments about the C++
2770 wrapper.
2773 Version 6.0 07-Jun-05
2774 ---------------------
2776 1. Some minor internal re-organization to help with my DFA experiments.
2778 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2779 didn't matter for the library itself when fully configured, but did matter
2780 when compiling without UCP support, or within Exim, where the ucp files are
2781 not imported.
2783 3. Refactoring of the library code to split up the various functions into
2784 different source modules. The addition of the new DFA matching code (see
2785 below) to a single monolithic source would have made it really too
2786 unwieldy, quite apart from causing all the code to be include in a
2787 statically linked application, when only some functions are used. This is
2788 relevant even without the DFA addition now that patterns can be compiled in
2789 one application and matched in another.
2791 The downside of splitting up is that there have to be some external
2792 functions and data tables that are used internally in different modules of
2793 the library but which are not part of the API. These have all had their
2794 names changed to start with "_pcre_" so that they are unlikely to clash
2795 with other external names.
2797 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2798 a different (DFA) algorithm. Although it is slower than the original
2799 function, it does have some advantages for certain types of matching
2800 problem.
2802 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2803 including restarting after a partial match.
2805 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2806 defined when compiling for Windows was sent to me. I have put it into the
2807 code, though I have no means of testing or verifying it.
2809 7. Added the pcre_refcount() auxiliary function.
2811 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2812 match before or at the first newline in the subject string. In pcretest,
2813 the /f option on a pattern can be used to set this.
2815 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2816 would behave wrongly. This has been present in PCRE since release 4.0.
2818 10. A number of changes to the pcregrep command:
2820 (a) Refactored how -x works; insert ^(...)$ instead of setting
2821 PCRE_ANCHORED and checking the length, in preparation for adding
2822 something similar for -w.
2824 (b) Added the -w (match as a word) option.
2826 (c) Refactored the way lines are read and buffered so as to have more
2827 than one at a time available.
2829 (d) Implemented a pcregrep test script.
2831 (e) Added the -M (multiline match) option. This allows patterns to match
2832 over several lines of the subject. The buffering ensures that at least
2833 8K, or the rest of the document (whichever is the shorter) is available
2834 for matching (and similarly the previous 8K for lookbehind assertions).
2836 (f) Changed the --help output so that it now says
2838 -w, --word-regex(p)
2840 instead of two lines, one with "regex" and the other with "regexp"
2841 because that confused at least one person since the short forms are the
2842 same. (This required a bit of code, as the output is generated
2843 automatically from a table. It wasn't just a text change.)
2845 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2846 option but starts with a hyphen. Could be a pattern or a path name
2847 starting with a hyphen, for instance.
2849 (h) "-" can be given as a file name to represent stdin.
2851 (i) When file names are being printed, "(standard input)" is used for
2852 the standard input, for compatibility with GNU grep. Previously
2853 "<stdin>" was used.
2855 (j) The option --label=xxx can be used to supply a name to be used for
2856 stdin when file names are being printed. There is no short form.
2858 (k) Re-factored the options decoding logic because we are going to add
2859 two more options that take data. Such options can now be given in four
2860 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2862 (l) Added the -A, -B, and -C options for requesting that lines of context
2863 around matches be printed.
2865 (m) Added the -L option to print the names of files that do not contain
2866 any matching lines, that is, the complement of -l.
2868 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2869 continue to scan other files.
2871 (o) The -s option was incorrectly implemented. For compatibility with other
2872 greps, it now suppresses the error message for a non-existent or non-
2873 accessible file (but not the return code). There is a new option called
2874 -q that suppresses the output of matching lines, which was what -s was
2875 previously doing.
2877 (p) Added --include and --exclude options to specify files for inclusion
2878 and exclusion when recursing.
2880 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2881 Hopefully, it now does.
2883 12. Missing cast in pcre_study().
2885 13. Added an "uninstall" target to the makefile.
2887 14. Replaced "extern" in the function prototypes in Makefile.in with
2888 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2889 world, but is set differently for Windows.
2891 15. Added a second compiling function called pcre_compile2(). The only
2892 difference is that it has an extra argument, which is a pointer to an
2893 integer error code. When there is a compile-time failure, this is set
2894 non-zero, in addition to the error test pointer being set to point to an
2895 error message. The new argument may be NULL if no error number is required
2896 (but then you may as well call pcre_compile(), which is now just a
2897 wrapper). This facility is provided because some applications need a
2898 numeric error indication, but it has also enabled me to tidy up the way
2899 compile-time errors are handled in the POSIX wrapper.
2901 16. Added VPATH=.libs to the makefile; this should help when building with one
2902 prefix path and installing with another. (Or so I'm told by someone who
2903 knows more about this stuff than I do.)
2905 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2906 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2907 match everything, including newlines. This is not POSIX-compatible, but
2908 somebody wanted the feature. From pcretest it can be activated by using
2909 both the P and the s flags.
2911 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2913 19. libpcre.pc was being incorrectly installed as executable.
2915 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2916 it now also looks for '\r' so that it will work unmodified on Windows.
2918 21. Added Google's contributed C++ wrapper to the distribution.
2920 22. Added some untidy missing memory free() calls in pcretest, to keep
2921 Electric Fence happy when testing.
2925 Version 5.0 13-Sep-04
2926 ---------------------
2928 1. Internal change: literal characters are no longer packed up into items
2929 containing multiple characters in a single byte-string. Each character
2930 is now matched using a separate opcode. However, there may be more than one
2931 byte in the character in UTF-8 mode.
2933 2. The pcre_callout_block structure has two new fields: pattern_position and
2934 next_item_length. These contain the offset in the pattern to the next match
2935 item, and its length, respectively.
2937 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2938 insertion of callouts before each pattern item. Added the /C option to
2939 pcretest to make use of this.
2941 4. On the advice of a Windows user, the lines
2943 #if defined(_WIN32) || defined(WIN32)
2944 _setmode( _fileno( stdout ), 0x8000 );
2945 #endif /* defined(_WIN32) || defined(WIN32) */
2947 have been added to the source of pcretest. This apparently does useful
2948 magic in relation to line terminators.
2950 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2951 for the benefit of those environments where the "b" makes a difference.
2953 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2954 to know about it. I have put a hack into configure.in that adds in code
2955 to set GCC=yes if CC=icc. This seems to end up at a point in the
2956 generated configure script that is early enough to affect the setting of
2957 compiler options, which is what is needed, but I have no means of testing
2958 whether it really works. (The user who reported this had patched the
2959 generated configure script, which of course I cannot do.)
2961 LATER: After change 22 below (new libtool files), the configure script
2962 seems to know about icc (and also ecc). Therefore, I have commented out
2963 this hack in configure.in.
2965 7. Added support for pkg-config (2 patches were sent in).
2967 8. Negated POSIX character classes that used a combination of internal tables
2968 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2969 [[:^ascii]]. Typically, they would match almost any characters. The other
2970 POSIX classes were not broken in this way.
2972 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2973 to find the match, as PCRE was deluded into thinking that the match had to
2974 start at the start point or following a newline. The same bug applied to
2975 patterns with negative forward assertions or any backward assertions
2976 preceding ".*" at the start, unless the pattern required a fixed first
2977 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2979 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2980 starting at the last subject character, bytes beyond the end of the subject
2981 string were read.
2983 11. Renamed the variable "class" as "classbits" to make life easier for C++
2984 users. (Previously there was a macro definition, but it apparently wasn't
2985 enough.)
2987 12. Added the new field "tables" to the extra data so that tables can be passed
2988 in at exec time, or the internal tables can be re-selected. This allows
2989 a compiled regex to be saved and re-used at a later time by a different
2990 program that might have everything at different addresses.
2992 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2993 -R library as well as a -L library.
2995 14. The debugging options of pcretest (-d on the command line or D on a
2996 pattern) showed incorrect output for anything following an extended class
2997 that contained multibyte characters and which was followed by a quantifier.
2999 15. Added optional support for general category Unicode character properties
3000 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3001 support. It adds about 90K to the size of the library. The meanings of the
3002 inbuilt class escapes such as \d and \s have NOT been changed.
3004 16. Updated pcredemo.c to include calls to free() to release the memory for the
3005 compiled pattern.
3007 17. The generated file chartables.c was being created in the source directory
3008 instead of in the building directory. This caused the build to fail if the
3009 source directory was different from the building directory, and was
3010 read-only.
3012 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3013 file. No doubt somebody will tell me if they don't make sense... Also added
3014 Dan Mooney's comments about building on OpenVMS.
3016 19. Added support for partial matching via the PCRE_PARTIAL option for
3017 pcre_exec() and the \P data escape in pcretest.
3019 20. Extended pcretest with 3 new pattern features:
3021 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3022 write the compiled pattern to the file whose name is "rest-of-line".
3023 This is a straight binary dump of the data, with the saved pointer to
3024 the character tables forced to be NULL. The study data, if any, is
3025 written too. After writing, pcretest reads a new pattern.
3027 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3028 compiled pattern from the given file. There must not be any
3029 occurrences of "<" in the file name (pretty unlikely); if there are,
3030 pcretest will instead treat the initial "<" as a pattern delimiter.
3031 After reading in the pattern, pcretest goes on to read data lines as
3032 usual.
3034 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3035 and 16-bit fields in a compiled pattern, to simulate a pattern that
3036 was compiled on a host of opposite endianness.
3038 21. The pcre-exec() function can now cope with patterns that were compiled on
3039 hosts of opposite endianness, with this restriction:
3041 As for any compiled expression that is saved and used later, the tables
3042 pointer field cannot be preserved; the extra_data field in the arguments
3043 to pcre_exec() should be used to pass in a tables address if a value
3044 other than the default internal tables were used at compile time.
3046 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3047 now diagnosed as an error. Previously, most of the time, a negative number
3048 would have been treated as zero, but if in addition "ovector" was passed as
3049 NULL, a crash could occur.
3051 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3052 new versions from the libtool 1.5 distribution (the last one is a copy of
3053 a file called libtool.m4). This seems to have fixed the need to patch
3054 "configure" to support Darwin 1.3 (which I used to do). However, I still
3055 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3056 workstation).
3058 24. Changed the PCRE licence to be the more standard "BSD" licence.
3061 Version 4.5 01-Dec-03
3062 ---------------------
3064 1. There has been some re-arrangement of the code for the match() function so
3065 that it can be compiled in a version that does not call itself recursively.
3066 Instead, it keeps those local variables that need separate instances for
3067 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3068 needs to "recurse". Keeping track of where control must go is done by means
3069 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3070 hide most of the details from the main code, and operates only if
3071 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3072 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3073 operating.
3075 To make it easier for callers to provide specially tailored get/free
3076 functions for this usage, two new functions, pcre_stack_malloc, and
3077 pcre_stack_free, are used. They are always called in strict stacking order,
3078 and the size of block requested is always the same.
3080 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3081 PCRE has been compiled to use the stack or the heap for recursion. The
3082 -C option of pcretest uses this to show which version is compiled.
3084 A new data escape \S, is added to pcretest; it causes the amounts of store
3085 obtained and freed by both kinds of malloc/free at match time to be added
3086 to the output.
3088 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3089 what's available on my current Linux desktop machine.
3091 3. When matching a UTF-8 string, the test for a valid string at the start has
3092 been extended. If start_offset is not zero, PCRE now checks that it points
3093 to a byte that is the start of a UTF-8 character. If not, it returns
3094 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3095 this is necessary because there may be backward assertions in the pattern.
3096 When matching the same subject several times, it may save resources to use
3097 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3099 4. The code for checking the validity of UTF-8 strings has been tightened so
3100 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3101 containing "overlong sequences".
3103 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3104 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3105 should have been "&", but it just so happened that all the cases this let
3106 through by mistake were picked up later in the function.
3108 6. I had used a variable called "isblank" - this is a C99 function, causing
3109 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3111 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3112 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3113 so the version is shown; (c) stop "make test" echoing "./RunTest".
3115 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3117 9. The prototype for memmove() for systems that don't have it was using
3118 size_t, but the inclusion of the header that defines size_t was later. I've
3119 moved the #includes for the C headers earlier to avoid this.
3121 10. Added some adjustments to the code to make it easier to compiler on certain
3122 special systems:
3124 (a) Some "const" qualifiers were missing.
3125 (b) Added the macro EXPORT before all exported functions; by default this
3126 is defined to be empty.
3127 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3128 that it reads its output file name as an argument instead of writing
3129 to the standard output and assuming this can be redirected.
3131 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3132 class containing characters with values greater than 255, PCRE compilation
3133 went into a loop.
3135 12. A recursive reference to a subpattern that was within another subpattern
3136 that had a minimum quantifier of zero caused PCRE to crash. For example,
3137 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3138 recursion. If the recursively-called subpattern itself had a zero repeat,
3139 that was OK.
3141 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3142 buffer into which it was copied (for escape processing) was still set at
3143 1024, so long lines caused crashes.
3145 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3146 "internal error: code overflow...". This applied to any character class
3147 that was followed by a possessive quantifier.
3149 15. Modified the Makefile to add libpcre.la as a prerequisite for
3150 libpcreposix.la because I was told this is needed for a parallel build to
3151 work.
3153 16. If a pattern that contained .* following optional items at the start was
3154 studied, the wrong optimizing data was generated, leading to matching
3155 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3156 matching string must start with a or b or c. The correct conclusion for
3157 this pattern is that a match can start with any character.
3160 Version 4.4 13-Aug-03
3161 ---------------------
3163 1. In UTF-8 mode, a character class containing characters with values between
3164 127 and 255 was not handled correctly if the compiled pattern was studied.
3165 In fixing this, I have also improved the studying algorithm for such
3166 classes (slightly).
3168 2. Three internal functions had redundant arguments passed to them. Removal
3169 might give a very teeny performance improvement.
3171 3. Documentation bug: the value of the capture_top field in a callout is *one
3172 more than* the number of the hightest numbered captured substring.
3174 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3175 in incorrectly linking with a previously installed version. They now link
3176 explicitly with libpcre.la.
3178 5. configure.in no longer needs to recognize Cygwin specially.
3180 6. A problem in pcre.in for Windows platforms is fixed.
3182 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3183 pcretest, it used to include the size of the study block as part of its
3184 output. Unfortunately, the structure contains a field that has a different
3185 size on different hardware architectures. This meant that the tests that
3186 showed this size failed. As the block is currently always of a fixed size,
3187 this information isn't actually particularly useful in pcretest output, so
3188 I have just removed it.
3190 8. Three pre-processor statements accidentally did not start in column 1.
3191 Sadly, there are *still* compilers around that complain, even though
3192 standard C has not required this for well over a decade. Sigh.
3194 9. In pcretest, the code for checking callouts passed small integers in the
3195 callout_data field, which is a void * field. However, some picky compilers
3196 complained about the casts involved for this on 64-bit systems. Now
3197 pcretest passes the address of the small integer instead, which should get
3198 rid of the warnings.
3200 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3201 both compile and run time, and gives an error if an invalid UTF-8 sequence
3202 is found. There is a option for disabling this check in cases where the
3203 string is known to be correct and/or the maximum performance is wanted.
3205 11. In response to a bug report, I changed one line in Makefile.in from
3207 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3208 to
3209 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3211 to look similar to other lines, but I have no way of telling whether this
3212 is the right thing to do, as I do not use Windows. No doubt I'll get told
3213 if it's wrong...
3216 Version 4.3 21-May-03
3217 ---------------------
3219 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3220 Makefile.
3222 2. Some refactoring to improve the quality of the code:
3224 (i) The utf8_table... variables are now declared "const".
3226 (ii) The code for \cx, which used the "case flipping" table to upper case
3227 lower case letters, now just substracts 32. This is ASCII-specific,
3228 but the whole concept of \cx is ASCII-specific, so it seems
3229 reasonable.
3231 (iii) PCRE was using its character types table to recognize decimal and
3232 hexadecimal digits in the pattern. This is silly, because it handles
3233 only 0-9, a-f, and A-F, but the character types table is locale-
3234 specific, which means strange things might happen. A private
3235 table is now used for this - though it costs 256 bytes, a table is
3236 much faster than multiple explicit tests. Of course, the standard
3237 character types table is still used for matching digits in subject
3238 strings against \d.
3240 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3241 ending in _t are). So I've renamed it as ESC_tee.
3243 3. The first argument for regexec() in the POSIX wrapper should have been
3244 defined as "const".
3246 4. Changed pcretest to use malloc() for its buffers so that they can be
3247 Electric Fenced for debugging.
3249 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3250 to read one or more bytes before the start of the subject string. Often this
3251 had no effect on PCRE's behaviour, but in some circumstances it could
3252 provoke a segmentation fault.
3254 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3255 to try to read one or more bytes before the start of the subject string.
3257 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3258 UTF-8 support could misbehave in various ways if the subject string
3259 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3260 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3261 back over UTF-8 characters.)
3264 Version 4.2 14-Apr-03
3265 ---------------------
3267 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3269 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3270 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3271 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3272 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3273 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3275 Note: automatic setting of the BUILD variables is not yet working
3276 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3277 compile-time but not at link-time
3278 [LINK]: use for linking executables only
3279 make different versions for Windows and non-Windows
3280 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3281 libraries
3282 [LINK_FOR_BUILD]: new variable
3283 [OBJEXT]: use throughout
3284 [EXEEXT]: use throughout
3285 <winshared>: new target
3286 <wininstall>: new target
3287 <dftables.o>: use native compiler
3288 <dftables>: use native linker
3289 <install>: handle Windows platform correctly
3290 <clean>: ditto
3291 <check>: ditto
3292 copy DLL to top builddir before testing
3294 As part of these changes, -no-undefined was removed again. This was reported
3295 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3296 in any case.
3298 3. Some tidies to get rid of compiler warnings:
3300 . In the match_data structure, match_limit was an unsigned long int, whereas
3301 match_call_count was an int. I've made them both unsigned long ints.
3303 . In pcretest the fact that a const uschar * doesn't automatically cast to
3304 a void * provoked a warning.
3306 . Turning on some more compiler warnings threw up some "shadow" variables
3307 and a few more missing casts.
3309 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3310 option, a class that contained a single character with a value between 128
3311 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3313 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3314 option, a class that contained several characters, but with at least one
3315 whose value was between 128 and 255 caused PCRE to crash.
3318 Version 4.1 12-Mar-03
3319 ---------------------
3321 1. Compiling with gcc -pedantic found a couple of places where casts were
3322 needed, and a string in dftables.c that was longer than standard compilers are
3323 required to support.
3325 2. Compiling with Sun's compiler found a few more places where the code could
3326 be tidied up in order to avoid warnings.
3328 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3329 first of these names is deprecated in the latest Autoconf in favour of the name
3330 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3331 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3332 analogy I have changed it to CFLAGS_FOR_BUILD.
3334 4. Added -no-undefined to the linking command in the Makefile, because this is
3335 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3336 linking step for the pcreposix library.
3338 5. PCRE was failing to diagnose the case of two named groups with the same
3339 name.
3341 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3342 literal character that is needed in the subject for a match, and scans along to
3343 ensure that it is present before embarking on the full matching process. This
3344 saves time in cases of nested unlimited repeats that are never going to match.
3345 Problem: the scan can take a lot of time if the subject is very long (e.g.
3346 megabytes), thus penalizing straightforward matches. It is now done only if the
3347 amount of subject to be scanned is less than 1000 bytes.
3349 7. A lesser problem with the same optimization is that it was recording the
3350 first character of an anchored pattern as "needed", thus provoking a search
3351 right along the subject, even when the first match of the pattern was going to
3352 fail. The "needed" character is now not set for anchored patterns, unless it
3353 follows something in the pattern that is of non-fixed length. Thus, it still
3354 fulfils its original purpose of finding quick non-matches in cases of nested
3355 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3358 Version 4.0 17-Feb-03
3359 ---------------------
3361 1. If a comment in an extended regex that started immediately after a meta-item
3362 extended to the end of string, PCRE compiled incorrect data. This could lead to
3363 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3365 2. Moved to autoconf 2.53 and libtool 1.4.2.
3367 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3368 the special perltest8 script is no longer needed - all the tests can be run
3369 from a single perltest script.
3371 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3372 by \s. It has now been removed in PCRE. This means it isn't recognized as
3373 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3374 class [:space:] *does* include VT, thereby creating a mess.
3376 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3377 space and tab.
3379 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3380 its new features into the main test script, reducing the number of scripts.
3382 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3383 were backward compatible, and made the (?i) apply to the whole pattern, as if
3384 /i were given. Now it behaves more logically, and applies the option setting
3385 only to what follows. PCRE has been changed to follow suit. However, if it
3386 finds options settings right at the start of the pattern, it extracts them into
3387 the global options, as before. Thus, they show up in the info data.
3389 8. Added support for the \Q...\E escape sequence. Characters in between are
3390 treated as literals. This is slightly different from Perl in that $ and @ are
3391 also handled as literals inside the quotes. In Perl, they will cause variable
3392 interpolation. Note the following examples:
3394 Pattern PCRE matches Perl matches
3396 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3397 \Qabc\$xyz\E abc\$xyz abc\$xyz
3398 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3400 For compatibility with Perl, \Q...\E sequences are recognized inside character
3401 classes as well as outside them.
3403 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3404 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3405 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3406 signed/unsigned warnings.
3408 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3409 option for pcretest, so I've replaced it by a simple function that does just
3410 that job.
3412 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3413 "pcregrep -".
3415 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3416 Java package. This provides some syntactic sugar for simple cases of what my
3417 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3418 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3419 item, you can use this simplified notation. Note that only makes sense with
3420 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3421 greediness, whatever the setting of the PCRE_UNGREEDY option.
3423 13. A change of greediness default within a pattern was not taking effect at
3424 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3425 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3426 was abstracted outside.
3428 14. PCRE now supports the \G assertion. It is true when the current matching
3429 position is at the start point of the match. This differs from \A when the
3430 starting offset is non-zero. Used with the /g option of pcretest (or similar
3431 code), it works in the same way as it does for Perl's /g option. If all
3432 alternatives of a regex begin with \G, the expression is anchored to the start
3433 match position, and the "anchored" flag is set in the compiled expression.
3435 15. Some bugs concerning the handling of certain option changes within patterns
3436 have been fixed. These applied to options other than (?ims). For example,
3437 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3438 been the other way round. Some of this was related to change 7 above.
3440 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3441 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3442 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3443 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3445 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3446 mode. Unlike ".", it always matches newline, whatever the setting of
3447 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3448 assertions. Perl allows it, but it doesn't (in general) work because it can't
3449 calculate the length of the lookbehind. At least, that's the case for Perl
3450 5.8.0 - I've been told they are going to document that it doesn't work in
3451 future.
3453 18. Added an error diagnosis for escapes that PCRE does not support: these are
3454 \L, \l, \N, \P, \p, \U, \u, and \X.
3456 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3457 reading past the end of the pattern in cases such as /[abcd/.
3459 20. PCRE was getting more memory than necessary for patterns with classes that
3460 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3462 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3463 compiling PCRE for use with Virtual Pascal.
3465 22. Small fix to the Makefile to make it work properly if the build is done
3466 outside the source tree.
3468 23. Added a new extension: a condition to go with recursion. If a conditional
3469 subpattern starts with (?(R) the "true" branch is used if recursion has
3470 happened, whereas the "false" branch is used only at the top level.
3472 24. When there was a very long string of literal characters (over 255 bytes
3473 without UTF support, over 250 bytes with UTF support), the computation of how
3474 much memory was required could be incorrect, leading to segfaults or other
3475 strange effects.
3477 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3478 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3479 there was a subsequent back reference to those brackets. This meant that, for
3480 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3481 possible to check for precisely this case. All we can do is abandon the
3482 optimization if .* occurs inside capturing brackets when there are any back
3483 references whatsoever. (See below for a better fix that came later.)
3485 26. The handling of the optimization for finding the first character of a
3486 non-anchored pattern, and for finding a character that is required later in the
3487 match were failing in some cases. This didn't break the matching; it just
3488 failed to optimize when it could. The way this is done has been re-implemented.
3490 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3492 28. Added a new feature that provides some of the functionality that Perl
3493 provides with (?{...}). The facility is termed a "callout". The way it is done
3494 in PCRE is for the caller to provide an optional function, by setting
3495 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3496 global variable. By default it is unset, which disables all calling out. To get
3497 the function called, the regex must include (?C) at appropriate points. This
3498 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3499 This provides a means of identifying different callout points. When PCRE
3500 reaches such a point in the regex, if pcre_callout has been set, the external
3501 function is called. It is provided with data in a structure called
3502 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3503 matching continues; if it returns a non-zero value, the match at the current
3504 point fails. However, backtracking will occur if possible. [This was changed
3505 later and other features added - see item 49 below.]
3507 29. pcretest is upgraded to test the callout functionality. It provides a
3508 callout function that displays information. By default, it shows the start of
3509 the match and the current position in the text. There are some new data escapes
3510 to vary what happens:
3512 \C+ in addition, show current contents of captured substrings
3513 \C- do not supply a callout function
3514 \C!n return 1 when callout number n is reached
3515 \C!n!m return 1 when callout number n is reached for the mth time
3517 30. If pcregrep was called with the -l option and just a single file name, it
3518 output "<stdin>" if a match was found, instead of the file name.
3520 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3521 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3522 pcre_exec(). This saves a malloc/free per call. The default value of
3523 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3524 when configuring.
3526 32. The default maximum size of a compiled pattern is 64K. There have been a
3527 few cases of people hitting this limit. The code now uses macros to handle the
3528 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3529 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3530 configuring. Tests 2 and 5 work only with 2-byte links because they output
3531 debugging information about compiled patterns.
3533 33. Internal code re-arrangements:
3535 (a) Moved the debugging function for printing out a compiled regex into
3536 its own source file (printint.c) and used #include to pull it into
3537 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3538 separate copies.
3540 (b) Defined the list of op-code names for debugging as a macro in
3541 internal.h so that it is next to the definition of the opcodes.
3543 (c) Defined a table of op-code lengths for simpler skipping along compiled
3544 code. This is again a macro in internal.h so that it is next to the
3545 definition of the opcodes.
3547 34. Added support for recursive calls to individual subpatterns, along the
3548 lines of Robin Houston's patch (but implemented somewhat differently).
3550 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3551 allow it to read and process whole directories in Win32. This code was
3552 contributed by Lionel Fourquaux; it has not been tested by me.
3554 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3555 used to name a group. Names consist of alphanumerics and underscores, and must
3556 be unique. Back references use the syntax (?P=name) and recursive calls use
3557 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3558 numbers. The function pcre_fullinfo() can be used after compilation to extract
3559 a name/number map. There are three relevant calls:
3561 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3562 PCRE_INFO_NAMECOUNT yields the number of entries
3563 PCRE_INFO_NAMETABLE yields a pointer to the map.
3565 The map is a vector of fixed-size entries. The size of each entry depends on
3566 the length of the longest name used. The first two bytes of each entry are the
3567 group number, most significant byte first. There follows the corresponding
3568 name, zero terminated. The names are in alphabetical order.
3570 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3571 case instead of 255. Making it the same both with and without UTF-8 support
3572 means that the same test output works with both.
3574 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3575 calling malloc() with a zero argument.
3577 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3578 optimization. I've improved things by keeping a bitmap of backreferences with
3579 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3580 fact referenced, the optimization can be applied. It is unlikely that a
3581 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3582 the match to follow \n) will appear inside brackets with a number greater than
3583 31, but if it does, any back reference > 31 suppresses the optimization.
3585 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3586 of disabling numbered capturing parentheses. Any opening parenthesis that is
3587 not followed by ? behaves as if it were followed by ?: but named parentheses
3588 can still be used for capturing (and they will acquire numbers in the usual
3589 way).
3591 41. Redesigned the return codes from the match() function into yes/no/error so
3592 that errors can be passed back from deep inside the nested calls. A malloc
3593 failure while inside a recursive subpattern call now causes the
3594 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3596 42. It is now possible to set a limit on the number of times the match()
3597 function is called in a call to pcre_exec(). This facility makes it possible to
3598 limit the amount of recursion and backtracking, though not in a directly
3599 obvious way, because the match() function is used in a number of different
3600 circumstances. The count starts from zero for each position in the subject
3601 string (for non-anchored patterns). The default limit is, for compatibility, a
3602 large number, namely 10 000 000. You can change this in two ways:
3604 (a) When configuring PCRE before making, you can use --with-match-limit=n
3605 to set a default value for the compiled library.
3607 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3608 a different value is set. See 45 below.
3610 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3612 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3613 of things that can be changed at compile time. The first argument specifies
3614 what is wanted and the second points to where the information is to be placed.
3615 The current list of available information is:
3619 The output is an integer that is set to one if UTF-8 support is available;
3620 otherwise it is set to zero.
3624 The output is an integer that it set to the value of the code that is used for
3625 newline. It is either LF (10) or CR (13).
3629 The output is an integer that contains the number of bytes used for internal
3630 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3634 The output is an integer that contains the threshold above which the POSIX
3635 interface uses malloc() for output vectors. See item 31 above.
3639 The output is an unsigned integer that contains the default limit of the number
3640 of match() calls in a pcre_exec() execution. See 42 above.
3642 44. pcretest has been upgraded by the addition of the -C option. This causes it
3643 to extract all the available output from the new pcre_config() function, and to
3644 output it. The program then exits immediately.
3646 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3647 order to support additional features. One way would have been to define
3648 pcre_exec2() (for example) with extra arguments, but this would not have been
3649 extensible, and would also have required all calls to the original function to
3650 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3651 is used for passing in "extra" data from pcre_study().
3653 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3654 contains the following fields:
3656 flags a bitmap indicating which of the following fields are set
3657 study_data opaque data from pcre_study()
3658 match_limit a way of specifying a limit on match() calls for a specific
3659 call to pcre_exec()
3660 callout_data data for callouts (see 49 below)
3662 The flag bits are also defined in pcre.h, and are
3668 The pcre_study() function now returns one of these new pcre_extra blocks, with
3669 the actual study data pointed to by the study_data field, and the
3670 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3671 before. That is, this change is entirely upwards-compatible and requires no
3672 change to existing code.
3674 If you want to pass in additional data to pcre_exec(), you can either place it
3675 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3676 block.
3678 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3679 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3680 times with different match limits, until it finds the minimum value needed for
3681 pcre_exec() to complete. The value is then output. This can be instructive; for
3682 most simple matches the number is quite small, but for pathological cases it
3683 gets very large very quickly.
3685 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3686 returns the size of the data block pointed to by the study_data field in a
3687 pcre_extra block, that is, the value that was passed as the argument to
3688 pcre_malloc() when PCRE was getting memory in which to place the information
3689 created by pcre_study(). The fourth argument should point to a size_t variable.
3690 pcretest has been extended so that this information is shown after a successful
3691 pcre_study() call when information about the compiled regex is being displayed.
3693 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3694 because what follows is always an absolute path. (Later: it turns out that this
3695 is more than cosmetic for MinGW, because it doesn't like empty path
3696 components.)
3698 49. Some changes have been made to the callout feature (see 28 above):
3700 (i) A callout function now has three choices for what it returns:
3702 0 => success, carry on matching
3703 > 0 => failure at this point, but backtrack if possible
3704 < 0 => serious error, return this value from pcre_exec()
3706 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3707 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3708 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3709 use by callout functions. It will never be used by PCRE itself.
3711 (ii) The pcre_extra structure (see 45 above) has a void * field called
3712 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3713 pcre_callout_block structure has a field of the same name. The contents of
3714 the field passed in the pcre_extra structure are passed to the callout
3715 function in the corresponding field in the callout block. This makes it
3716 easier to use the same callout-containing regex from multiple threads. For
3717 testing, the pcretest program has a new data escape
3719 \C*n pass the number n (may be negative) as callout_data
3721 If the callout function in pcretest receives a non-zero value as
3722 callout_data, it returns that value.
3724 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3725 there were some redundant $(CFLAGS) in commands that are now specified as
3726 $(LINK), which already includes $(CFLAGS).
3728 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3729 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3730 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3731 one-byte characters throughout. Note that case-insensitive matching applies
3732 only to characters whose values are less than 256. PCRE doesn't support the
3733 notion of cases for higher-valued characters.
3735 (i) A character class whose characters are all within 0-255 is handled as
3736 a bit map, and the map is inverted for negative classes. Previously, a
3737 character > 255 always failed to match such a class; however it should
3738 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3740 (ii) A negated character class with a single character < 255 is coded as
3741 "not this character" (OP_NOT). This wasn't working properly when the test
3742 character was multibyte, either singly or repeated.
3744 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3745 mode, for example: \x{100}{2,3}.
3747 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3748 singly or repeated) now correctly test multibyte characters. However,
3749 PCRE doesn't recognize any characters with values greater than 255 as
3750 digits, spaces, or word characters. Such characters always match \D, \S,
3751 and \W, and never match \d, \s, or \w.
3753 (v) Classes may now contain characters and character ranges with values
3754 greater than 255. For example: [ab\x{100}-\x{400}].
3756 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3757 PCRE in UTF-8 mode.
3759 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3760 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3761 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3762 value.)
3764 53. The single man page has become too large. I have therefore split it up into
3765 a number of separate man pages. These also give rise to individual HTML pages;
3766 these are now put in a separate directory, and there is an index.html page that
3767 lists them all. Some hyperlinking between the pages has been installed.
3769 54. Added convenience functions for handling named capturing parentheses.
3771 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3772 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3773 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3774 are faulted.
3776 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3777 calling configure. These values are used when compiling the dftables.c program
3778 which is run to generate the source of the default character tables. They
3779 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3780 you will need to set these values.
3782 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3785 Version 3.9 02-Jan-02
3786 ---------------------
3788 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3790 2. If --disable-static was given, the building process failed when trying to
3791 build pcretest and pcregrep. (For some reason it was using libtool to compile
3792 them, which is not right, as they aren't part of the library.)
3795 Version 3.8 18-Dec-01
3796 ---------------------
3798 1. The experimental UTF-8 code was completely screwed up. It was packing the
3799 bytes in the wrong order. How dumb can you get?
3802 Version 3.7 29-Oct-01
3803 ---------------------
3805 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3806 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3807 this didn't happen under Solaris 8, where I normally test things.
3809 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3810 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3811 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3812 this fix an hour or so after the initial 3.7 release.)
3815 Version 3.6 23-Oct-01
3816 ---------------------
3818 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3819 offsets passed as NULL with zero offset count.
3821 2. The config.guess and config.sub files had not been updated when I moved to
3822 the latest autoconf.
3825 Version 3.5 15-Aug-01
3826 ---------------------
3828 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3829 had been forgotten.
3831 2. By using declared but undefined structures, we can avoid using "void"
3832 definitions in pcre.h while keeping the internal definitions of the structures
3833 private.
3835 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3836 user point of view, this means that both static and shared libraries are built
3837 by default, but this can be individually controlled. More of the work of
3838 handling this static/shared cases is now inside libtool instead of PCRE's make
3839 file.
3841 4. The pcretest utility is now installed along with pcregrep because it is
3842 useful for users (to test regexs) and by doing this, it automatically gets
3843 relinked by libtool. The documentation has been turned into a man page, so
3844 there are now .1, .txt, and .html versions in /doc.
3846 5. Upgrades to pcregrep:
3847 (i) Added long-form option names like gnu grep.
3848 (ii) Added --help to list all options with an explanatory phrase.
3849 (iii) Added -r, --recursive to recurse into sub-directories.
3850 (iv) Added -f, --file to read patterns from a file.
3852 6. pcre_exec() was referring to its "code" argument before testing that
3853 argument for NULL (and giving an error if it was NULL).
3855 7. Upgraded Makefile.in to allow for compiling in a different directory from
3856 the source directory.
3858 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3859 options bits, the pointer it was passed was to an int instead of to an unsigned
3860 long int. This mattered only on 64-bit systems.
3862 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3863 generated) instead of pcre.in, which it its source. Also made the same change
3864 in several of the .c files.
3866 10. A new release of gcc defines printf() as a macro, which broke pcretest
3867 because it had an ifdef in the middle of a string argument for printf(). Fixed
3868 by using separate calls to printf().
3870 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3871 script, to force use of CR or LF instead of \n in the source. On non-Unix
3872 systems, the value can be set in config.h.
3874 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3875 absolute limit. Changed the text of the error message to make this clear, and
3876 likewise updated the man page.
3878 13. The limit of 99 on the number of capturing subpatterns has been removed.
3879 The new limit is 65535, which I hope will not be a "real" limit.
3882 Version 3.4 22-Aug-00
3883 ---------------------
3885 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3887 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3890 Version 3.3 01-Aug-00
3891 ---------------------
3893 1. If an octal character was given, but the value was greater than \377, it
3894 was not getting masked to the least significant bits, as documented. This could
3895 lead to crashes in some systems.
3897 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3898 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3900 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3901 These just pass their arguments on to (pcre_free)(), but they are provided
3902 because some uses of PCRE bind it to non-C systems that can call its functions,
3903 but cannot call free() or pcre_free() directly.
3905 4. Add "make test" as a synonym for "make check". Corrected some comments in
3906 the Makefile.
3908 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3909 Makefile.
3911 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3912 command called pgrep for grepping around the active processes.
3914 7. Added the beginnings of support for UTF-8 character strings.
3916 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3917 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3918 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3919 out for the ar command.)
3922 Version 3.2 12-May-00
3923 ---------------------
3925 This is purely a bug fixing release.
3927 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3928 of ZA. This was just one example of several cases that could provoke this bug,
3929 which was introduced by change 9 of version 2.00. The code for breaking
3930 infinite loops after an iteration that matches an empty string was't working
3931 correctly.
3933 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3934 when matched against abbab (for example). After matching an empty string, it
3935 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3936 caused it to match further down the string than it should.
3938 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3939 was there because it doesn't seem to be needed, and it causes trouble on some
3940 systems, as it is not a Standard C header. It has been removed.
3942 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3943 were reported on the Macintosh. The changes were from
3945 while ((c = *(++ptr)) != 0 && c != '\n');
3946 to
3947 while ((c = *(++ptr)) != 0 && c != '\n') ;
3949 Totally extraordinary, but if that's what it takes...
3951 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3952 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3953 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3954 assumes the way PCRE uses memmove() (always moving upwards).
3956 6. PCRE is being used in one environment where strchr() is not available. There
3957 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3958 faster code anyway.
3961 Version 3.1 09-Feb-00
3962 ---------------------
3964 The only change in this release is the fixing of some bugs in Makefile.in for
3965 the "install" target:
3967 (1) It was failing to install pcreposix.h.
3969 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3972 Version 3.0 01-Feb-00
3973 ---------------------
3975 1. Add support for the /+ modifier to perltest (to output $` like it does in
3976 pcretest).
3978 2. Add support for the /g modifier to perltest.
3980 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3981 matches null strings.
3983 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3984 pattern. Perl treats empty patterns specially - it reuses the most recent
3985 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3986 effect.
3988 5. The POSIX interface was broken in that it was just handing over the POSIX
3989 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3990 required a bigger vector, with some working space on the end. This means that
3991 the POSIX wrapper now has to get and free some memory, and copy the results.
3993 6. Added some simple autoconf support, placing the test data and the
3994 documentation in separate directories, re-organizing some of the
3995 information files, and making it build pcre-config (a GNU standard). Also added
3996 libtool support for building PCRE as a shared library, which is now the
3997 default.
3999 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4000 09 are not valid octal constants. Single digits will be used for minor values
4001 less than 10.
4003 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4004 existing programs that set these in the POSIX interface can use PCRE without
4005 modification.
4007 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4008 return all that pcre_info() returns, plus additional data. The pcre_info()
4009 function is retained for compatibility, but is considered to be obsolete.
4011 10. Added experimental recursion feature (?R) to handle one common case that
4012 Perl 5.6 will be able to do with (?p{...}).
4014 11. Added support for POSIX character classes like [:alpha:], which Perl is
4015 adopting.
4018 Version 2.08 31-Aug-99
4019 ----------------------
4021 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4022 trying to match at the startoffset position, but instead was moving forward to
4023 the next newline as if a previous match had failed.
4025 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4026 and could get into a loop if a null string was matched other than at the start
4027 of the subject.
4029 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4030 be distinguished at compile time, and for completeness also added PCRE_DATE.
4032 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4033 in GnuWin32 environments.
4036 Version 2.07 29-Jul-99
4037 ----------------------
4039 1. The documentation is now supplied in plain text form and HTML as well as in
4040 the form of man page sources.
4042 2. C++ compilers don't like assigning (void *) values to other pointer types.
4043 In particular this affects malloc(). Although there is no problem in Standard
4044 C, I've put in casts to keep C++ compilers happy.
4046 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4047 should be (const char *).
4049 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4050 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4051 However, I haven't made this a standard facility. The documentation doesn't
4052 mention it, and the Makefile doesn't support it.
4054 5. The Makefile now contains an "install" target, with editable destinations at
4055 the top of the file. The pcretest program is not installed.
4057 6. pgrep -V now gives the PCRE version number and date.
4059 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4060 causing the entire string to be ignored, instead of just the last character.
4062 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4063 non-matching string, it can take a very, very long time, even for strings of
4064 quite modest length, because of the nested recursion. PCRE now does better in
4065 some of these cases. It does this by remembering the last required literal
4066 character in the pattern, and pre-searching the subject to ensure it is present
4067 before running the real match. In other words, it applies a heuristic to detect
4068 some types of certain failure quickly, and in the above example, if presented
4069 with a string that has no trailing " it gives "no match" very quickly.
4071 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4072 other alternatives are tried instead.
4075 Version 2.06 09-Jun-99
4076 ----------------------
4078 1. Change pcretest's output for amount of store used to show just the code
4079 space, because the remainder (the data block) varies in size between 32-bit and
4080 64-bit systems.
4082 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4083 start matching at. This allows lookbehinds to work when searching for multiple
4084 occurrences in a string.
4086 3. Added additional options to pcretest for testing multiple occurrences:
4088 /+ outputs the rest of the string that follows a match
4089 /g loops for multiple occurrences, using the new startoffset argument
4090 /G loops for multiple occurrences by passing an incremented pointer
4092 4. PCRE wasn't doing the "first character" optimization for patterns starting
4093 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4094 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4095 the letter 'x'. On long subject strings, this gives a significant speed-up.
4098 Version 2.05 21-Apr-99
4099 ----------------------
4101 1. Changed the type of magic_number from int to long int so that it works
4102 properly on 16-bit systems.
4104 2. Fixed a bug which caused patterns starting with .* not to work correctly
4105 when the subject string contained newline characters. PCRE was assuming
4106 anchoring for such patterns in all cases, which is not correct because .* will
4107 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4108 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4109 must be retried after every newline in the subject.
4112 Version 2.04 18-Feb-99
4113 ----------------------
4115 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4116 computation of the store needed to hold the pattern was incorrect (too large).
4117 If such patterns were nested a few deep, this could multiply and become a real
4118 problem.
4120 2. Added /M option to pcretest to show the memory requirement of a specific
4121 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4123 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4124 compiled in such a way that the backtracking after subsequent failure was
4125 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4126 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4129 Version 2.03 02-Feb-99
4130 ----------------------
4132 1. Fixed typo and small mistake in man page.
4134 2. Added 4th condition (GPL supersedes if conflict) and created separate
4135 LICENCE file containing the conditions.
4137 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4138 Perl, that is the internal \ allows the delimiter to be included in the
4139 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4140 the final delimiter, add \ to the end of the pattern (to test the error).
4142 4. Added the convenience functions for extracting substrings after a successful
4143 match. Updated pcretest to make it able to test these functions.
4146 Version 2.02 14-Jan-99
4147 ----------------------
4149 1. Initialized the working variables associated with each extraction so that
4150 their saving and restoring doesn't refer to uninitialized store.
4152 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4153 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4154 fix the problem.
4156 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4157 calls, and wasn't printing the correct value for compiling calls. Increased the
4158 default value of LOOPREPEAT, and the number of significant figures in the
4159 times.
4161 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4163 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4164 a building problem on Windows NT with a FAT file system.
4167 Version 2.01 21-Oct-98
4168 ----------------------
4170 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4171 to character tables built by pcre_maketables() in the current locale. If NULL
4172 is passed, the default tables are used.
4175 Version 2.00 24-Sep-98
4176 ----------------------
4178 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4179 it any more.
4181 2. Allow quantification of (?>) groups, and make it work correctly.
4183 3. The first character computation wasn't working for (?>) groups.
4185 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4186 end of the subject) and add 5.005's \z, which really does match only at the
4187 very end of the subject.
4189 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4191 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4192 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4193 localized options. All options to pcre_study() were also removed.
4195 7. Add other new features from 5.005:
4197 $(?<= positive lookbehind
4198 $(?<! negative lookbehind
4199 (?imsx-imsx) added the unsetting capability
4200 such a setting is global if at outer level; local otherwise
4201 (?imsx-imsx:) non-capturing groups with option setting
4202 (?(cond)re|re) conditional pattern matching
4204 A backreference to itself in a repeated group matches the previous
4205 captured string.
4207 8. General tidying up of studying (both automatic and via "study")
4208 consequential on the addition of new assertions.
4210 9. As in 5.005, unlimited repeated groups that could match an empty substring
4211 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4212 runtime if any iteration does actually match an empty substring.
4214 10. Include the RunTest script in the distribution.
4216 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4217 discrepancies, some of which were old and were also with respect to 5.004. They
4218 have now been fixed.
4221 Version 1.09 28-Apr-98
4222 ----------------------
4224 1. A negated single character class followed by a quantifier with a minimum
4225 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4226 program crashes, or just wrong answers. This did not apply to negated classes
4227 containing more than one character, or to minima other than one.
4230 Version 1.08 27-Mar-98
4231 ----------------------
4233 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4235 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4236 latter must appear before anything that relies on it in the pattern.
4239 Version 1.07 16-Feb-98
4240 ----------------------
4242 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4243 repeat of a potentially empty string).
4246 Version 1.06 23-Jan-98
4247 ----------------------
4249 1. Added Markus Oberhumer's little patches for C++.
4251 2. Literal strings longer than 255 characters were broken.
4254 Version 1.05 23-Dec-97
4255 ----------------------
4257 1. Negated character classes containing more than one character were failing if
4258 PCRE_CASELESS was set at run time.
4261 Version 1.04 19-Dec-97
4262 ----------------------
4264 1. Corrected the man page, where some "const" qualifiers had been omitted.
4266 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4267 input syntax.
4269 3. Fixed memory leak which occurred when a regex with back references was
4270 matched with an offsets vector that wasn't big enough. The temporary memory
4271 that is used in this case wasn't being freed if the match failed.
4273 4. Tidied pcretest to ensure it frees memory that it gets.
4275 5. Temporary memory was being obtained in the case where the passed offsets
4276 vector was exactly big enough.
4278 6. Corrected definition of offsetof() from change 5 below.
4280 7. I had screwed up change 6 below and broken the rules for the use of
4281 setjmp(). Now fixed.
4284 Version 1.03 18-Dec-97
4285 ----------------------
4287 1. A erroneous regex with a missing opening parenthesis was correctly
4288 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4289 on some systems.
4291 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4292 it was reported that one broken compiler failed on the former because "code" is
4293 also an independent variable.
4295 3. The erroneous regex a[]b caused an array overrun reference.
4297 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4298 fail on data ending with that character. (It was going on too far, and checking
4299 the next character, typically a binary zero.) This was specific to the
4300 optimized code for single-character negative classes.
4302 5. Added a contributed patch from the TIN world which does the following:
4304 + Add an undef for memmove, in case the the system defines a macro for it.
4306 + Add a definition of offsetof(), in case there isn't one. (I don't know
4307 the reason behind this - offsetof() is part of the ANSI standard - but
4308 it does no harm).
4310 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4311 most of the places where whitespace preceded '#'. I have given up and
4312 allowed the remaining 2 cases to be at the margin.
4314 + Rename some variables in pcre to eliminate shadowing. This seems very
4315 pedantic, but does no harm, of course.
4317 6. Moved the call to setjmp() into its own function, to get rid of warnings
4318 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4320 7. Constructs such as \d{8,} were compiling into the equivalent of
4321 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4322 outcome, but in this particular case used more store than had been allocated,
4323 which caused the bug to be discovered because it threw up an internal error.
4325 8. The debugging code in both pcre and pcretest for outputting the compiled
4326 form of a regex was going wrong in the case of back references followed by
4327 curly-bracketed repeats.
4330 Version 1.02 12-Dec-97
4331 ----------------------
4333 1. Typos in pcre.3 and comments in the source fixed.
4335 2. Applied a contributed patch to get rid of places where it used to remove
4336 'const' from variables, and fixed some signed/unsigned and uninitialized
4337 variable warnings.
4339 3. Added the "runtest" target to Makefile.
4341 4. Set default compiler flag to -O2 rather than just -O.
4344 Version 1.01 19-Nov-97
4345 ----------------------
4347 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4348 like /([ab]*)*/, that is, for classes with more than one character in them.
4350 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4351 as /((?>a*))*/ (a PCRE_EXTRA facility).
4354 Version 1.00 18-Nov-97
4355 ----------------------
4357 1. Added compile-time macros to support systems such as SunOS4 which don't have
4358 memmove() or strerror() but have other things that can be used instead.
4360 2. Arranged that "make clean" removes the executables.
4363 Version 0.99 27-Oct-97
4364 ----------------------
4366 1. Fixed bug in code for optimizing classes with only one character. It was
4367 initializing a 32-byte map regardless, which could cause it to run off the end
4368 of the memory it had got.
4370 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4373 Version 0.98 22-Oct-97
4374 ----------------------
4376 1. Fixed bug in code for handling temporary memory usage when there are more
4377 back references than supplied space in the ovector. This could cause segfaults.
4380 Version 0.97 21-Oct-97
4381 ----------------------
4383 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4385 2. Optimized negated single characters not to use a bit map.
4387 3. Brought error texts together as macro definitions; clarified some of them;
4388 fixed one that was wrong - it said "range out of order" when it meant "invalid
4389 escape sequence".
4391 4. Changed some char * arguments to const char *.
4393 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4395 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4396 pcretest.
4399 Version 0.96 16-Oct-97
4400 ----------------------
4402 1. Added a simple "pgrep" utility to the distribution.
4404 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4405 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4406 where "ddd" means "one or more decimal digits".
4408 3. Fixed serious bug. If a pattern had a back reference, but the call to
4409 pcre_exec() didn't supply a large enough ovector to record the related
4410 identifying subpattern, the match always failed. PCRE now remembers the number
4411 of the largest back reference, and gets some temporary memory in which to save
4412 the offsets during matching if necessary, in order to ensure that
4413 backreferences always work.
4415 4. Increased the compatibility with Perl in a number of ways:
4417 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4418 to request this handling. The option can be set at compile or exec time.
4420 (b) $ matches before a terminating newline by default; an option
4421 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4422 mode). The option can be set at compile or exec time.
4424 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4425 the same as Perl's. If the decimal number it represents is less than 10
4426 or there aren't that many previous left capturing parentheses, an octal
4427 escape is read. Inside a character class, it's always an octal escape,
4428 even if it is a single digit.
4430 (d) An escaped but undefined alphabetic character is taken as a literal,
4431 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4432 escapes.
4434 (e) {0} is now permitted. (The previous item is removed from the compiled
4435 pattern).
4437 5. Changed all the names of code files so that the basic parts are no longer
4438 than 10 characters, and abolished the teeny "globals.c" file.
4440 6. Changed the handling of character classes; they are now done with a 32-byte
4441 bit map always.
4443 7. Added the -d and /D options to pcretest to make it possible to look at the
4444 internals of compilation without having to recompile pcre.
4447 Version 0.95 23-Sep-97
4448 ----------------------
4450 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4451 \x20 at the start of a run of normal characters. These were being treated as
4452 real characters, instead of the source characters being re-checked.
4455 Version 0.94 18-Sep-97
4456 ----------------------
4458 1. The functions are now thread-safe, with the caveat that the global variables
4459 containing pointers to malloc() and free() or alternative functions are the
4460 same for all threads.
4462 2. Get pcre_study() to generate a bitmap of initial characters for non-
4463 anchored patterns when this is possible, and use it if passed to pcre_exec().
4466 Version 0.93 15-Sep-97
4467 ----------------------
4469 1. /(b)|(:+)/ was computing an incorrect first character.
4471 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4472 but not actually doing anything yet.
4474 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4475 as Perl does (e.g. [-az] or [az-]).
4477 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4478 all possible positions.
4480 5. Split up into different modules to avoid including unneeded functions in a
4481 compiled binary. However, compile and exec are still in one module. The "study"
4482 function is split off.
4484 6. The character tables are now in a separate module whose source is generated
4485 by an auxiliary program - but can then be edited by hand if required. There are
4486 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4487 toupper() in the code.
4489 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4490 make them global. Abolish the function for setting them, as the caller can now
4491 set them directly.
4494 Version 0.92 11-Sep-97
4495 ----------------------
4497 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4498 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4500 2. Caseless matching was not working in character classes if the characters in
4501 the pattern were in upper case.
4503 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4505 4. Make PCRE_ANCHORED public and accept as a compile option.
4507 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4508 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4509 pass them.
4511 6. Give an error if bad option bits passed at compile or run time.
4513 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4514 pcretest to cause it to pass that flag.
4516 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4517 options, and the first character, if set.
4519 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4522 Version 0.91 10-Sep-97
4523 ----------------------
4525 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4526 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4528 2. PCRE was looping on encountering an indefinitely repeated back reference to
4529 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4530 Perl does - treats the match as successful.
4532 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5