ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 672 - (show annotations)
Tue Aug 23 16:45:55 2011 UTC (4 years, 1 month ago) by ph10
File size: 212930 byte(s)
Update non-manpage documentation for JIT.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.20
5 ------------
7 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
8 a POSIX class. After further experiments with Perl, which convinced me that
9 Perl has bugs and confusions, a closing square bracket is no longer allowed
10 in a POSIX name.
12 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no captured
13 substring, but while checking the failing first alternative, substring 1 is
14 temporarily captured. If the output vector supplied to pcre_exec() was not
15 big enough for this capture, the yield of the function was still zero
16 ("insufficient space for captured substrings"). This cannot be totally fixed
17 without adding another stack variable, which seems a lot of expense for a
18 edge case. However, I have improved the situation in cases such as
19 /(a)(b)x|abc/ matched against "abc", where the return code indicates that
20 fewer than the maximum number of slots in the ovector have been set.
22 3. Related to (2) above: when there are more back references in a pattern than
23 slots in the output vector, pcre_exec() uses temporary memory during
24 matching, and copies in the captures as far as possible afterwards. It was
25 using the entire output vector, but this conflicts with the specification
26 that only 2/3 is used for passing back captured substrings. Now it uses only
27 the first 2/3, for compatibility. This is, of course, another edge case.
29 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
30 main code base, and can be used by building with --enable-jit.
33 Version 8.13 16-Aug-2011
34 ------------------------
36 1. The Unicode data tables have been updated to Unicode 6.0.0.
38 2. Two minor typos in pcre_internal.h have been fixed.
40 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
41 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
42 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
44 4. There were a number of related bugs in the code for matching backrefences
45 caselessly in UTF-8 mode when codes for the characters concerned were
46 different numbers of bytes. For example, U+023A and U+2C65 are an upper
47 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
48 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
49 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
50 2-byte code at the end of the subject (it thought there wasn't enough data
51 left).
53 5. Comprehensive information about what went wrong is now returned by
54 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
55 as the output vector has at least 2 elements. The offset of the start of
56 the failing character and a reason code are placed in the vector.
58 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
59 now returned is for the first byte of the failing character, instead of the
60 last byte inspected. This is an incompatible change, but I hope it is small
61 enough not to be a problem. It makes the returned offset consistent with
62 pcre_exec() and pcre_dfa_exec().
64 7. pcretest now gives a text phrase as well as the error number when
65 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
66 failure, the offset and reason code are output.
68 8. When \R was used with a maximizing quantifier it failed to skip backwards
69 over a \r\n pair if the subsequent match failed. Instead, it just skipped
70 back over a single character (\n). This seems wrong (because it treated the
71 two characters as a single entity when going forwards), conflicts with the
72 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
73 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
74 has been changed.
76 9. Some internal refactoring has changed the processing so that the handling
77 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
78 time (the PCRE_DOTALL option was changed this way some time ago: version
79 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
80 which was always a bit of a fudge. It also means that there is one less
81 argument for the match() function, which reduces its stack requirements
82 slightly. This change also fixes an incompatibility with Perl: the pattern
83 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
85 10. More internal refactoring has drastically reduced the number of recursive
86 calls to match() for possessively repeated groups such as (abc)++ when
87 using pcre_exec().
89 11. While implementing 10, a number of bugs in the handling of groups were
90 discovered and fixed:
92 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
93 (a|)*(?1) gave a compile-time internal error.
94 ((a|)+)+ did not notice that the outer group could match an empty string.
95 (^a|^)+ was not marked as anchored.
96 (.*a|.*)+ was not marked as matching at start or after a newline.
98 12. Yet more internal refactoring has removed another argument from the match()
99 function. Special calls to this function are now indicated by setting a
100 value in a variable in the "match data" data block.
102 13. Be more explicit in pcre_study() instead of relying on "default" for
103 opcodes that mean there is no starting character; this means that when new
104 ones are added and accidentally left out of pcre_study(), testing should
105 pick them up.
107 14. The -s option of pcretest has been documented for ages as being an old
108 synonym of -m (show memory usage). I have changed it to mean "force study
109 for every regex", that is, assume /S for every regex. This is similar to -i
110 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
111 using it. It makes it easier to run collections of tests with and without
112 study enabled, and thereby test pcre_study() more easily. All the standard
113 tests are now run with and without -s (but some patterns can be marked as
114 "never study" - see 20 below).
116 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
117 restoration of the capturing data to the outer values was not happening
118 correctly.
120 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
121 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
122 pattern had matched an empty string, and so incorrectly returned a no
123 match.
125 17. There was optimizing code for the last branch of non-capturing parentheses,
126 and also for the obeyed branch of a conditional subexpression, which used
127 tail recursion to cut down on stack usage. Unfortunately, now that there is
128 the possibility of (*THEN) occurring in these branches, tail recursion is
129 no longer possible because the return has to be checked for (*THEN). These
130 two optimizations have therefore been removed.
132 18. If a pattern containing \R was studied, it was assumed that \R always
133 matched two bytes, thus causing the minimum subject length to be
134 incorrectly computed because \R can also match just one byte.
136 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
137 was incorrectly computed.
139 20. If /S is present twice on a test pattern in pcretest input, it now
140 *disables* studying, thereby overriding the use of -s on the command line
141 (see 14 above). This is necessary for one or two tests to keep the output
142 identical in both cases.
144 21. When (*ACCEPT) was used in an assertion that matched an empty string and
145 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
147 22. When an atomic group that contained a capturing parenthesis was
148 successfully matched, but the branch in which it appeared failed, the
149 capturing was not being forgotten if a higher numbered group was later
150 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
151 group 1 to "a", when in fact it should be unset. This applied to multi-
152 branched capturing and non-capturing groups, repeated or not, and also to
153 positive assertions (capturing in negative assertions does not happen
154 in PCRE) and also to nested atomic groups.
156 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
157 subject after a captured substring, to make it easier to tell which of a
158 number of identical substrings has been captured.
160 24. The way atomic groups are processed by pcre_exec() has been changed so that
161 if they are repeated, backtracking one repetition now resets captured
162 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
163 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
164 "aaa". Previously, it would have been "a". As part of this code
165 refactoring, the way recursive calls are handled has also been changed.
167 25. If an assertion condition captured any substrings, they were not passed
168 back unless some other capturing happened later. For example, if
169 (?(?=(a))a) was matched against "a", no capturing was returned.
171 26. When studying a pattern that contained subroutine calls or assertions,
172 the code for finding the minimum length of a possible match was handling
173 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
174 group 1 called group 2 while simultaneously a separate group 2 called group
175 1). A stack overflow occurred in this case. I have fixed this by limiting
176 the recursion depth to 10.
178 27. Updated RunTest.bat in the distribution to the version supplied by Tom
179 Fortmann. This supports explicit test numbers on the command line, and has
180 argument validation and error reporting.
182 28. An instance of \X with an unlimited repeat could fail if at any point the
183 first character it looked at was a mark character.
185 29. Some minor code refactoring concerning Unicode properties and scripts
186 should reduce the stack requirement of match() slightly.
188 30. Added the '=' option to pcretest to check the setting of unused capturing
189 slots at the end of the pattern, which are documented as being -1, but are
190 not included in the return count.
192 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
193 compiled something random. Now it gives a compile-time error (as does
194 Perl).
196 32. A *MARK encountered during the processing of a positive assertion is now
197 recorded and passed back (compatible with Perl).
199 33. If --only-matching or --colour was set on a pcregrep call whose pattern
200 had alternative anchored branches, the search for a second match in a line
201 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
202 matched the line "0102" twice. The same bug affected patterns that started
203 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
204 twice.
206 34. Previously, PCRE did not allow quantification of assertions. However, Perl
207 does, and because of capturing effects, quantifying parenthesized
208 assertions may at times be useful. Quantifiers are now allowed for
209 parenthesized assertions.
211 35. A minor code tidy in pcre_compile() when checking options for \R usage.
213 36. \g was being checked for fancy things in a character class, when it should
214 just be a literal "g".
216 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
217 appearance of a nested POSIX class supersedes an apparent external class.
218 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
219 unescaped square brackets may also appear as part of class names. For
220 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
221 more like Perl. (But see 8.20/1 above.)
223 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
224 was because it thought it was \N{name}, which is not supported).
226 39. Add minix to OS list not supporting the -S option in pcretest.
228 40. PCRE tries to detect cases of infinite recursion at compile time, but it
229 cannot analyze patterns in sufficient detail to catch mutual recursions
230 such as ((?1))((?2)). There is now a runtime test that gives an error if a
231 subgroup is called recursively as a subpattern for a second time at the
232 same position in the subject string. In previous releases this might have
233 been caught by the recursion limit, or it might have run out of stack.
235 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
236 happen only once. PCRE was, however incorrectly giving a compile time error
237 "recursive call could loop indefinitely" because it cannot analyze the
238 pattern in sufficient detail. The compile time test no longer happens when
239 PCRE is compiling a conditional subpattern, but actual runaway loops are
240 now caught at runtime (see 40 above).
242 42. It seems that Perl allows any characters other than a closing parenthesis
243 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
244 has been changed to be the same.
246 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
247 as not to get warnings when autogen.sh is called. Also changed
248 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
250 44. To help people who use pcregrep to scan files containing exceedingly long
251 lines, the following changes have been made:
253 (a) The default value of the buffer size parameter has been increased from
254 8K to 20K. (The actual buffer used is three times this size.)
256 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
257 PCRE is built.
259 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
260 to be set at run time.
262 (d) Numerical values in pcregrep options can be followed by K or M, for
263 example --buffer-size=50K.
265 (e) If a line being scanned overflows pcregrep's buffer, an error is now
266 given and the return code is set to 2.
268 45. Add a pointer to the latest mark to the callout data block.
270 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
271 partial match of an empty string instead of no match. This was specific to
272 the use of ".".
274 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
275 complete match instead of a partial match. This bug was dependent on both
276 the PCRE_UTF8 and PCRE_DOTALL options being set.
278 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
279 starting byte set, because \b was not being ignored.
282 Version 8.12 15-Jan-2011
283 ------------------------
285 1. Fixed some typos in the markup of the man pages, and wrote a script that
286 checks for such things as part of the documentation building process.
288 2. On a big-endian 64-bit system, pcregrep did not correctly process the
289 --match-limit and --recursion-limit options (added for 8.11). In
290 particular, this made one of the standard tests fail. (The integer value
291 went into the wrong half of a long int.)
293 3. If the --colour option was given to pcregrep with -v (invert match), it
294 did strange things, either producing crazy output, or crashing. It should,
295 of course, ignore a request for colour when reporting lines that do not
296 match.
298 4. Another pcregrep bug caused similar problems if --colour was specified with
299 -M (multiline) and the pattern match finished with a line ending.
301 5. In pcregrep, when a pattern that ended with a literal newline sequence was
302 matched in multiline mode, the following line was shown as part of the
303 match. This seems wrong, so I have changed it.
305 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
306 the check for further matches in the same line (so they could be coloured)
307 to overrun the end of the current line. If another match was found, it was
308 incorrectly shown (and then shown again when found in the next line).
310 7. If pcregrep was compiled under Windows, there was a reference to the
311 function pcregrep_exit() before it was defined. I am assuming this was
312 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
313 reported by a user. I've moved the definition above the reference.
316 Version 8.11 10-Dec-2010
317 ------------------------
319 1. (*THEN) was not working properly if there were untried alternatives prior
320 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
321 backtracked to try for "b" instead of moving to the next alternative branch
322 at the same level (in this case, to look for "c"). The Perl documentation
323 is clear that when (*THEN) is backtracked onto, it goes to the "next
324 alternative in the innermost enclosing group".
326 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
327 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
328 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
329 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
330 (*THEN).
332 3. If \s appeared in a character class, it removed the VT character from
333 the class, even if it had been included by some previous item, for example
334 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
335 of \s, but is part of the POSIX "space" class.)
337 4. A partial match never returns an empty string (because you can always
338 match an empty string at the end of the subject); however the checking for
339 an empty string was starting at the "start of match" point. This has been
340 changed to the "earliest inspected character" point, because the returned
341 data for a partial match starts at this character. This means that, for
342 example, /(?<=abc)def/ gives a partial match for the subject "abc"
343 (previously it gave "no match").
345 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
346 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
347 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
348 has an implication that the given string is incomplete (because a partial
349 match is preferred over a full match). For this reason, these items now
350 give a partial match in this situation. [Aside: previously, the one case
351 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
352 match rather than a full match, which was wrong by the old rules, but is
353 now correct.]
355 6. There was a bug in the handling of #-introduced comments, recognized when
356 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
357 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
358 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
359 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
360 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
361 places in pcre_compile().
363 7. Related to (6) above, when pcre_compile() was skipping #-introduced
364 comments when looking ahead for named forward references to subpatterns,
365 the only newline sequence it recognized was NL. It now handles newlines
366 according to the set newline convention.
368 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
369 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
370 cater for a lack of strerror(). These oversights have been fixed.
372 9. Added --match-limit and --recursion-limit to pcregrep.
374 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
376 11. When the -o option was used, pcregrep was setting a return code of 1, even
377 when matches were found, and --line-buffered was not being honoured.
379 12. Added an optional parentheses number to the -o and --only-matching options
380 of pcregrep.
382 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
383 can match an empty string. The code to do it in pcretest and pcredemo
384 needed fixing:
386 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
387 only one byte after an empty string match just before CRLF (this case
388 just got forgotten; "any" and "anycrlf" were OK).
390 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
391 mode when an empty string match preceded an ASCII character followed by
392 a non-ASCII character. (The code for advancing by one character rather
393 than one byte was nonsense.)
395 (c) The pcredemo.c sample program did not have any code at all to handle
396 the cases when CRLF is a valid newline sequence.
398 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
399 as a starting offset was within the subject string. There is now a new
400 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
401 negative or greater than the length of the string. In order to test this,
402 pcretest is extended to allow the setting of negative starting offsets.
404 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
405 starting offset points to the beginning of a UTF-8 character was
406 unnecessarily clumsy. I tidied it up.
408 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
409 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
411 17. Nobody had reported that the --include_dir option, which was added in
412 release 7.7 should have been called --include-dir (hyphen, not underscore)
413 for compatibility with GNU grep. I have changed it to --include-dir, but
414 left --include_dir as an undocumented synonym, and the same for
415 --exclude-dir, though that is not available in GNU grep, at least as of
416 release 2.5.4.
418 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
419 characters from a string of bytes) have been redefined so as not to use
420 loops, in order to improve performance in some environments. At the same
421 time, I abstracted some of the common code into auxiliary macros to save
422 repetition (this should not affect the compiled code).
424 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
425 compile-time error is now given if \c is not followed by an ASCII
426 character, that is, a byte less than 128. (In EBCDIC mode, the code is
427 different, and any byte value is allowed.)
429 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
430 START_OPTIMIZE option, which is now allowed at compile time - but just
431 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
432 to pcregrep and other applications that have no direct access to PCRE
433 options. The new /Y option in pcretest sets this option when calling
434 pcre_compile().
436 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
437 back references. Groups containing recursive back references were forced to
438 be atomic by that change, but in the case of named groups, the amount of
439 memory required was incorrectly computed, leading to "Failed: internal
440 error: code overflow". This has been fixed.
442 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
443 pcretest.c, to avoid build problems in some Borland environments.
446 Version 8.10 25-Jun-2010
447 ------------------------
449 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
450 THEN.
452 2. (*ACCEPT) was not working when inside an atomic group.
454 3. Inside a character class, \B is treated as a literal by default, but
455 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
456 causes the error). The code is unchanged, but I tidied the documentation.
458 4. Inside a character class, PCRE always treated \R and \X as literals,
459 whereas Perl faults them if its -w option is set. I have changed PCRE so
460 that it faults them when PCRE_EXTRA is set.
462 5. Added support for \N, which always matches any character other than
463 newline. (It is the same as "." when PCRE_DOTALL is not set.)
465 6. When compiling pcregrep with newer versions of gcc which may have
466 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
467 declared with attribute warn_unused_result" were given. Just casting the
468 result to (void) does not stop the warnings; a more elaborate fudge is
469 needed. I've used a macro to implement this.
471 7. Minor change to pcretest.c to avoid a compiler warning.
473 8. Added four artifical Unicode properties to help with an option to make
474 \s etc use properties (see next item). The new properties are: Xan
475 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
477 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
478 use Unicode properties. (*UCP) at the start of a pattern can be used to set
479 this option. Modified pcretest to add /W to test this facility. Added
480 REG_UCP to make it available via the POSIX interface.
482 10. Added --line-buffered to pcregrep.
484 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
485 studied, and the match started with a letter with a code point greater than
486 127 whose first byte was different to the first byte of the other case of
487 the letter, the other case of this starting letter was not recognized
488 (#976).
490 12. If a pattern that was studied started with a repeated Unicode property
491 test, for example, \p{Nd}+, there was the theoretical possibility of
492 setting up an incorrect bitmap of starting bytes, but fortunately it could
493 not have actually happened in practice until change 8 above was made (it
494 added property types that matched character-matching opcodes).
496 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
497 possible starting bytes for non-anchored patterns.
499 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
500 \R, and also a number of cases that involve Unicode properties, both
501 explicit and implicit when PCRE_UCP is set.
503 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
504 input, it could crash or give wrong results if characters with values
505 greater than 0xc0 were present in the subject string. (Detail: it assumed
506 UTF-8 input when processing these items.)
508 16. Added a lot of (int) casts to avoid compiler warnings in systems where
509 size_t is 64-bit (#991).
511 17. Added a check for running out of memory when PCRE is compiled with
512 --disable-stack-for-recursion (#990).
514 18. If the last data line in a file for pcretest does not have a newline on
515 the end, a newline was missing in the output.
517 19. The default pcre_chartables.c file recognizes only ASCII characters (values
518 less than 128) in its various bitmaps. However, there is a facility for
519 generating tables according to the current locale when PCRE is compiled. It
520 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
521 characters, are recognized by isspace() and therefore were getting set in
522 these tables, and indeed these tables seem to approximate to ISO 8859. This
523 caused a problem in UTF-8 mode when pcre_study() was used to create a list
524 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
525 which of course cannot start UTF-8 characters. I have changed the code so
526 that only real ASCII characters (less than 128) and the correct starting
527 bytes for UTF-8 encodings are set for characters greater than 127 when in
528 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
529 altogether.)
531 20. Added the /T option to pcretest so as to be able to run tests with non-
532 standard character tables, thus making it possible to include the tests
533 used for 19 above in the standard set of tests.
535 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
536 reference to a subpattern the other side of a comment that contains an
537 opening parenthesis caused either an internal compiling error, or a
538 reference to the wrong subpattern.
541 Version 8.02 19-Mar-2010
542 ------------------------
544 1. The Unicode data tables have been updated to Unicode 5.2.0.
546 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
547 configured.
549 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
550 original author of that file, following a query about its status.
552 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
553 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
555 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
556 quantifier applied to a forward-referencing subroutine call, could compile
557 incorrect code or give the error "internal error: previously-checked
558 referenced subpattern not found".
560 6. Both MS Visual Studio and Symbian OS have problems with initializing
561 variables to point to external functions. For these systems, therefore,
562 pcre_malloc etc. are now initialized to local functions that call the
563 relevant global functions.
565 7. There were two entries missing in the vectors called coptable and poptable
566 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
567 I've fixed the data, and added a kludgy way of testing at compile time that
568 the lengths are correct (equal to the number of opcodes).
570 8. Following on from 7, I added a similar kludge to check the length of the
571 eint vector in pcreposix.c.
573 9. Error texts for pcre_compile() are held as one long string to avoid too
574 much relocation at load time. To find a text, the string is searched,
575 counting zeros. There was no check for running off the end of the string,
576 which could happen if a new error number was added without updating the
577 string.
579 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
581 11. \K was not working if it appeared in an atomic group or in a group that
582 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
583 \K is "not well defined" if used in an assertion. PCRE now accepts it if
584 the assertion is positive, but not if it is negative.
586 12. Change 11 fortuitously reduced the size of the stack frame used in the
587 "match()" function of pcre_exec.c by one pointer. Forthcoming
588 implementation of support for (*MARK) will need an extra pointer on the
589 stack; I have reserved it now, so that the stack frame size does not
590 decrease.
592 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
593 item in branch that calls a recursion is a subroutine call - as in the
594 second branch in the above example - was incorrectly given the compile-
595 time error "recursive call could loop indefinitely" because pcre_compile()
596 was not correctly checking the subroutine for matching a non-empty string.
598 14. The checks for overrunning compiling workspace could trigger after an
599 overrun had occurred. This is a "should never occur" error, but it can be
600 triggered by pathological patterns such as hundreds of nested parentheses.
601 The checks now trigger 100 bytes before the end of the workspace.
603 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
606 Version 8.01 19-Jan-2010
607 ------------------------
609 1. If a pattern contained a conditional subpattern with only one branch (in
610 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
611 computed the wrong minimum data length (which is of course zero for such
612 subpatterns). This could cause incorrect "no match" results.
614 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
615 the pattern is reset in the first branch, pcre_compile() failed with
616 "internal error: code overflow at offset...". This happened only when
617 the reset was to the original external option setting. (An optimization
618 abstracts leading options settings into an external setting, which was the
619 cause of this.)
621 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
622 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
623 assertion pattern did not match (meaning that the assertion was true), it
624 was incorrectly treated as false if the SKIP had been reached during the
625 matching. This also applied to assertions used as conditions.
627 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
628 assertion subpattern, including such a pattern used as a condition,
629 unpredictable results occurred, instead of the error return
632 5. The C++ GlobalReplace function was not working like Perl for the special
633 situation when an empty string is matched. It now does the fancy magic
634 stuff that is necessary.
636 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
637 removed. (These were left over from very, very early versions of PCRE.)
639 7. Some cosmetic changes to the code to make life easier when compiling it
640 as part of something else:
642 (a) Change DEBUG to PCRE_DEBUG.
644 (b) In pcre_compile(), rename the member of the "branch_chain" structure
645 called "current" as "current_branch", to prevent a collision with the
646 Linux macro when compiled as a kernel module.
648 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
649 prevent a collision with the Linux macro when compiled as a kernel
650 module.
652 8. In pcre_compile() there are some checks for integer overflows that used to
653 cast potentially large values to (double). This has been changed to that
654 when building, a check for int64_t is made, and if it is found, it is used
655 instead, thus avoiding the use of floating point arithmetic. (There is no
656 other use of FP in PCRE.) If int64_t is not found, the fallback is to
657 double.
659 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
660 2005 (difference between two addresses compared to an unsigned value).
662 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
663 custom one, because of the following reported problem in Windows:
665 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
666 under Win32.
667 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
668 therefore missing the function definition.
669 - The compiler thus generates a "C" signature for the test function.
670 - The linker fails to find the "C" function.
671 - PCRE fails to configure if asked to do so against libbz2.
673 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
674 messages were output:
676 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
677 rerunning libtoolize, to keep the correct libtool macros in-tree.
678 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
680 I have done both of these things.
682 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
683 most of the time, it *can* run out if it is given a pattern that contains a
684 runaway infinite recursion. I updated the discussion in the pcrestack man
685 page.
687 13. Now that we have gone to the x.xx style of version numbers, the minor
688 version may start with zero. Using 08 or 09 is a bad idea because users
689 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
690 interpreted as invalid octal numbers. I've updated the previous comment in
691 configure.ac, and also added a check that gives an error if 08 or 09 are
692 used.
694 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
695 causing partial matching to fail when the end of the subject matched \W
696 in a UTF-8 pattern where \W was quantified with a minimum of 3.
698 15. There were some discrepancies between the declarations in pcre_internal.h
699 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
700 their definitions. The declarations used "const uschar *" and the
701 definitions used USPTR. Even though USPTR is normally defined as "const
702 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
703 reported that: "This difference in casting confuses some C++ compilers, for
704 example, SunCC recognizes above declarations as different functions and
705 generates broken code for hbpcre." I have changed the declarations to use
706 USPTR.
708 16. GNU libtool is named differently on some systems. The autogen.sh script now
709 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
710 (FreeBSD).
712 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
713 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
714 comment: "Figure out how to create a longlong from a string: strtoll and
715 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
716 instance, but it only takes 2 args instead of 3!"
718 18. A subtle bug concerned with back references has been fixed by a change of
719 specification, with a corresponding code fix. A pattern such as
720 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
721 refers, was giving matches when it shouldn't. For example, xa=xaaa would
722 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
723 same bug. Such groups have to be quantified to be useful, or contained
724 inside another quantified group. (If there's no repetition, the reference
725 can never match.) The problem arises because, having left the group and
726 moved on to the rest of the pattern, a later failure that backtracks into
727 the group uses the captured value from the final iteration of the group
728 rather than the correct earlier one. I have fixed this in PCRE by forcing
729 any group that contains a reference to itself to be an atomic group; that
730 is, there cannot be any backtracking into it once it has completed. This is
731 similar to recursive and subroutine calls.
734 Version 8.00 19-Oct-09
735 ----------------------
737 1. The table for translating pcre_compile() error codes into POSIX error codes
738 was out-of-date, and there was no check on the pcre_compile() error code
739 being within the table. This could lead to an OK return being given in
740 error.
742 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
743 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
744 in a Windows environment.
746 3. The pcregrep --count option prints the count for each file even when it is
747 zero, as does GNU grep. However, pcregrep was also printing all files when
748 --files-with-matches was added. Now, when both options are given, it prints
749 counts only for those files that have at least one match. (GNU grep just
750 prints the file name in this circumstance, but including the count seems
751 more useful - otherwise, why use --count?) Also ensured that the
752 combination -clh just lists non-zero counts, with no names.
754 4. The long form of the pcregrep -F option was incorrectly implemented as
755 --fixed_strings instead of --fixed-strings. This is an incompatible change,
756 but it seems right to fix it, and I didn't think it was worth preserving
757 the old behaviour.
759 5. The command line items --regex=pattern and --regexp=pattern were not
760 recognized by pcregrep, which required --regex pattern or --regexp pattern
761 (with a space rather than an '='). The man page documented the '=' forms,
762 which are compatible with GNU grep; these now work.
764 6. No libpcreposix.pc file was created for pkg-config; there was just
765 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
767 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
768 when UCP support is not needed, by modifying the Python script that
769 generates it from Unicode data files. This should not matter if the module
770 is correctly used as a library, but I received one complaint about 50K of
771 unwanted data. My guess is that the person linked everything into his
772 program rather than using a library. Anyway, it does no harm.
774 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
775 was a minimum greater than 1 for a wide character in a possessive
776 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
777 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
778 character. Chaos in the form of incorrect output or a compiling loop could
779 result.
781 9. The restrictions on what a pattern can contain when partial matching is
782 requested for pcre_exec() have been removed. All patterns can now be
783 partially matched by this function. In addition, if there are at least two
784 slots in the offset vector, the offset of the earliest inspected character
785 for the match and the offset of the end of the subject are set in them when
786 PCRE_ERROR_PARTIAL is returned.
788 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
789 synonymous with PCRE_PARTIAL, for backwards compatibility, and
790 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
791 and may be more useful for multi-segment matching.
793 11. Partial matching with pcre_exec() is now more intuitive. A partial match
794 used to be given if ever the end of the subject was reached; now it is
795 given only if matching could not proceed because another character was
796 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
797 string "Z", which now yields "no match" instead of "partial match". In the
798 case of pcre_dfa_exec(), "no match" is given if every matching path for the
799 final character ended with (*FAIL).
801 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
802 if the pattern had a "must contain" character that was already found in the
803 earlier partial match, unless partial matching was again requested. For
804 example, with the pattern /dog.(body)?/, the "must contain" character is
805 "g". If the first part-match was for the string "dog", restarting with
806 "sbody" failed. This bug has been fixed.
808 13. The string returned by pcre_dfa_exec() after a partial match has been
809 changed so that it starts at the first inspected character rather than the
810 first character of the match. This makes a difference only if the pattern
811 starts with a lookbehind assertion or \b or \B (\K is not supported by
812 pcre_dfa_exec()). It's an incompatible change, but it makes the two
813 matching functions compatible, and I think it's the right thing to do.
815 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
816 so that the demonstration program is easily available in environments where
817 PCRE has not been installed from source.
819 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
820 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
821 library.
823 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
824 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
825 is not the first non-POSIX option to be added. Clearly some people find
826 these options useful.
828 17. If a caller to the POSIX matching function regexec() passes a non-zero
829 value for nmatch with a NULL value for pmatch, the value of
830 nmatch is forced to zero.
832 18. RunGrepTest did not have a test for the availability of the -u option of
833 the diff command, as RunTest does. It now checks in the same way as
834 RunTest, and also checks for the -b option.
836 19. If an odd number of negated classes containing just a single character
837 interposed, within parentheses, between a forward reference to a named
838 subpattern and the definition of the subpattern, compilation crashed with
839 an internal error, complaining that it could not find the referenced
840 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
841 [The bug was that it was starting one character too far in when skipping
842 over the character class, thus treating the ] as data rather than
843 terminating the class. This meant it could skip too much.]
845 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
846 /g option in pcretest when the pattern contains \K, which makes it possible
847 to have an empty string match not at the start, even when the pattern is
848 anchored. Updated pcretest and pcredemo to use this option.
850 21. If the maximum number of capturing subpatterns in a recursion was greater
851 than the maximum at the outer level, the higher number was returned, but
852 with unset values at the outer level. The correct (outer level) value is
853 now given.
855 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
856 PCRE did not set those parentheses (unlike Perl). I have now found a way to
857 make it do so. The string so far is captured, making this feature
858 compatible with Perl.
860 23. The tests have been re-organized, adding tests 11 and 12, to make it
861 possible to check the Perl 5.10 features against Perl 5.10.
863 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
864 pattern matches a fixed length string. PCRE did not allow this; now it
865 does. Neither allows recursion.
867 25. I finally figured out how to implement a request to provide the minimum
868 length of subject string that was needed in order to match a given pattern.
869 (It was back references and recursion that I had previously got hung up
870 on.) This code has now been added to pcre_study(); it finds a lower bound
871 to the length of subject needed. It is not necessarily the greatest lower
872 bound, but using it to avoid searching strings that are too short does give
873 some useful speed-ups. The value is available to calling programs via
874 pcre_fullinfo().
876 26. While implementing 25, I discovered to my embarrassment that pcretest had
877 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
878 study optimizations had never been tested with that matching function.
879 Oops. What is worse, even when it was passed study data, there was a bug in
880 pcre_dfa_exec() that meant it never actually used it. Double oops. There
881 were also very few tests of studied patterns with pcre_dfa_exec().
883 27. If (?| is used to create subpatterns with duplicate numbers, they are now
884 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
885 on the other side of the coin, they are no longer allowed to have different
886 names, because these cannot be distinguished in PCRE, and this has caused
887 confusion. (This is a difference from Perl.)
889 28. When duplicate subpattern names are present (necessarily with different
890 numbers, as required by 27 above), and a test is made by name in a
891 conditional pattern, either for a subpattern having been matched, or for
892 recursion in such a pattern, all the associated numbered subpatterns are
893 tested, and the overall condition is true if the condition is true for any
894 one of them. This is the way Perl works, and is also more like the way
895 testing by number works.
898 Version 7.9 11-Apr-09
899 ---------------------
901 1. When building with support for bzlib/zlib (pcregrep) and/or readline
902 (pcretest), all targets were linked against these libraries. This included
903 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
904 libraries. This caused unwanted dependencies to be created. This problem
905 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
906 pcretest is linked with readline.
908 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
909 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
910 moved outside it again, because FALSE and TRUE are already defined in AIX,
911 but BOOL is not.
913 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
914 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
916 4. The pcregrep documentation said spaces were inserted as well as colons (or
917 hyphens) following file names and line numbers when outputting matching
918 lines. This is not true; no spaces are inserted. I have also clarified the
919 wording for the --colour (or --color) option.
921 5. In pcregrep, when --colour was used with -o, the list of matching strings
922 was not coloured; this is different to GNU grep, so I have changed it to be
923 the same.
925 6. When --colo(u)r was used in pcregrep, only the first matching substring in
926 each matching line was coloured. Now it goes on to look for further matches
927 of any of the test patterns, which is the same behaviour as GNU grep.
929 7. A pattern that could match an empty string could cause pcregrep to loop; it
930 doesn't make sense to accept an empty string match in pcregrep, so I have
931 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
932 seems to be how GNU grep behaves.
934 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
935 start or after a newline", because the conditional assertion was not being
936 correctly handled. The rule now is that both the assertion and what follows
937 in the first alternative must satisfy the test.
939 9. If auto-callout was enabled in a pattern with a conditional group whose
940 condition was an assertion, PCRE could crash during matching, both with
941 pcre_exec() and pcre_dfa_exec().
943 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
944 used for matching.
946 11. Unicode property support in character classes was not working for
947 characters (bytes) greater than 127 when not in UTF-8 mode.
949 12. Added the -M command line option to pcretest.
951 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
953 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
955 16. Added comments and documentation about mis-use of no_arg in the C++
956 wrapper.
958 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
959 from Martin Jerabek that uses macro names for all relevant character and
960 string constants.
962 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
963 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
964 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
965 these, but not everybody uses configure.
967 19. A conditional group that had only one branch was not being correctly
968 recognized as an item that could match an empty string. This meant that an
969 enclosing group might also not be so recognized, causing infinite looping
970 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
971 with the subject "ab", where knowledge that the repeated group can match
972 nothing is needed in order to break the loop.
974 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
975 exec(), but without supplying a callout function, matching went wrong.
977 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
978 leak if the size of the offset vector was greater than 30. When the vector
979 is smaller, the saved offsets during recursion go onto a local stack
980 vector, but for larger vectors malloc() is used. It was failing to free
981 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
982 error, in fact).
984 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
985 heapframe that is used only when UTF-8 support is enabled. This caused no
986 problem, but was untidy.
988 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
989 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
990 included within another project.
992 24. Steven Van Ingelgem's patches to add more options to the CMake support,
993 slightly modified by me:
995 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
996 not building pcregrep.
998 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
999 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1001 25. Forward references, both numeric and by name, in patterns that made use of
1002 duplicate group numbers, could behave incorrectly or give incorrect errors,
1003 because when scanning forward to find the reference group, PCRE was not
1004 taking into account the duplicate group numbers. A pattern such as
1005 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1007 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1008 the feature of a custom pointer more persuasive (as requested by a user).
1010 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1011 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1012 pre-defined.
1014 28. Added support for (*UTF8) at the start of a pattern.
1016 29. Arrange for flags added by the "release type" setting in CMake to be shown
1017 in the configuration summary.
1020 Version 7.8 05-Sep-08
1021 ---------------------
1023 1. Replaced UCP searching code with optimized version as implemented for Ad
1024 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1025 stage table and inline lookup instead of a function, giving speed ups of 2
1026 to 5 times on some simple patterns that I tested. Permission was given to
1027 distribute the MultiStage2.py script that generates the tables (it's not in
1028 the tarball, but is in the Subversion repository).
1030 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1031 scripts.
1033 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1034 a group with a zero qualifier. The result of the study could be incorrect,
1035 or the function might crash, depending on the pattern.
1037 4. Caseless matching was not working for non-ASCII characters in back
1038 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1039 It now works when Unicode Property Support is available.
1041 5. In pcretest, an escape such as \x{de} in the data was always generating
1042 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1043 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1044 truncation.
1046 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1048 7. Added two (int) casts to pcregrep when printing the difference of two
1049 pointers, in case they are 64-bit values.
1051 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1052 test 2 if it fails.
1054 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1055 and a #define of that name to empty if it is not externally set. This is to
1056 allow users of MSVC to set it if necessary.
1058 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1059 the convenience functions in the pcre_get.c source file.
1061 11. An option change at the start of a pattern that had top-level alternatives
1062 could cause overwriting and/or a crash. This command provoked a crash in
1063 some environments:
1065 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1067 This potential security problem was recorded as CVE-2008-2371.
1069 12. For a pattern where the match had to start at the beginning or immediately
1070 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1071 pcre_dfa_exec() could read past the end of the passed subject if there was
1072 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1073 pcretest so that it places the subject at the end of its malloc-ed buffer.
1075 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1076 exec() might read past the end of the data buffer in UTF-8 mode.
1078 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1079 the data contained the byte 0x85 as part of a UTF-8 character within its
1080 first line. This applied both to normal and DFA matching.
1082 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1083 /^[^d]*?$/8 failed to match "abc".
1085 16. Added a missing copyright notice to pcrecpp_internal.h.
1087 17. Make it more clear in the documentation that values returned from
1088 pcre_exec() in ovector are byte offsets, not character counts.
1090 18. Tidied a few places to stop certain compilers from issuing warnings.
1092 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1093 supplied by Stefan Weber. I made a further small update for 7.8 because
1094 there is a change of source arrangements: the pcre_searchfuncs.c module is
1095 replaced by pcre_ucd.c.
1098 Version 7.7 07-May-08
1099 ---------------------
1101 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1102 a string to a long long, pretend we don't even have a long long." This is
1103 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1105 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1106 pre-7.6 versions, which defined a global no_arg variable instead of putting
1107 it in the RE class. (See also #8 below.)
1109 3. Remove a line of dead code, identified by coverity and reported by Nuno
1110 Lopes.
1112 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1114 (1) The include/exclude patterns were being applied to the whole pathnames
1115 of files, instead of just to the final components.
1117 (2) If there was more than one level of directory, the subdirectories were
1118 skipped unless they satisfied the include/exclude conditions. This is
1119 inconsistent with GNU grep (and could even be seen as contrary to the
1120 pcregrep specification - which I improved to make it absolutely clear).
1121 The action now is always to scan all levels of directory, and just
1122 apply the include/exclude patterns to regular files.
1124 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1125 --exclude_dir in the tests to avoid scanning .svn directories.
1127 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1128 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1129 doesn't support NULs in patterns.
1131 7. Added some missing "const"s to declarations of static tables in
1132 pcre_compile.c and pcre_dfa_exec.c.
1134 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1135 caused by fix #2 above. (Subsequently also a second patch to fix the
1136 first patch. And a third patch - this was a messy problem.)
1138 9. Applied Craig's patch to remove the use of push_back().
1140 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1141 matching function regexec().
1143 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1144 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1145 references. PCRE supports relative numbers with this syntax (I don't think
1146 Oniguruma does).
1148 12. Previously, a group with a zero repeat such as (...){0} was completely
1149 omitted from the compiled regex. However, this means that if the group
1150 was called as a subroutine from elsewhere in the pattern, things went wrong
1151 (an internal error was given). Such groups are now left in the compiled
1152 pattern, with a new opcode that causes them to be skipped at execution
1153 time.
1155 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1156 to the way PCRE behaves:
1158 (a) A lone ] character is dis-allowed (Perl treats it as data).
1160 (b) A back reference to an unmatched subpattern matches an empty string
1161 (Perl fails the current match path).
1163 (c) A data ] in a character class must be notated as \] because if the
1164 first data character in a class is ], it defines an empty class. (In
1165 Perl it is not possible to have an empty class.) The empty class []
1166 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1167 The negative empty class [^] matches any one character, independently
1168 of the DOTALL setting.
1170 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1171 non-existent subpattern following a character class starting with ']' and
1172 containing () gave an internal compiling error instead of "reference to
1173 non-existent subpattern". Fortunately, when the pattern did exist, the
1174 compiled code was correct. (When scanning forwards to check for the
1175 existencd of the subpattern, it was treating the data ']' as terminating
1176 the class, so got the count wrong. When actually compiling, the reference
1177 was subsequently set up correctly.)
1179 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1180 it was being rejected as not supported by pcre_dfa_exec(), even though
1181 other assertions are supported. I have made pcre_dfa_exec() support
1182 (*FAIL).
1184 16. The implementation of 13c above involved the invention of a new opcode,
1185 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1186 cannot be changed at match time, I realized I could make a small
1187 improvement to matching performance by compiling OP_ALLANY instead of
1188 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1189 on the OP_ANY path.
1191 17. Compiling pcretest on Windows with readline support failed without the
1192 following two fixes: (1) Make the unistd.h include conditional on
1193 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1195 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1196 ncurses library to be included for pcretest when ReadLine support is
1197 requested, but also to allow for it to be overridden. This patch came from
1198 Daniel Bergström.
1200 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1201 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1202 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1203 spotting this.
1206 Version 7.6 28-Jan-08
1207 ---------------------
1209 1. A character class containing a very large number of characters with
1210 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1211 overflow.
1213 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1214 HAVE_LONG_LONG is not defined.
1216 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1217 bring them up to date and include new features. This patch includes:
1219 - Fixed PH's badly added libz and libbz2 support.
1220 - Fixed a problem with static linking.
1221 - Added pcredemo. [But later removed - see 7 below.]
1222 - Fixed dftables problem and added an option.
1223 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1225 - Added readline support for pcretest.
1226 - Added an listing of the option settings after cmake has run.
1228 4. A user submitted a patch to Makefile that makes it easy to create
1229 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1230 Makefile.am that cause it to include this special target, without
1231 affecting anything else. Note that the same mingw target plus all
1232 the other distribution libraries and programs are now supported
1233 when configuring with CMake (see 6 below) instead of with
1234 Configure/Make.
1236 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1237 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1238 exported in the Windows port". It has not yet been confirmed that the patch
1239 solves the problem, but it does no harm.
1241 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1242 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1243 with CMake, and also correct the comment about stack recursion.
1245 7. Remove the automatic building of pcredemo from the ./configure system and
1246 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1247 of a program that users should build themselves after PCRE is installed, so
1248 building it automatically is not really right. What is more, it gave
1249 trouble in some build environments.
1251 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1254 Version 7.5 10-Jan-08
1255 ---------------------
1257 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1258 values in parens when parsing an RE using the C++ wrapper."
1260 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1261 Characters greater than 255 were excluded from the class instead of being
1262 included.
1264 3. The same bug as (2) above applied to negated POSIX classes such as
1265 [:^space:].
1267 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1268 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1269 I have changed it.
1271 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1272 first named subpattern) and a construct such as (?&a) would reference the
1273 first named subpattern whose name started with "a" (in other words, the
1274 length check was missing). Both these problems are fixed. "Subpattern name
1275 expected" is now given for (?&) (a zero-length name), and this patch also
1276 makes it give the same error for \k'' (previously it complained that that
1277 was a reference to a non-existent subpattern).
1279 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1280 this is right because (?- can be followed by option settings as well as by
1281 digits. I have, however, made the messages clearer.
1283 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1284 than the number used in the conditional) now cause a compile-time error.
1285 This is actually not compatible with Perl, which accepts such patterns, but
1286 treats the conditional as always being FALSE (as PCRE used to), but it
1287 seems to me that giving a diagnostic is better.
1289 8. Change "alphameric" to the more common word "alphanumeric" in comments
1290 and messages.
1292 9. Fix two occurrences of "backslash" in comments that should have been
1293 "backspace".
1295 10. Remove two redundant lines of code that can never be obeyed (their function
1296 was moved elsewhere).
1298 11. The program that makes PCRE's Unicode character property table had a bug
1299 which caused it to generate incorrect table entries for sequences of
1300 characters that have the same character type, but are in different scripts.
1301 It amalgamated them into a single range, with the script of the first of
1302 them. In other words, some characters were in the wrong script. There were
1303 thirteen such cases, affecting characters in the following ranges:
1305 U+002b0 - U+002c1
1306 U+0060c - U+0060d
1307 U+0061e - U+00612
1308 U+0064b - U+0065e
1309 U+0074d - U+0076d
1310 U+01800 - U+01805
1311 U+01d00 - U+01d77
1312 U+01d9b - U+01dbf
1313 U+0200b - U+0200f
1314 U+030fc - U+030fe
1315 U+03260 - U+0327f
1316 U+0fb46 - U+0fbb1
1317 U+10450 - U+1049d
1319 12. The -o option (show only the matching part of a line) for pcregrep was not
1320 compatible with GNU grep in that, if there was more than one match in a
1321 line, it showed only the first of them. It now behaves in the same way as
1322 GNU grep.
1324 13. If the -o and -v options were combined for pcregrep, it printed a blank
1325 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1326 does the same. The return code can be used to tell if there were any
1327 non-matching lines.
1329 14. Added --file-offsets and --line-offsets to pcregrep.
1331 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1332 infinitely looping recursion. The bug was that positive lookaheads were not
1333 being skipped when checking for a possible empty match (negative lookaheads
1334 and both kinds of lookbehind were skipped).
1336 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1337 inclusion of <windows.h> to before rather than after the definition of
1338 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1340 17. Specifying a possessive quantifier with a specific limit for a Unicode
1341 character property caused pcre_compile() to compile bad code, which led at
1342 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1343 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1344 caused the error; without that there was no problem.
1346 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1348 19. Added --enable-pcretest-libreadline.
1350 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1351 RE::GlobalReplace(). As a result, the number of replacements returned was
1352 double what it should be. I removed one of the increments, but Craig sent a
1353 later patch that removed the other one (the right fix) and added unit tests
1354 that check the return values (which was not done before).
1356 21. Several CMake things:
1358 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1359 the names libpcre and libpcreposix, not just pcre and pcreposix.
1361 (2) The above change means that pcretest and pcregrep are now correctly
1362 linked with the newly-built libraries, not previously installed ones.
1366 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1367 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1368 UTF-8 newline character). The key issue is that the pattern starts .*;
1369 this means that the match must be either at the beginning, or after a
1370 newline. The bug was in the code for advancing after a failed match and
1371 checking that the new position followed a newline. It was not taking
1372 account of UTF-8 characters correctly.
1374 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1375 character classes. PCRE was not treating the sequence [:...:] as a
1376 character class unless the ... were all letters. Perl, however, seems to
1377 allow any characters between [: and :], though of course it rejects as
1378 unknown any "names" that contain non-letters, because all the known class
1379 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1380 for example, whereas PCRE did not - it did not recognize a POSIX character
1381 class. This seemed a bit dangerous, so the code has been changed to be
1382 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1383 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1384 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1385 Perl does, and where it didn't before.
1387 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1388 Windows environments %n is disabled by default.
1391 Version 7.4 21-Sep-07
1392 ---------------------
1394 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1395 means that a class such as [\s] counted as "explicit reference to CR or
1396 LF". That isn't really right - the whole point of the change was to try to
1397 help when there was an actual mention of one of the two characters. So now
1398 the change happens only if \r or \n (or a literal CR or LF) character is
1399 encountered.
1401 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1402 of both had grown to the point where there were only 3 bits left.
1403 Fortunately, there was spare space in the data structure, and so I have
1404 moved the internal flags into a new 16-bit field to free up more option
1405 bits.
1407 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1408 but did not set the internal JCHANGED flag - either of these is enough to
1409 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1410 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1411 start sets both bits.
1413 4. Added options (at build time, compile time, exec time) to change \R from
1414 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1416 5. doc/pcresyntax.html was missing from the distribution.
1418 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1419 compatibility, even though it is no longer used.
1421 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1422 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1423 windows.h file is present (where different names are used). [This was
1424 reversed later after testing - see 16 below.]
1426 8. Changed all #include <config.h> to #include "config.h". There were also
1427 some further <pcre.h> cases that I changed to "pcre.h".
1429 9. When pcregrep was used with the --colour option, it missed the line ending
1430 sequence off the lines that it output.
1432 10. It was pointed out to me that arrays of string pointers cause lots of
1433 relocations when a shared library is dynamically loaded. A technique of
1434 using a single long string with a table of offsets can drastically reduce
1435 these. I have refactored PCRE in four places to do this. The result is
1436 dramatic:
1438 Originally: 290
1439 After changing UCP table: 187
1440 After changing error message table: 43
1441 After changing table of "verbs" 36
1442 After changing table of Posix names 22
1444 Thanks to the folks working on Gregex for glib for this insight.
1446 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1447 unicode-properties was also set.
1449 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1451 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1452 checked only for CRLF.
1454 14. Added casts to pcretest.c to avoid compiler warnings.
1456 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1458 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1459 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1460 entirely. This removes changes made in 7 above.
1462 17. The CMake files have been updated, and there is now more information about
1463 building with CMake in the NON-UNIX-USE document.
1466 Version 7.3 28-Aug-07
1467 ---------------------
1469 1. In the rejigging of the build system that eventually resulted in 7.1, the
1470 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1471 brackets there is not right, since it causes compilers to look for an
1472 installed pcre.h, not the version that is in the source that is being
1473 compiled (which of course may be different). I have changed it back to:
1475 #include "pcre.h"
1477 I have a vague recollection that the change was concerned with compiling in
1478 different directories, but in the new build system, that is taken care of
1479 by the VPATH setting the Makefile.
1481 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1482 when the subject happened to end in the byte 0x85 (e.g. if the last
1483 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1484 characters but of course it shouldn't be taken as a newline when it is part
1485 of another character. The bug was that, for an unlimited repeat of . in
1486 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1487 characters when looking for a newline.
1489 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1491 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1492 in debug output.
1494 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1495 long printing in the pcrecpp unittest when running under MinGW.
1497 6. ESC_K was left out of the EBCDIC table.
1499 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1500 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1501 limit also applies to "virtual nesting" when a pattern is recursive, and in
1502 this case 1000 isn't so big. I have been able to remove this limit at the
1503 expense of backing off one optimization in certain circumstances. Normally,
1504 when pcre_exec() would call its internal match() function recursively and
1505 immediately return the result unconditionally, it uses a "tail recursion"
1506 feature to save stack. However, when a subpattern that can match an empty
1507 string has an unlimited repetition quantifier, it no longer makes this
1508 optimization. That gives it a stack frame in which to save the data for
1509 checking that an empty string has been matched. Previously this was taken
1510 from the 1000-entry workspace that had been reserved. So now there is no
1511 explicit limit, but more stack is used.
1513 8. Applied Daniel's patches to solve problems with the import/export magic
1514 syntax that is required for Windows, and which was going wrong for the
1515 pcreposix and pcrecpp parts of the library. These were overlooked when this
1516 problem was solved for the main library.
1518 9. There were some crude static tests to avoid integer overflow when computing
1519 the size of patterns that contain repeated groups with explicit upper
1520 limits. As the maximum quantifier is 65535, the maximum group length was
1521 set at 30,000 so that the product of these two numbers did not overflow a
1522 32-bit integer. However, it turns out that people want to use groups that
1523 are longer than 30,000 bytes (though not repeat them that many times).
1524 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1525 made it possible to implement the integer overflow checks in a much more
1526 dynamic way, which I have now done. The artificial limitation on group
1527 length has been removed - we now have only the limit on the total length of
1528 the compiled pattern, which depends on the LINK_SIZE setting.
1530 10. Fixed a bug in the documentation for get/copy named substring when
1531 duplicate names are permitted. If none of the named substrings are set, the
1532 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1533 empty string.
1535 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1536 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1537 because the ] is interpreted as the first data character and the
1538 terminating ] is not found. PCRE has been made compatible with Perl in this
1539 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1540 cause memory overwriting.
1542 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1543 string has been matched (to stop an infinite loop). It was not recognizing
1544 a conditional subpattern that could match an empty string if that
1545 subpattern was within another subpattern. For example, it looped when
1546 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1547 condition was not nested. This bug has been fixed.
1549 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1550 past the start of the subject in the presence of bytes with the top bit
1551 set, for example "\x8aBCD".
1553 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1554 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1556 14. Optimized (?!) to (*FAIL).
1558 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1559 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1560 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1561 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1562 does: it's just the validity check that is more restrictive.
1564 16. Inserted checks for integer overflows during escape sequence (backslash)
1565 processing, and also fixed erroneous offset values for syntax errors during
1566 backslash processing.
1568 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1569 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1571 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1572 caused an overrun.
1574 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1575 something other than just ASCII characters) inside a group that had an
1576 unlimited repeat caused a loop at compile time (while checking to see
1577 whether the group could match an empty string).
1579 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1580 [\P{Any}] did so. (Error in the code for printing property names.)
1582 21. An orphan \E inside a character class could cause a crash.
1584 22. A repeated capturing bracket such as (A)? could cause a wild memory
1585 reference during compilation.
1587 23. There are several functions in pcre_compile() that scan along a compiled
1588 expression for various reasons (e.g. to see if it's fixed length for look
1589 behind). There were bugs in these functions when a repeated \p or \P was
1590 present in the pattern. These operators have additional parameters compared
1591 with \d, etc, and these were not being taken into account when moving along
1592 the compiled data. Specifically:
1594 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1595 length.
1597 (b) An item such as \pL+ within a repeated group could cause crashes or
1598 loops.
1600 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1601 "reference to non-existent subpattern" error.
1603 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1605 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1606 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1608 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1610 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1611 character were causing crashes (broken optimization).
1613 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1614 \p or \P) caused a compile-time loop.
1616 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1617 break. For example, the unstudied pattern [\r\n]A does not match the string
1618 "\r\nA" because change 7.0/46 below moves the current point on by two
1619 characters after failing to match at the start. However, the pattern \nA
1620 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1621 the same is true. There doesn't seem any very clean way out of this, but
1622 what I have chosen to do makes the common cases work: PCRE now takes note
1623 of whether there can be an explicit match for \r or \n anywhere in the
1624 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1625 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1626 pattern has explicit CR or LF references.
1628 29. Added (*CR) etc for changing newline setting at start of pattern.
1631 Version 7.2 19-Jun-07
1632 ---------------------
1634 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1635 which is apparently normally available under Windows.
1637 2. Re-jig the pcregrep tests with different newline settings in an attempt
1638 to make them independent of the local environment's newline setting.
1640 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1642 4. Some of the "internals" tests were previously cut out when the link size
1643 was not 2, because the output contained actual offsets. The recent new
1644 "Z" feature of pcretest means that these can be cut out, making the tests
1645 usable with all link sizes.
1647 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1648 stack recursion. This gives a massive performance boost under BSD, but just
1649 a small improvement under Linux. However, it saves one field in the frame
1650 in all cases.
1652 6. Added more features from the forthcoming Perl 5.10:
1654 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1655 recursion call. It refers to the nth most recently opened parentheses.
1657 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1658 to be opened parentheses.
1660 (c) Conditions that refer to capturing parentheses can be specified
1661 relatively, for example, (?(-2)... or (?(+3)...
1663 (d) \K resets the start of the current match so that everything before
1664 is not part of it.
1666 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1668 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1669 reference syntax.
1671 (g) (?| introduces a group in which the numbering of parentheses in each
1672 alternative starts with the same number.
1674 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1676 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1679 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1680 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1681 for detecting groups that can match an empty string.
1683 9. A pattern with a very large number of alternatives (more than several
1684 hundred) was running out of internal workspace during the pre-compile
1685 phase, where pcre_compile() figures out how much memory will be needed. A
1686 bit of new cunning has reduced the workspace needed for groups with
1687 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1688 workspace instead of running out of the 4096 that are available.
1690 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1692 11. Applied patch from Google to remove an optimization that didn't quite work.
1693 The report of the bug said:
1695 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1696 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1697 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1699 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1700 it matched the wrong number of bytes.
1703 Version 7.1 24-Apr-07
1704 ---------------------
1706 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1707 that is more "standard", making use of automake and other Autotools. There
1708 is some re-arrangement of the files and adjustment of comments consequent
1709 on this.
1711 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1712 for recursive directory scanning broke on some systems because the files
1713 are not scanned in any specific order and on different systems the order
1714 was different. A call to "sort" has been inserted into RunGrepTest for the
1715 approprate test as a short-term fix. In the longer term there may be an
1716 alternative.
1718 3. I had an email from Eric Raymond about problems translating some of PCRE's
1719 man pages to HTML (despite the fact that I distribute HTML pages, some
1720 people do their own conversions for various reasons). The problems
1721 concerned the use of low-level troff macros .br and .in. I have therefore
1722 removed all such uses from the man pages (some were redundant, some could
1723 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1724 HTML has been updated to handle .nf/.fi and to complain if it encounters
1725 .br or .in.
1727 4. Updated comments in configure.ac that get placed in config.h.in and also
1728 arranged for config.h to be included in the distribution, with the name
1729 config.h.generic, for the benefit of those who have to compile without
1730 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1732 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1733 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1734 makevp.bat for the current PCRE, using the additional files
1735 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1737 6. A Windows user reported a minor discrepancy with test 2, which turned out
1738 to be caused by a trailing space on an input line that had got lost in his
1739 copy. The trailing space was an accident, so I've just removed it.
1741 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1742 that is needed.
1744 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1745 as "const" (a) because they are and (b) because it helps the PHP
1746 maintainers who have recently made a script to detect big data structures
1747 in the php code that should be moved to the .rodata section. I remembered
1748 to update Builducptable as well, so it won't revert if ucptable.h is ever
1749 re-created.
1751 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1752 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1753 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1754 support is not required. This saves 1.5-2K of code, which is important in
1755 some applications.
1757 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1758 so as not to refer to the tables, even though these functions will never be
1759 called when UTF-8 support is disabled. Otherwise there are problems with a
1760 shared library.
1762 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1764 (a) It was defining its arguments as char * instead of void *.
1766 (b) It was assuming that all moves were upwards in memory; this was true
1767 a long time ago when I wrote it, but is no longer the case.
1769 The emulated memove() is provided for those environments that have neither
1770 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1771 is clearly not the case, as these two bugs were recently reported.
1773 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1774 and Detrail to create the HTML documentation, the .txt form of the man
1775 pages, and it removes trailing spaces from listed files. It also creates
1776 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1777 case, it wraps all the #defines with #ifndefs. This script should be run
1778 before "make dist".
1780 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1781 with Unicode property support.
1783 (a) For a maximizing quantifier, if the two different cases of the
1784 character were of different lengths in their UTF-8 codings (there are
1785 some cases like this - I found 11), and the matching function had to
1786 back up over a mixture of the two cases, it incorrectly assumed they
1787 were both the same length.
1789 (b) When PCRE was configured to use the heap rather than the stack for
1790 recursion during matching, it was not correctly preserving the data for
1791 the other case of a UTF-8 character when checking ahead for a match
1792 while processing a minimizing repeat. If the check also involved
1793 matching a wide character, but failed, corruption could cause an
1794 erroneous result when trying to check for a repeat of the original
1795 character.
1797 13. Some tidying changes to the testing mechanism:
1799 (a) The RunTest script now detects the internal link size and whether there
1800 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1801 values substituted by "configure". (The RunGrepTest script already did
1802 this for UTF-8.) The configure.ac script no longer substitutes the
1803 relevant variables.
1805 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1806 with length and offset values. This means that the output is different
1807 for different internal link sizes. Test 2 is skipped for link sizes
1808 other than 2 because of this, bypassing the problem. Unfortunately,
1809 there was also a test in test 3 (the locale tests) that used /B and
1810 failed for link sizes other than 2. Rather than cut the whole test out,
1811 I have added a new /Z option to pcretest that replaces the length and
1812 offset values with spaces. This is now used to make test 3 independent
1813 of link size. (Test 2 will be tidied up later.)
1815 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1816 segmentation fault instead of returning the appropriate error message.
1818 15. In multiline mode when the newline sequence was set to "any", the pattern
1819 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1820 This doesn't seem right; it now treats the CRLF combination as the line
1821 ending, and so does not match in that case. It's only a pattern such as ^$
1822 that would hit this one: something like ^ABC$ would have failed after \r
1823 and then tried again after \r\n.
1825 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1826 in an attempt to make files that differ only in their line terminators
1827 compare equal. This works on Linux.
1829 17. Under certain error circumstances pcregrep might try to free random memory
1830 as it exited. This is now fixed, thanks to valgrind.
1832 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1833 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1834 was because its rules for how to advance for /g after matching an empty
1835 string at the end of a line did not allow for this case. They now check for
1836 it specially.
1838 20. pcretest is supposed to handle patterns and data of any length, by
1839 extending its buffers when necessary. It was getting this wrong when the
1840 buffer for a data line had to be extended.
1842 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1843 CRLF as a newline sequence.
1845 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1846 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1847 I have nevertheless tidied it up.
1849 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1851 24. Added a man page for pcre-config.
1854 Version 7.0 19-Dec-06
1855 ---------------------
1857 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1858 moving to gcc 4.1.1.
1860 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1861 sys/time.h, which is documented as needed for this function. It doesn't
1862 seem to matter on Linux, but it showed up on some releases of OS X.
1864 3. It seems that there are systems where bytes whose values are greater than
1865 127 match isprint() in the "C" locale. The "C" locale should be the
1866 default when a C program starts up. In most systems, only ASCII printing
1867 characters match isprint(). This difference caused the output from pcretest
1868 to vary, making some of the tests fail. I have changed pcretest so that:
1870 (a) When it is outputting text in the compiled version of a pattern, bytes
1871 other than 32-126 are always shown as hex escapes.
1873 (b) When it is outputting text that is a matched part of a subject string,
1874 it does the same, unless a different locale has been set for the match
1875 (using the /L modifier). In this case, it uses isprint() to decide.
1877 4. Fixed a major bug that caused incorrect computation of the amount of memory
1878 required for a compiled pattern when options that changed within the
1879 pattern affected the logic of the preliminary scan that determines the
1880 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1881 that the computed length was too small. The symptoms of this bug were
1882 either the PCRE error "internal error: code overflow" from pcre_compile(),
1883 or a glibc crash with a message such as "pcretest: free(): invalid next
1884 size (fast)". Examples of patterns that provoked this bug (shown in
1885 pcretest format) are:
1887 /(?-x: )/x
1888 /(?x)(?-x: \s*#\s*)/
1889 /((?i)[\x{c0}])/8
1890 /(?i:[\x{c0}])/8
1892 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1893 is now done differently.
1895 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1896 wrapper classes; (b) implement a new function in the C++ scanner that is
1897 more efficient than the old way of doing things because it avoids levels of
1898 recursion in the regex matching; (c) add a paragraph to the documentation
1899 for the FullMatch() function.
1901 6. The escape sequence \n was being treated as whatever was defined as
1902 "newline". Not only was this contrary to the documentation, which states
1903 that \n is character 10 (hex 0A), but it also went horribly wrong when
1904 "newline" was defined as CRLF. This has been fixed.
1906 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1907 was being set to -1 for the "end of line" case (supposedly a value that no
1908 character can have). Though this value is never used (the check for end of
1909 line is "zero bytes in current character"), it caused compiler complaints.
1910 I've changed it to 0xffffffff.
1912 8. In pcre_version.c, the version string was being built by a sequence of
1913 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1914 string (as it is for production releases) called a macro with an empty
1915 argument. The C standard says the result of this is undefined. The gcc
1916 compiler treats it as an empty string (which was what was wanted) but it is
1917 reported that Visual C gives an error. The source has been hacked around to
1918 avoid this problem.
1920 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1921 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1922 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1923 of them did).
1925 10. Originally, pcretest opened its input and output without "b"; then I was
1926 told that "b" was needed in some environments, so it was added for release
1927 5.0 to both the input and output. (It makes no difference on Unix-like
1928 systems.) Later I was told that it is wrong for the input on Windows. I've
1929 now abstracted the modes into two macros, to make it easier to fiddle with
1930 them, and removed "b" from the input mode under Windows.
1932 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1934 12. Added -help and --help to pcretest as an official way of being reminded
1935 of the options.
1937 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1938 and pcrecpp.cc because they annoy compilers at high warning levels.
1940 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1942 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1943 scripts are not C programs :-) and which was not noticed because it works
1944 on Linux.
1946 16. pcretest is supposed to handle any length of pattern and data line (as one
1947 line or as a continued sequence of lines) by extending its input buffer if
1948 necessary. This feature was broken for very long pattern lines, leading to
1949 a string of junk being passed to pcre_compile() if the pattern was longer
1950 than about 50K.
1952 17. I have done a major re-factoring of the way pcre_compile() computes the
1953 amount of memory needed for a compiled pattern. Previously, there was code
1954 that made a preliminary scan of the pattern in order to do this. That was
1955 OK when PCRE was new, but as the facilities have expanded, it has become
1956 harder and harder to keep it in step with the real compile phase, and there
1957 have been a number of bugs (see for example, 4 above). I have now found a
1958 cunning way of running the real compile function in a "fake" mode that
1959 enables it to compute how much memory it would need, while actually only
1960 ever using a few hundred bytes of working memory and without too many
1961 tests of the mode. This should make future maintenance and development
1962 easier. A side effect of this work is that the limit of 200 on the nesting
1963 depth of parentheses has been removed (though this was never a serious
1964 limitation, I suspect). However, there is a downside: pcre_compile() now
1965 runs more slowly than before (30% or more, depending on the pattern). I
1966 hope this isn't a big issue. There is no effect on runtime performance.
1968 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1969 newline (only possible for the last line of a file) and it was a
1970 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1972 19. Added additional timing features to pcretest. (1) The -tm option now times
1973 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1974 separate command line item, by a number that specifies the number of
1975 repeats to use when timing. The default is 50000; this gives better
1976 precision, but takes uncomfortably long for very large patterns.
1978 20. Extended pcre_study() to be more clever in cases where a branch of a
1979 subpattern has no definite first character. For example, (a*|b*)[cd] would
1980 previously give no result from pcre_study(). Now it recognizes that the
1981 first character must be a, b, c, or d.
1983 21. There was an incorrect error "recursive call could loop indefinitely" if
1984 a subpattern (or the entire pattern) that was being tested for matching an
1985 empty string contained only one non-empty item after a nested subpattern.
1986 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1987 incorrectly, because the \d was being skipped in the check.
1989 22. The pcretest program now has a new pattern option /B and a command line
1990 option -b, which is equivalent to adding /B to every pattern. This causes
1991 it to show the compiled bytecode, without the additional information that
1992 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1993 is the same as /B/I).
1995 23. A new optimization is now able automatically to treat some sequences such
1996 as a*b as a*+b. More specifically, if something simple (such as a character
1997 or a simple class like \d) has an unlimited quantifier, and is followed by
1998 something that cannot possibly match the quantified thing, the quantifier
1999 is automatically "possessified".
2001 24. A recursive reference to a subpattern whose number was greater than 39
2002 went wrong under certain circumstances in UTF-8 mode. This bug could also
2003 have affected the operation of pcre_study().
2005 25. Realized that a little bit of performance could be had by replacing
2006 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2008 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2010 27. Possessive quantifiers such as a++ were previously implemented by turning
2011 them into atomic groups such as ($>a+). Now they have their own opcodes,
2012 which improves performance. This includes the automatically created ones
2013 from 23 above.
2015 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2016 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2017 the first matched character to be a colon. This applied both to named and
2018 numbered groups.
2020 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2022 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2023 building PCRE on Windows easier, so I have included it in the distribution.
2025 31. There is now a check in pcretest against a ridiculously large number being
2026 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2027 loop, the loop is abandoned.
2029 32. Forward references to subpatterns in conditions such as (?(2)...) where
2030 subpattern 2 is defined later cause pcre_compile() to search forwards in
2031 the pattern for the relevant set of parentheses. This search went wrong
2032 when there were unescaped parentheses in a character class, parentheses
2033 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2035 33. "Subroutine" calls and backreferences were previously restricted to
2036 referencing subpatterns earlier in the regex. This restriction has now
2037 been removed.
2039 34. Added a number of extra features that are going to be in Perl 5.10. On the
2040 whole, these are just syntactic alternatives for features that PCRE had
2041 previously implemented using the Python syntax or my own invention. The
2042 other formats are all retained for compatibility.
2044 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2045 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2046 also .NET compatible.
2048 (b) A recursion or subroutine call to a named group can now be defined as
2049 (?&name) as well as (?P>name).
2051 (c) A backreference to a named group can now be defined as \k<name> or
2052 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2053 5.10, are also .NET compatible.
2055 (d) A conditional reference to a named group can now use the syntax
2056 (?(<name>) or (?('name') as well as (?(name).
2058 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2059 groups (named and numbered) that are never evaluated inline, but can be
2060 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2061 is always false. There may be only one alternative in such a group.
2063 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2064 as the simple (?(R). The condition is true only if the most recent
2065 recursion is that of the given number or name. It does not search out
2066 through the entire recursion stack.
2068 (g) The escape \gN or \g{N} has been added, where N is a positive or
2069 negative number, specifying an absolute or relative reference.
2071 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2072 some "unreachable code" warnings.
2074 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2075 things, this adds five new scripts.
2077 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2078 There were also incompatibilities regarding the handling of \Q..\E inside
2079 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2080 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2082 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2083 matches an empty string, and forcibly breaks the loop. There were bugs in
2084 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2085 against aaaa the result was just "a" rather than "aaaa", for example. Two
2086 separate and independent bugs (that affected different cases) have been
2087 fixed.
2089 39. Refactored the code to abolish the use of different opcodes for small
2090 capturing bracket numbers. This is a tidy that I avoided doing when I
2091 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2092 The new approach is not only tidier, it makes it possible to reduce the
2093 memory needed to fix the previous bug (38).
2095 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2096 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2097 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2098 mode.
2100 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2101 report.
2103 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2104 copy construction and assignment in the C++ wrapper.
2106 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2107 couple of bugs that could have given wrong results in the "--newline=crlf"
2108 case.
2110 44. Added a number of casts and did some reorganization of signed/unsigned int
2111 variables following suggestions from Dair Grant. Also renamed the variable
2112 "this" as "item" because it is a C++ keyword.
2114 45. Arranged for dftables to add
2116 #include "pcre_internal.h"
2118 to pcre_chartables.c because without it, gcc 4.x may remove the array
2119 definition from the final binary if PCRE is built into a static library and
2120 dead code stripping is activated.
2122 46. For an unanchored pattern, if a match attempt fails at the start of a
2123 newline sequence, and the newline setting is CRLF or ANY, and the next two
2124 characters are CRLF, advance by two characters instead of one.
2127 Version 6.7 04-Jul-06
2128 ---------------------
2130 1. In order to handle tests when input lines are enormously long, pcretest has
2131 been re-factored so that it automatically extends its buffers when
2132 necessary. The code is crude, but this _is_ just a test program. The
2133 default size has been increased from 32K to 50K.
2135 2. The code in pcre_study() was using the value of the re argument before
2136 testing it for NULL. (Of course, in any sensible call of the function, it
2137 won't be NULL.)
2139 3. The memmove() emulation function in pcre_internal.h, which is used on
2140 systems that lack both memmove() and bcopy() - that is, hardly ever -
2141 was missing a "static" storage class specifier.
2143 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2144 containing an extended class (one that cannot be represented by a bitmap
2145 because it contains high-valued characters or Unicode property items, e.g.
2146 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2147 pattern, but PCRE should not loop if you do not (it no longer does).
2148 [Detail: two cases were found: (a) a repeated subpattern containing an
2149 extended class; (b) a recursive reference to a subpattern that followed a
2150 previous extended class. It wasn't skipping over the extended class
2151 correctly when UTF-8 mode was not set.]
2153 5. A negated single-character class was not being recognized as fixed-length
2154 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2155 compile error "lookbehind assertion is not fixed length".
2157 6. The RunPerlTest auxiliary script was showing an unexpected difference
2158 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2159 write a Perl script that can interpret lines of an input file either as
2160 byte characters or as UTF-8, which is what "perltest" was being required to
2161 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2162 can't do is switch easily at run time between having the "use utf8;" pragma
2163 or not. In the end, I fudged it by using the RunPerlTest script to insert
2164 "use utf8;" explicitly for the UTF-8 tests.
2166 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2167 the end of the subject string, contrary to the documentation and to what
2168 Perl does. This was true of both matching functions. Now it matches only at
2169 the start of the subject and immediately after *internal* newlines.
2171 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2172 a pointer to an int instead of a pointer to an unsigned long int. This
2173 caused problems on 64-bit systems.
2175 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2176 instance of the 'standard' template library not being so standard".
2178 10. There was no check on the number of named subpatterns nor the maximum
2179 length of a subpattern name. The product of these values is used to compute
2180 the size of the memory block for a compiled pattern. By supplying a very
2181 long subpattern name and a large number of named subpatterns, the size
2182 computation could be caused to overflow. This is now prevented by limiting
2183 the length of names to 32 characters, and the number of named subpatterns
2184 to 10,000.
2186 11. Subpatterns that are repeated with specific counts have to be replicated in
2187 the compiled pattern. The size of memory for this was computed from the
2188 length of the subpattern and the repeat count. The latter is limited to
2189 65535, but there was no limit on the former, meaning that integer overflow
2190 could in principle occur. The compiled length of a repeated subpattern is
2191 now limited to 30,000 bytes in order to prevent this.
2193 12. Added the optional facility to have named substrings with the same name.
2195 13. Added the ability to use a named substring as a condition, using the
2196 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2197 are numbers (not recommended). Forward references are permitted.
2199 14. Added forward references in named backreferences (if you see what I mean).
2201 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2202 pattern could run off the end of the subject. For example, the pattern
2203 "(?s)(.{1,5})"8 did this with the subject "ab".
2205 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2206 PCRE_CASELESS was set when matching characters that were quantified with ?
2207 or *.
2209 17. A character class other than a single negated character that had a minimum
2210 but no maximum quantifier - for example [ab]{6,} - was not handled
2211 correctly by pce_dfa_exec(). It would match only one character.
2213 18. A valid (though odd) pattern that looked like a POSIX character
2214 class but used an invalid character after [ (for example [[,abc,]]) caused
2215 pcre_compile() to give the error "Failed: internal error: code overflow" or
2216 in some cases to crash with a glibc free() error. This could even happen if
2217 the pattern terminated after [[ but there just happened to be a sequence of
2218 letters, a binary zero, and a closing ] in the memory that followed.
2220 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2221 over the years. Originally (before any Unicode support), just the bottom 8
2222 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2223 output from "man perlunicode" includes this:
2225 The regular expression compiler produces polymorphic opcodes. That
2226 is, the pattern adapts to the data and automatically switches to
2227 the Unicode character scheme when presented with Unicode data--or
2228 instead uses a traditional byte scheme when presented with byte
2229 data.
2231 Sadly, a wide octal escape does not cause a switch, and in a string with
2232 no other multibyte characters, these octal escapes are treated as before.
2233 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2234 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2235 Unicode string.
2237 I have not perpetrated such confusion in PCRE. Up till now, it took just
2238 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2239 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2240 translate to the appropriate multibyte character.
2242 29. Applied some refactoring to reduce the number of warnings from Microsoft
2243 and Borland compilers. This has included removing the fudge introduced
2244 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2245 a warning about an unused variable.
2247 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2248 characters since release 4.0, because Perl (from release 5.004) does not.
2249 [Or at least, is documented not to: some releases seem to be in conflict
2250 with the documentation.] However, when a pattern was studied with
2251 pcre_study() and all its branches started with \s, PCRE still included VT
2252 as a possible starting character. Of course, this did no harm; it just
2253 caused an unnecessary match attempt.
2255 22. Removed a now-redundant internal flag bit that recorded the fact that case
2256 dependency changed within the pattern. This was once needed for "required
2257 byte" processing, but is no longer used. This recovers a now-scarce options
2258 bit. Also moved the least significant internal flag bit to the most-
2259 significant bit of the word, which was not previously used (hangover from
2260 the days when it was an int rather than a uint) to free up another bit for
2261 the future.
2263 23. Added support for CRLF line endings as well as CR and LF. As well as the
2264 default being selectable at build time, it can now be changed at runtime
2265 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2266 specify that it is scanning data with non-default line endings.
2268 24. Changed the definition of CXXLINK to make it agree with the definition of
2269 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2271 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2272 recursions. This makes a big different to stack usage for some patterns.
2274 26. If a subpattern containing a named recursion or subroutine reference such
2275 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2276 the space required for the compiled pattern went wrong and gave too small a
2277 value. Depending on the environment, this could lead to "Failed: internal
2278 error: code overflow at offset 49" or "glibc detected double free or
2279 corruption" errors.
2281 27. Applied patches from Google (a) to support the new newline modes and (b) to
2282 advance over multibyte UTF-8 characters in GlobalReplace.
2284 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2285 difference for some implementation of PCRE in some Windows version.
2287 29. Added some extra testing facilities to pcretest:
2289 \q<number> in a data line sets the "match limit" value
2290 \Q<number> in a data line sets the "match recursion limt" value
2291 -S <number> sets the stack size, where <number> is in megabytes
2293 The -S option isn't available for Windows.
2296 Version 6.6 06-Feb-06
2297 ---------------------
2299 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2300 in pcreposix.h. I have copied the definition from pcre.h.
2302 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2303 because pcre.h is no longer a built file.
2305 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2306 not normally included in the compiled code.
2309 Version 6.5 01-Feb-06
2310 ---------------------
2312 1. When using the partial match feature with pcre_dfa_exec(), it was not
2313 anchoring the second and subsequent partial matches at the new starting
2314 point. This could lead to incorrect results. For example, with the pattern
2315 /1234/, partially matching against "123" and then "a4" gave a match.
2317 2. Changes to pcregrep:
2319 (a) All non-match returns from pcre_exec() were being treated as failures
2320 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2321 error message is output. Some extra information is given for the
2323 probably the only errors that are likely to be caused by users (by
2324 specifying a regex that has nested indefinite repeats, for instance).
2325 If there are more than 20 of these errors, pcregrep is abandoned.
2327 (b) A binary zero was treated as data while matching, but terminated the
2328 output line if it was written out. This has been fixed: binary zeroes
2329 are now no different to any other data bytes.
2331 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2332 used to set a locale for matching. The --locale=xxxx long option has
2333 been added (no short equivalent) to specify a locale explicitly on the
2334 pcregrep command, overriding the environment variables.
2336 (d) When -B was used with -n, some line numbers in the output were one less
2337 than they should have been.
2339 (e) Added the -o (--only-matching) option.
2341 (f) If -A or -C was used with -c (count only), some lines of context were
2342 accidentally printed for the final match.
2344 (g) Added the -H (--with-filename) option.
2346 (h) The combination of options -rh failed to suppress file names for files
2347 that were found from directory arguments.
2349 (i) Added the -D (--devices) and -d (--directories) options.
2351 (j) Added the -F (--fixed-strings) option.
2353 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2355 (l) Added the --colo(u)r option.
2357 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2358 is not present by default.
2360 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2361 items such as (?R) or (?1), when the recursion could match a number of
2362 alternatives. If it matched one of the alternatives, but subsequently,
2363 outside the recursion, there was a failure, the code tried to back up into
2364 the recursion. However, because of the way PCRE is implemented, this is not
2365 possible, and the result was an incorrect result from the match.
2367 In order to prevent this happening, the specification of recursion has
2368 been changed so that all such subpatterns are automatically treated as
2369 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2371 4. I had overlooked the fact that, in some locales, there are characters for
2372 which isalpha() is true but neither isupper() nor islower() are true. In
2373 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2374 and ordfeminine) are like this. This affected the treatment of \w and \W
2375 when they appeared in character classes, but not when they appeared outside
2376 a character class. The bit map for "word" characters is now created
2377 separately from the results of isalnum() instead of just taking it from the
2378 upper, lower, and digit maps. (Plus the underscore character, of course.)
2380 5. The above bug also affected the handling of POSIX character classes such as
2381 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2382 permanent tables. Instead, the bit maps for such a class were previously
2383 created as the appropriate unions of the upper, lower, and digit bitmaps.
2384 Now they are created by subtraction from the [[:word:]] class, which has
2385 its own bitmap.
2387 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2388 It is created by subtracting the vertical space characters (\x09, \x0a,
2389 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2390 subtraction was done in the overall bitmap for a character class, meaning
2391 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2392 be recognized. This bug has been fixed.
2394 7. Patches from the folks at Google:
2396 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2397 real life, but is still worth protecting against".
2399 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2400 regular expressions".
2402 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2403 have it.
2405 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2406 "configure" and the latter not, in order to fix a problem somebody had
2407 with compiling the Arg class on HP-UX.
2409 (e) Improve the error-handling of the C++ wrapper a little bit.
2411 (f) New tests for checking recursion limiting.
2413 8. The pcre_memmove() function, which is used only if the environment does not
2414 have a standard memmove() function (and is therefore rarely compiled),
2415 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2416 returning a result (though PCRE never actually uses the result).
2418 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2419 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2420 returned instead of calling malloc() with an overflowing number that would
2421 most likely cause subsequent chaos.
2423 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2425 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2426 with this option is matched, the nmatch and pmatch options of regexec() are
2427 ignored.
2429 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2430 provided in case anyone wants to the the POSIX interface with UTF-8
2431 strings.
2433 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2434 C++ linking (needed for some HP-UX environments).
2436 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2437 (unused parameter) and in the pcre_printint() function (omitted "default"
2438 switch label when the default is to do nothing).
2440 15. Added some code to make it possible, when PCRE is compiled as a C++
2441 library, to replace subject pointers for pcre_exec() with a smart pointer
2442 class, thus making it possible to process discontinuous strings.
2444 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2445 much the same function. They were added by different people who were trying
2446 to make PCRE easy to compile on non-Unix systems. It has been suggested
2447 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2448 for compiling on Windows systems. I have therefore replaced it with
2449 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2450 defaults to "extern" for C or "extern C" for C++, which works fine on
2451 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2452 SCOPE with something explicit in config.h. In addition:
2454 (a) pcreposix.h still had just "extern" instead of either of these macros;
2455 I have replaced it with PCRE_DATA_SCOPE.
2457 (b) Functions such as _pcre_xclass(), which are internal to the library,
2458 but external in the C sense, all had PCRE_EXPORT in their definitions.
2459 This is apparently wrong for the Windows case, so I have removed it.
2460 (It makes no difference on Unix-like systems.)
2462 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2463 of recursive calls to match(). This is different to MATCH_LIMIT because
2464 that limits the total number of calls to match(), not all of which increase
2465 the depth of recursion. Limiting the recursion depth limits the amount of
2466 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2467 when PCRE is compiled, and changed at run time. A patch from Google adds
2468 this functionality to the C++ interface.
2470 18. Changes to the handling of Unicode character properties:
2472 (a) Updated the table to Unicode 4.1.0.
2474 (b) Recognize characters that are not in the table as "Cn" (undefined).
2476 (c) I revised the way the table is implemented to a much improved format
2477 which includes recognition of ranges. It now supports the ranges that
2478 are defined in UnicodeData.txt, and it also amalgamates other
2479 characters into ranges. This has reduced the number of entries in the
2480 table from around 16,000 to around 3,000, thus reducing its size
2481 considerably. I realized I did not need to use a tree structure after
2482 all - a binary chop search is just as efficient. Having reduced the
2483 number of entries, I extended their size from 6 bytes to 8 bytes to
2484 allow for more data.
2486 (d) Added support for Unicode script names via properties such as \p{Han}.
2488 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2489 matching that character.
2491 20. When matching a repeated Unicode property with a minimum greater than zero,
2492 (for example \pL{2,}), PCRE could look past the end of the subject if it
2493 reached it while seeking the minimum number of characters. This could
2494 happen only if some of the characters were more than one byte long, because
2495 there is a check for at least the minimum number of bytes.
2497 21. Refactored the implementation of \p and \P so as to be more general, to
2498 allow for more different types of property in future. This has changed the
2499 compiled form incompatibly. Anybody with saved compiled patterns that use
2500 \p or \P will have to recompile them.
2502 22. Added "Any" and "L&" to the supported property types.
2504 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2505 but give a compile time error if the value is greater than 0xff.
2507 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2508 accidentally not being installed or uninstalled.
2510 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2511 made were to insert the current release number. This seemed silly, because
2512 it made things harder for people building PCRE on systems that don't run
2513 "configure". I have turned pcre.h into a distributed file, no longer built
2514 by "configure", with the version identification directly included. There is
2515 no longer a pcre.h.in file.
2517 However, this change necessitated a change to the pcre-config script as
2518 well. It is built from pcre-config.in, and one of the substitutions was the
2519 release number. I have updated configure.ac so that ./configure now finds
2520 the release number by grepping pcre.h.
2522 26. Added the ability to run the tests under valgrind.
2525 Version 6.4 05-Sep-05
2526 ---------------------
2528 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2529 "--" to be printed when multiple files were scanned, even when none of the
2530 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2531 consider it to be a bug, and have restored the previous behaviour.
2533 2. A couple of code tidies to get rid of compiler warnings.
2535 3. The pcretest program used to cheat by referring to symbols in the library
2536 whose names begin with _pcre_. These are internal symbols that are not
2537 really supposed to be visible externally, and in some environments it is
2538 possible to suppress them. The cheating is now confined to including
2539 certain files from the library's source, which is a bit cleaner.
2541 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2542 file's purpose clearer.
2544 5. Reorganized pcre_ucp_findchar().
2547 Version 6.3 15-Aug-05
2548 ---------------------
2550 1. The file libpcre.pc.in did not have general read permission in the tarball.
2552 2. There were some problems when building without C++ support:
2554 (a) If C++ support was not built, "make install" and "make test" still
2555 tried to test it.
2557 (b) There were problems when the value of CXX was explicitly set. Some
2558 changes have been made to try to fix these, and ...
2560 (c) --disable-cpp can now be used to explicitly disable C++ support.
2562 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2563 backslash in a target when C++ was disabled. This confuses some
2564 versions of "make", apparently. Using an intermediate variable solves
2565 this. (Same for CPP_LOBJ.)
2567 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2568 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2569 necessary on certain architectures.
2571 4. Added a setting of -export-symbols-regex to the link command to remove
2572 those symbols that are exported in the C sense, but actually are local
2573 within the library, and not documented. Their names all begin with
2574 "_pcre_". This is not a perfect job, because (a) we have to except some
2575 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2576 available (and never for static libraries). I have made a note to try to
2577 find a way round (a) in the future.
2580 Version 6.2 01-Aug-05
2581 ---------------------
2583 1. There was no test for integer overflow of quantifier values. A construction
2584 such as {1111111111111111} would give undefined results. What is worse, if
2585 a minimum quantifier for a parenthesized subpattern overflowed and became
2586 negative, the calculation of the memory size went wrong. This could have
2587 led to memory overwriting.
2589 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2591 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2592 operating environments where this matters.
2594 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2595 PCRE options from within the C++ wrapper.
2597 5. Named capturing subpatterns were not being correctly counted when a pattern
2598 was compiled. This caused two problems: (a) If there were more than 100
2599 such subpatterns, the calculation of the memory needed for the whole
2600 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2601 back references of the form \12, where the number was greater than 9, were
2602 not recognized as back references, even though there were sufficient
2603 previous subpatterns.
2605 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2606 versions of gcc, e.g. 2.95.4.
2609 Version 6.1 21-Jun-05
2610 ---------------------
2612 1. There was one reference to the variable "posix" in pcretest.c that was not
2613 surrounded by "#if !defined NOPOSIX".
2615 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2616 the cross-check on the old pcre_info() function, for the benefit of the
2617 cut-down version of PCRE that is currently imported into Exim.
2619 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2620 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2621 patterns that start (?i) but I don't think that matters. The use of (?i) is
2622 just an example; this all applies to the other options as well.
2624 4. Since libtool seems to echo the compile commands it is issuing, the output
2625 from "make" can be reduced a bit by putting "@" in front of each libtool
2626 compile command.
2628 5. Patch from the folks at Google for configure.in to be a bit more thorough
2629 in checking for a suitable C++ installation before trying to compile the
2630 C++ stuff. This should fix a reported problem when a compiler was present,
2631 but no suitable headers.
2633 6. The man pages all had just "PCRE" as their title. I have changed them to
2634 be the relevant file name. I have also arranged that these names are
2635 retained in the file doc/pcre.txt, which is a concatenation in text format
2636 of all the man pages except the little individual ones for each function.
2638 7. The NON-UNIX-USE file had not been updated for the different set of source
2639 files that come with release 6. I also added a few comments about the C++
2640 wrapper.
2643 Version 6.0 07-Jun-05
2644 ---------------------
2646 1. Some minor internal re-organization to help with my DFA experiments.
2648 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2649 didn't matter for the library itself when fully configured, but did matter
2650 when compiling without UCP support, or within Exim, where the ucp files are
2651 not imported.
2653 3. Refactoring of the library code to split up the various functions into
2654 different source modules. The addition of the new DFA matching code (see
2655 below) to a single monolithic source would have made it really too
2656 unwieldy, quite apart from causing all the code to be include in a
2657 statically linked application, when only some functions are used. This is
2658 relevant even without the DFA addition now that patterns can be compiled in
2659 one application and matched in another.
2661 The downside of splitting up is that there have to be some external
2662 functions and data tables that are used internally in different modules of
2663 the library but which are not part of the API. These have all had their
2664 names changed to start with "_pcre_" so that they are unlikely to clash
2665 with other external names.
2667 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2668 a different (DFA) algorithm. Although it is slower than the original
2669 function, it does have some advantages for certain types of matching
2670 problem.
2672 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2673 including restarting after a partial match.
2675 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2676 defined when compiling for Windows was sent to me. I have put it into the
2677 code, though I have no means of testing or verifying it.
2679 7. Added the pcre_refcount() auxiliary function.
2681 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2682 match before or at the first newline in the subject string. In pcretest,
2683 the /f option on a pattern can be used to set this.
2685 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2686 would behave wrongly. This has been present in PCRE since release 4.0.
2688 10. A number of changes to the pcregrep command:
2690 (a) Refactored how -x works; insert ^(...)$ instead of setting
2691 PCRE_ANCHORED and checking the length, in preparation for adding
2692 something similar for -w.
2694 (b) Added the -w (match as a word) option.
2696 (c) Refactored the way lines are read and buffered so as to have more
2697 than one at a time available.
2699 (d) Implemented a pcregrep test script.
2701 (e) Added the -M (multiline match) option. This allows patterns to match
2702 over several lines of the subject. The buffering ensures that at least
2703 8K, or the rest of the document (whichever is the shorter) is available
2704 for matching (and similarly the previous 8K for lookbehind assertions).
2706 (f) Changed the --help output so that it now says
2708 -w, --word-regex(p)
2710 instead of two lines, one with "regex" and the other with "regexp"
2711 because that confused at least one person since the short forms are the
2712 same. (This required a bit of code, as the output is generated
2713 automatically from a table. It wasn't just a text change.)
2715 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2716 option but starts with a hyphen. Could be a pattern or a path name
2717 starting with a hyphen, for instance.
2719 (h) "-" can be given as a file name to represent stdin.
2721 (i) When file names are being printed, "(standard input)" is used for
2722 the standard input, for compatibility with GNU grep. Previously
2723 "<stdin>" was used.
2725 (j) The option --label=xxx can be used to supply a name to be used for
2726 stdin when file names are being printed. There is no short form.
2728 (k) Re-factored the options decoding logic because we are going to add
2729 two more options that take data. Such options can now be given in four
2730 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2732 (l) Added the -A, -B, and -C options for requesting that lines of context
2733 around matches be printed.
2735 (m) Added the -L option to print the names of files that do not contain
2736 any matching lines, that is, the complement of -l.
2738 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2739 continue to scan other files.
2741 (o) The -s option was incorrectly implemented. For compatibility with other
2742 greps, it now suppresses the error message for a non-existent or non-
2743 accessible file (but not the return code). There is a new option called
2744 -q that suppresses the output of matching lines, which was what -s was
2745 previously doing.
2747 (p) Added --include and --exclude options to specify files for inclusion
2748 and exclusion when recursing.
2750 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2751 Hopefully, it now does.
2753 12. Missing cast in pcre_study().
2755 13. Added an "uninstall" target to the makefile.
2757 14. Replaced "extern" in the function prototypes in Makefile.in with
2758 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2759 world, but is set differently for Windows.
2761 15. Added a second compiling function called pcre_compile2(). The only
2762 difference is that it has an extra argument, which is a pointer to an
2763 integer error code. When there is a compile-time failure, this is set
2764 non-zero, in addition to the error test pointer being set to point to an
2765 error message. The new argument may be NULL if no error number is required
2766 (but then you may as well call pcre_compile(), which is now just a
2767 wrapper). This facility is provided because some applications need a
2768 numeric error indication, but it has also enabled me to tidy up the way
2769 compile-time errors are handled in the POSIX wrapper.
2771 16. Added VPATH=.libs to the makefile; this should help when building with one
2772 prefix path and installing with another. (Or so I'm told by someone who
2773 knows more about this stuff than I do.)
2775 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2776 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2777 match everything, including newlines. This is not POSIX-compatible, but
2778 somebody wanted the feature. From pcretest it can be activated by using
2779 both the P and the s flags.
2781 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2783 19. libpcre.pc was being incorrectly installed as executable.
2785 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2786 it now also looks for '\r' so that it will work unmodified on Windows.
2788 21. Added Google's contributed C++ wrapper to the distribution.
2790 22. Added some untidy missing memory free() calls in pcretest, to keep
2791 Electric Fence happy when testing.
2795 Version 5.0 13-Sep-04
2796 ---------------------
2798 1. Internal change: literal characters are no longer packed up into items
2799 containing multiple characters in a single byte-string. Each character
2800 is now matched using a separate opcode. However, there may be more than one
2801 byte in the character in UTF-8 mode.
2803 2. The pcre_callout_block structure has two new fields: pattern_position and
2804 next_item_length. These contain the offset in the pattern to the next match
2805 item, and its length, respectively.
2807 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2808 insertion of callouts before each pattern item. Added the /C option to
2809 pcretest to make use of this.
2811 4. On the advice of a Windows user, the lines
2813 #if defined(_WIN32) || defined(WIN32)
2814 _setmode( _fileno( stdout ), 0x8000 );
2815 #endif /* defined(_WIN32) || defined(WIN32) */
2817 have been added to the source of pcretest. This apparently does useful
2818 magic in relation to line terminators.
2820 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2821 for the benefit of those environments where the "b" makes a difference.
2823 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2824 to know about it. I have put a hack into configure.in that adds in code
2825 to set GCC=yes if CC=icc. This seems to end up at a point in the
2826 generated configure script that is early enough to affect the setting of
2827 compiler options, which is what is needed, but I have no means of testing
2828 whether it really works. (The user who reported this had patched the
2829 generated configure script, which of course I cannot do.)
2831 LATER: After change 22 below (new libtool files), the configure script
2832 seems to know about icc (and also ecc). Therefore, I have commented out
2833 this hack in configure.in.
2835 7. Added support for pkg-config (2 patches were sent in).
2837 8. Negated POSIX character classes that used a combination of internal tables
2838 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2839 [[:^ascii]]. Typically, they would match almost any characters. The other
2840 POSIX classes were not broken in this way.
2842 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2843 to find the match, as PCRE was deluded into thinking that the match had to
2844 start at the start point or following a newline. The same bug applied to
2845 patterns with negative forward assertions or any backward assertions
2846 preceding ".*" at the start, unless the pattern required a fixed first
2847 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2849 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2850 starting at the last subject character, bytes beyond the end of the subject
2851 string were read.
2853 11. Renamed the variable "class" as "classbits" to make life easier for C++
2854 users. (Previously there was a macro definition, but it apparently wasn't
2855 enough.)
2857 12. Added the new field "tables" to the extra data so that tables can be passed
2858 in at exec time, or the internal tables can be re-selected. This allows
2859 a compiled regex to be saved and re-used at a later time by a different
2860 program that might have everything at different addresses.
2862 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2863 -R library as well as a -L library.
2865 14. The debugging options of pcretest (-d on the command line or D on a
2866 pattern) showed incorrect output for anything following an extended class
2867 that contained multibyte characters and which was followed by a quantifier.
2869 15. Added optional support for general category Unicode character properties
2870 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2871 support. It adds about 90K to the size of the library. The meanings of the
2872 inbuilt class escapes such as \d and \s have NOT been changed.
2874 16. Updated pcredemo.c to include calls to free() to release the memory for the
2875 compiled pattern.
2877 17. The generated file chartables.c was being created in the source directory
2878 instead of in the building directory. This caused the build to fail if the
2879 source directory was different from the building directory, and was
2880 read-only.
2882 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2883 file. No doubt somebody will tell me if they don't make sense... Also added
2884 Dan Mooney's comments about building on OpenVMS.
2886 19. Added support for partial matching via the PCRE_PARTIAL option for
2887 pcre_exec() and the \P data escape in pcretest.
2889 20. Extended pcretest with 3 new pattern features:
2891 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2892 write the compiled pattern to the file whose name is "rest-of-line".
2893 This is a straight binary dump of the data, with the saved pointer to
2894 the character tables forced to be NULL. The study data, if any, is
2895 written too. After writing, pcretest reads a new pattern.
2897 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2898 compiled pattern from the given file. There must not be any
2899 occurrences of "<" in the file name (pretty unlikely); if there are,
2900 pcretest will instead treat the initial "<" as a pattern delimiter.
2901 After reading in the pattern, pcretest goes on to read data lines as
2902 usual.
2904 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2905 and 16-bit fields in a compiled pattern, to simulate a pattern that
2906 was compiled on a host of opposite endianness.
2908 21. The pcre-exec() function can now cope with patterns that were compiled on
2909 hosts of opposite endianness, with this restriction:
2911 As for any compiled expression that is saved and used later, the tables
2912 pointer field cannot be preserved; the extra_data field in the arguments
2913 to pcre_exec() should be used to pass in a tables address if a value
2914 other than the default internal tables were used at compile time.
2916 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2917 now diagnosed as an error. Previously, most of the time, a negative number
2918 would have been treated as zero, but if in addition "ovector" was passed as
2919 NULL, a crash could occur.
2921 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2922 new versions from the libtool 1.5 distribution (the last one is a copy of
2923 a file called libtool.m4). This seems to have fixed the need to patch
2924 "configure" to support Darwin 1.3 (which I used to do). However, I still
2925 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2926 workstation).
2928 24. Changed the PCRE licence to be the more standard "BSD" licence.
2931 Version 4.5 01-Dec-03
2932 ---------------------
2934 1. There has been some re-arrangement of the code for the match() function so
2935 that it can be compiled in a version that does not call itself recursively.
2936 Instead, it keeps those local variables that need separate instances for
2937 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2938 needs to "recurse". Keeping track of where control must go is done by means
2939 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2940 hide most of the details from the main code, and operates only if
2941 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2942 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2943 operating.
2945 To make it easier for callers to provide specially tailored get/free
2946 functions for this usage, two new functions, pcre_stack_malloc, and
2947 pcre_stack_free, are used. They are always called in strict stacking order,
2948 and the size of block requested is always the same.
2950 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2951 PCRE has been compiled to use the stack or the heap for recursion. The
2952 -C option of pcretest uses this to show which version is compiled.
2954 A new data escape \S, is added to pcretest; it causes the amounts of store
2955 obtained and freed by both kinds of malloc/free at match time to be added
2956 to the output.
2958 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2959 what's available on my current Linux desktop machine.
2961 3. When matching a UTF-8 string, the test for a valid string at the start has
2962 been extended. If start_offset is not zero, PCRE now checks that it points
2963 to a byte that is the start of a UTF-8 character. If not, it returns
2964 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2965 this is necessary because there may be backward assertions in the pattern.
2966 When matching the same subject several times, it may save resources to use
2967 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2969 4. The code for checking the validity of UTF-8 strings has been tightened so
2970 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2971 containing "overlong sequences".
2973 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2974 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2975 should have been "&", but it just so happened that all the cases this let
2976 through by mistake were picked up later in the function.
2978 6. I had used a variable called "isblank" - this is a C99 function, causing
2979 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2981 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2982 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2983 so the version is shown; (c) stop "make test" echoing "./RunTest".
2985 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2987 9. The prototype for memmove() for systems that don't have it was using
2988 size_t, but the inclusion of the header that defines size_t was later. I've
2989 moved the #includes for the C headers earlier to avoid this.
2991 10. Added some adjustments to the code to make it easier to compiler on certain
2992 special systems:
2994 (a) Some "const" qualifiers were missing.
2995 (b) Added the macro EXPORT before all exported functions; by default this
2996 is defined to be empty.
2997 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2998 that it reads its output file name as an argument instead of writing
2999 to the standard output and assuming this can be redirected.
3001 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3002 class containing characters with values greater than 255, PCRE compilation
3003 went into a loop.
3005 12. A recursive reference to a subpattern that was within another subpattern
3006 that had a minimum quantifier of zero caused PCRE to crash. For example,
3007 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3008 recursion. If the recursively-called subpattern itself had a zero repeat,
3009 that was OK.
3011 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3012 buffer into which it was copied (for escape processing) was still set at
3013 1024, so long lines caused crashes.
3015 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3016 "internal error: code overflow...". This applied to any character class
3017 that was followed by a possessive quantifier.
3019 15. Modified the Makefile to add libpcre.la as a prerequisite for
3020 libpcreposix.la because I was told this is needed for a parallel build to
3021 work.
3023 16. If a pattern that contained .* following optional items at the start was
3024 studied, the wrong optimizing data was generated, leading to matching
3025 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3026 matching string must start with a or b or c. The correct conclusion for
3027 this pattern is that a match can start with any character.
3030 Version 4.4 13-Aug-03
3031 ---------------------
3033 1. In UTF-8 mode, a character class containing characters with values between
3034 127 and 255 was not handled correctly if the compiled pattern was studied.
3035 In fixing this, I have also improved the studying algorithm for such
3036 classes (slightly).
3038 2. Three internal functions had redundant arguments passed to them. Removal
3039 might give a very teeny performance improvement.
3041 3. Documentation bug: the value of the capture_top field in a callout is *one
3042 more than* the number of the hightest numbered captured substring.
3044 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3045 in incorrectly linking with a previously installed version. They now link
3046 explicitly with libpcre.la.
3048 5. configure.in no longer needs to recognize Cygwin specially.
3050 6. A problem in pcre.in for Windows platforms is fixed.
3052 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3053 pcretest, it used to include the size of the study block as part of its
3054 output. Unfortunately, the structure contains a field that has a different
3055 size on different hardware architectures. This meant that the tests that
3056 showed this size failed. As the block is currently always of a fixed size,
3057 this information isn't actually particularly useful in pcretest output, so
3058 I have just removed it.
3060 8. Three pre-processor statements accidentally did not start in column 1.
3061 Sadly, there are *still* compilers around that complain, even though
3062 standard C has not required this for well over a decade. Sigh.
3064 9. In pcretest, the code for checking callouts passed small integers in the
3065 callout_data field, which is a void * field. However, some picky compilers
3066 complained about the casts involved for this on 64-bit systems. Now
3067 pcretest passes the address of the small integer instead, which should get
3068 rid of the warnings.
3070 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3071 both compile and run time, and gives an error if an invalid UTF-8 sequence
3072 is found. There is a option for disabling this check in cases where the
3073 string is known to be correct and/or the maximum performance is wanted.
3075 11. In response to a bug report, I changed one line in Makefile.in from
3077 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3078 to
3079 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3081 to look similar to other lines, but I have no way of telling whether this
3082 is the right thing to do, as I do not use Windows. No doubt I'll get told
3083 if it's wrong...
3086 Version 4.3 21-May-03
3087 ---------------------
3089 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3090 Makefile.
3092 2. Some refactoring to improve the quality of the code:
3094 (i) The utf8_table... variables are now declared "const".
3096 (ii) The code for \cx, which used the "case flipping" table to upper case
3097 lower case letters, now just substracts 32. This is ASCII-specific,
3098 but the whole concept of \cx is ASCII-specific, so it seems
3099 reasonable.
3101 (iii) PCRE was using its character types table to recognize decimal and
3102 hexadecimal digits in the pattern. This is silly, because it handles
3103 only 0-9, a-f, and A-F, but the character types table is locale-
3104 specific, which means strange things might happen. A private
3105 table is now used for this - though it costs 256 bytes, a table is
3106 much faster than multiple explicit tests. Of course, the standard
3107 character types table is still used for matching digits in subject
3108 strings against \d.
3110 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3111 ending in _t are). So I've renamed it as ESC_tee.
3113 3. The first argument for regexec() in the POSIX wrapper should have been
3114 defined as "const".
3116 4. Changed pcretest to use malloc() for its buffers so that they can be
3117 Electric Fenced for debugging.
3119 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3120 to read one or more bytes before the start of the subject string. Often this
3121 had no effect on PCRE's behaviour, but in some circumstances it could
3122 provoke a segmentation fault.
3124 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3125 to try to read one or more bytes before the start of the subject string.
3127 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3128 UTF-8 support could misbehave in various ways if the subject string
3129 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3130 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3131 back over UTF-8 characters.)
3134 Version 4.2 14-Apr-03
3135 ---------------------
3137 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3139 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3140 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3141 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3142 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3143 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3145 Note: automatic setting of the BUILD variables is not yet working
3146 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3147 compile-time but not at link-time
3148 [LINK]: use for linking executables only
3149 make different versions for Windows and non-Windows
3150 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3151 libraries
3152 [LINK_FOR_BUILD]: new variable
3153 [OBJEXT]: use throughout
3154 [EXEEXT]: use throughout
3155 <winshared>: new target
3156 <wininstall>: new target
3157 <dftables.o>: use native compiler
3158 <dftables>: use native linker
3159 <install>: handle Windows platform correctly
3160 <clean>: ditto
3161 <check>: ditto
3162 copy DLL to top builddir before testing
3164 As part of these changes, -no-undefined was removed again. This was reported
3165 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3166 in any case.
3168 3. Some tidies to get rid of compiler warnings:
3170 . In the match_data structure, match_limit was an unsigned long int, whereas
3171 match_call_count was an int. I've made them both unsigned long ints.
3173 . In pcretest the fact that a const uschar * doesn't automatically cast to
3174 a void * provoked a warning.
3176 . Turning on some more compiler warnings threw up some "shadow" variables
3177 and a few more missing casts.
3179 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3180 option, a class that contained a single character with a value between 128
3181 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3183 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3184 option, a class that contained several characters, but with at least one
3185 whose value was between 128 and 255 caused PCRE to crash.
3188 Version 4.1 12-Mar-03
3189 ---------------------
3191 1. Compiling with gcc -pedantic found a couple of places where casts were
3192 needed, and a string in dftables.c that was longer than standard compilers are
3193 required to support.
3195 2. Compiling with Sun's compiler found a few more places where the code could
3196 be tidied up in order to avoid warnings.
3198 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3199 first of these names is deprecated in the latest Autoconf in favour of the name
3200 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3201 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3202 analogy I have changed it to CFLAGS_FOR_BUILD.
3204 4. Added -no-undefined to the linking command in the Makefile, because this is
3205 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3206 linking step for the pcreposix library.
3208 5. PCRE was failing to diagnose the case of two named groups with the same
3209 name.
3211 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3212 literal character that is needed in the subject for a match, and scans along to
3213 ensure that it is present before embarking on the full matching process. This
3214 saves time in cases of nested unlimited repeats that are never going to match.
3215 Problem: the scan can take a lot of time if the subject is very long (e.g.
3216 megabytes), thus penalizing straightforward matches. It is now done only if the
3217 amount of subject to be scanned is less than 1000 bytes.
3219 7. A lesser problem with the same optimization is that it was recording the
3220 first character of an anchored pattern as "needed", thus provoking a search
3221 right along the subject, even when the first match of the pattern was going to
3222 fail. The "needed" character is now not set for anchored patterns, unless it
3223 follows something in the pattern that is of non-fixed length. Thus, it still
3224 fulfils its original purpose of finding quick non-matches in cases of nested
3225 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3228 Version 4.0 17-Feb-03
3229 ---------------------
3231 1. If a comment in an extended regex that started immediately after a meta-item
3232 extended to the end of string, PCRE compiled incorrect data. This could lead to
3233 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3235 2. Moved to autoconf 2.53 and libtool 1.4.2.
3237 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3238 the special perltest8 script is no longer needed - all the tests can be run
3239 from a single perltest script.
3241 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3242 by \s. It has now been removed in PCRE. This means it isn't recognized as
3243 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3244 class [:space:] *does* include VT, thereby creating a mess.
3246 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3247 space and tab.
3249 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3250 its new features into the main test script, reducing the number of scripts.
3252 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3253 were backward compatible, and made the (?i) apply to the whole pattern, as if
3254 /i were given. Now it behaves more logically, and applies the option setting
3255 only to what follows. PCRE has been changed to follow suit. However, if it
3256 finds options settings right at the start of the pattern, it extracts them into
3257 the global options, as before. Thus, they show up in the info data.
3259 8. Added support for the \Q...\E escape sequence. Characters in between are
3260 treated as literals. This is slightly different from Perl in that $ and @ are
3261 also handled as literals inside the quotes. In Perl, they will cause variable
3262 interpolation. Note the following examples:
3264 Pattern PCRE matches Perl matches
3266 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3267 \Qabc\$xyz\E abc\$xyz abc\$xyz
3268 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3270 For compatibility with Perl, \Q...\E sequences are recognized inside character
3271 classes as well as outside them.
3273 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3274 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3275 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3276 signed/unsigned warnings.
3278 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3279 option for pcretest, so I've replaced it by a simple function that does just
3280 that job.
3282 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3283 "pcregrep -".
3285 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3286 Java package. This provides some syntactic sugar for simple cases of what my
3287 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3288 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3289 item, you can use this simplified notation. Note that only makes sense with
3290 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3291 greediness, whatever the setting of the PCRE_UNGREEDY option.
3293 13. A change of greediness default within a pattern was not taking effect at
3294 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3295 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3296 was abstracted outside.
3298 14. PCRE now supports the \G assertion. It is true when the current matching
3299 position is at the start point of the match. This differs from \A when the
3300 starting offset is non-zero. Used with the /g option of pcretest (or similar
3301 code), it works in the same way as it does for Perl's /g option. If all
3302 alternatives of a regex begin with \G, the expression is anchored to the start
3303 match position, and the "anchored" flag is set in the compiled expression.
3305 15. Some bugs concerning the handling of certain option changes within patterns
3306 have been fixed. These applied to options other than (?ims). For example,
3307 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3308 been the other way round. Some of this was related to change 7 above.
3310 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3311 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3312 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3313 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3315 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3316 mode. Unlike ".", it always matches newline, whatever the setting of
3317 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3318 assertions. Perl allows it, but it doesn't (in general) work because it can't
3319 calculate the length of the lookbehind. At least, that's the case for Perl
3320 5.8.0 - I've been told they are going to document that it doesn't work in
3321 future.
3323 18. Added an error diagnosis for escapes that PCRE does not support: these are
3324 \L, \l, \N, \P, \p, \U, \u, and \X.
3326 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3327 reading past the end of the pattern in cases such as /[abcd/.
3329 20. PCRE was getting more memory than necessary for patterns with classes that
3330 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3332 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3333 compiling PCRE for use with Virtual Pascal.
3335 22. Small fix to the Makefile to make it work properly if the build is done
3336 outside the source tree.
3338 23. Added a new extension: a condition to go with recursion. If a conditional
3339 subpattern starts with (?(R) the "true" branch is used if recursion has
3340 happened, whereas the "false" branch is used only at the top level.
3342 24. When there was a very long string of literal characters (over 255 bytes
3343 without UTF support, over 250 bytes with UTF support), the computation of how
3344 much memory was required could be incorrect, leading to segfaults or other
3345 strange effects.
3347 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3348 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3349 there was a subsequent back reference to those brackets. This meant that, for
3350 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3351 possible to check for precisely this case. All we can do is abandon the
3352 optimization if .* occurs inside capturing brackets when there are any back
3353 references whatsoever. (See below for a better fix that came later.)
3355 26. The handling of the optimization for finding the first character of a
3356 non-anchored pattern, and for finding a character that is required later in the
3357 match were failing in some cases. This didn't break the matching; it just
3358 failed to optimize when it could. The way this is done has been re-implemented.
3360 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3362 28. Added a new feature that provides some of the functionality that Perl
3363 provides with (?{...}). The facility is termed a "callout". The way it is done
3364 in PCRE is for the caller to provide an optional function, by setting
3365 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3366 global variable. By default it is unset, which disables all calling out. To get
3367 the function called, the regex must include (?C) at appropriate points. This
3368 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3369 This provides a means of identifying different callout points. When PCRE
3370 reaches such a point in the regex, if pcre_callout has been set, the external
3371 function is called. It is provided with data in a structure called
3372 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3373 matching continues; if it returns a non-zero value, the match at the current
3374 point fails. However, backtracking will occur if possible. [This was changed
3375 later and other features added - see item 49 below.]
3377 29. pcretest is upgraded to test the callout functionality. It provides a
3378 callout function that displays information. By default, it shows the start of
3379 the match and the current position in the text. There are some new data escapes
3380 to vary what happens:
3382 \C+ in addition, show current contents of captured substrings
3383 \C- do not supply a callout function
3384 \C!n return 1 when callout number n is reached
3385 \C!n!m return 1 when callout number n is reached for the mth time
3387 30. If pcregrep was called with the -l option and just a single file name, it
3388 output "<stdin>" if a match was found, instead of the file name.
3390 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3391 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3392 pcre_exec(). This saves a malloc/free per call. The default value of
3393 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3394 when configuring.
3396 32. The default maximum size of a compiled pattern is 64K. There have been a
3397 few cases of people hitting this limit. The code now uses macros to handle the
3398 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3399 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3400 configuring. Tests 2 and 5 work only with 2-byte links because they output
3401 debugging information about compiled patterns.
3403 33. Internal code re-arrangements:
3405 (a) Moved the debugging function for printing out a compiled regex into
3406 its own source file (printint.c) and used #include to pull it into
3407 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3408 separate copies.
3410 (b) Defined the list of op-code names for debugging as a macro in
3411 internal.h so that it is next to the definition of the opcodes.
3413 (c) Defined a table of op-code lengths for simpler skipping along compiled
3414 code. This is again a macro in internal.h so that it is next to the
3415 definition of the opcodes.
3417 34. Added support for recursive calls to individual subpatterns, along the
3418 lines of Robin Houston's patch (but implemented somewhat differently).
3420 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3421 allow it to read and process whole directories in Win32. This code was
3422 contributed by Lionel Fourquaux; it has not been tested by me.
3424 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3425 used to name a group. Names consist of alphanumerics and underscores, and must
3426 be unique. Back references use the syntax (?P=name) and recursive calls use
3427 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3428 numbers. The function pcre_fullinfo() can be used after compilation to extract
3429 a name/number map. There are three relevant calls:
3431 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3432 PCRE_INFO_NAMECOUNT yields the number of entries
3433 PCRE_INFO_NAMETABLE yields a pointer to the map.
3435 The map is a vector of fixed-size entries. The size of each entry depends on
3436 the length of the longest name used. The first two bytes of each entry are the
3437 group number, most significant byte first. There follows the corresponding
3438 name, zero terminated. The names are in alphabetical order.
3440 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3441 case instead of 255. Making it the same both with and without UTF-8 support
3442 means that the same test output works with both.
3444 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3445 calling malloc() with a zero argument.
3447 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3448 optimization. I've improved things by keeping a bitmap of backreferences with
3449 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3450 fact referenced, the optimization can be applied. It is unlikely that a
3451 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3452 the match to follow \n) will appear inside brackets with a number greater than
3453 31, but if it does, any back reference > 31 suppresses the optimization.
3455 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3456 of disabling numbered capturing parentheses. Any opening parenthesis that is
3457 not followed by ? behaves as if it were followed by ?: but named parentheses
3458 can still be used for capturing (and they will acquire numbers in the usual
3459 way).
3461 41. Redesigned the return codes from the match() function into yes/no/error so
3462 that errors can be passed back from deep inside the nested calls. A malloc
3463 failure while inside a recursive subpattern call now causes the
3464 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3466 42. It is now possible to set a limit on the number of times the match()
3467 function is called in a call to pcre_exec(). This facility makes it possible to
3468 limit the amount of recursion and backtracking, though not in a directly
3469 obvious way, because the match() function is used in a number of different
3470 circumstances. The count starts from zero for each position in the subject
3471 string (for non-anchored patterns). The default limit is, for compatibility, a
3472 large number, namely 10 000 000. You can change this in two ways:
3474 (a) When configuring PCRE before making, you can use --with-match-limit=n
3475 to set a default value for the compiled library.
3477 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3478 a different value is set. See 45 below.
3480 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3482 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3483 of things that can be changed at compile time. The first argument specifies
3484 what is wanted and the second points to where the information is to be placed.
3485 The current list of available information is:
3489 The output is an integer that is set to one if UTF-8 support is available;
3490 otherwise it is set to zero.
3494 The output is an integer that it set to the value of the code that is used for
3495 newline. It is either LF (10) or CR (13).
3499 The output is an integer that contains the number of bytes used for internal
3500 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3504 The output is an integer that contains the threshold above which the POSIX
3505 interface uses malloc() for output vectors. See item 31 above.
3509 The output is an unsigned integer that contains the default limit of the number
3510 of match() calls in a pcre_exec() execution. See 42 above.
3512 44. pcretest has been upgraded by the addition of the -C option. This causes it
3513 to extract all the available output from the new pcre_config() function, and to
3514 output it. The program then exits immediately.
3516 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3517 order to support additional features. One way would have been to define
3518 pcre_exec2() (for example) with extra arguments, but this would not have been
3519 extensible, and would also have required all calls to the original function to
3520 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3521 is used for passing in "extra" data from pcre_study().
3523 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3524 contains the following fields:
3526 flags a bitmap indicating which of the following fields are set
3527 study_data opaque data from pcre_study()
3528 match_limit a way of specifying a limit on match() calls for a specific
3529 call to pcre_exec()
3530 callout_data data for callouts (see 49 below)
3532 The flag bits are also defined in pcre.h, and are
3538 The pcre_study() function now returns one of these new pcre_extra blocks, with
3539 the actual study data pointed to by the study_data field, and the
3540 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3541 before. That is, this change is entirely upwards-compatible and requires no
3542 change to existing code.
3544 If you want to pass in additional data to pcre_exec(), you can either place it
3545 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3546 block.
3548 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3549 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3550 times with different match limits, until it finds the minimum value needed for
3551 pcre_exec() to complete. The value is then output. This can be instructive; for
3552 most simple matches the number is quite small, but for pathological cases it
3553 gets very large very quickly.
3555 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3556 returns the size of the data block pointed to by the study_data field in a
3557 pcre_extra block, that is, the value that was passed as the argument to
3558 pcre_malloc() when PCRE was getting memory in which to place the information
3559 created by pcre_study(). The fourth argument should point to a size_t variable.
3560 pcretest has been extended so that this information is shown after a successful
3561 pcre_study() call when information about the compiled regex is being displayed.
3563 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3564 because what follows is always an absolute path. (Later: it turns out that this
3565 is more than cosmetic for MinGW, because it doesn't like empty path
3566 components.)
3568 49. Some changes have been made to the callout feature (see 28 above):
3570 (i) A callout function now has three choices for what it returns:
3572 0 => success, carry on matching
3573 > 0 => failure at this point, but backtrack if possible
3574 < 0 => serious error, return this value from pcre_exec()
3576 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3577 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3578 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3579 use by callout functions. It will never be used by PCRE itself.
3581 (ii) The pcre_extra structure (see 45 above) has a void * field called
3582 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3583 pcre_callout_block structure has a field of the same name. The contents of
3584 the field passed in the pcre_extra structure are passed to the callout
3585 function in the corresponding field in the callout block. This makes it
3586 easier to use the same callout-containing regex from multiple threads. For
3587 testing, the pcretest program has a new data escape
3589 \C*n pass the number n (may be negative) as callout_data
3591 If the callout function in pcretest receives a non-zero value as
3592 callout_data, it returns that value.
3594 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3595 there were some redundant $(CFLAGS) in commands that are now specified as
3596 $(LINK), which already includes $(CFLAGS).
3598 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3599 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3600 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3601 one-byte characters throughout. Note that case-insensitive matching applies
3602 only to characters whose values are less than 256. PCRE doesn't support the
3603 notion of cases for higher-valued characters.
3605 (i) A character class whose characters are all within 0-255 is handled as
3606 a bit map, and the map is inverted for negative classes. Previously, a
3607 character > 255 always failed to match such a class; however it should
3608 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3610 (ii) A negated character class with a single character < 255 is coded as
3611 "not this character" (OP_NOT). This wasn't working properly when the test
3612 character was multibyte, either singly or repeated.
3614 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3615 mode, for example: \x{100}{2,3}.
3617 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3618 singly or repeated) now correctly test multibyte characters. However,
3619 PCRE doesn't recognize any characters with values greater than 255 as
3620 digits, spaces, or word characters. Such characters always match \D, \S,
3621 and \W, and never match \d, \s, or \w.
3623 (v) Classes may now contain characters and character ranges with values
3624 greater than 255. For example: [ab\x{100}-\x{400}].
3626 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3627 PCRE in UTF-8 mode.
3629 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3630 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3631 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3632 value.)
3634 53. The single man page has become too large. I have therefore split it up into
3635 a number of separate man pages. These also give rise to individual HTML pages;
3636 these are now put in a separate directory, and there is an index.html page that
3637 lists them all. Some hyperlinking between the pages has been installed.
3639 54. Added convenience functions for handling named capturing parentheses.
3641 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3642 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3643 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3644 are faulted.
3646 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3647 calling configure. These values are used when compiling the dftables.c program
3648 which is run to generate the source of the default character tables. They
3649 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3650 you will need to set these values.
3652 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3655 Version 3.9 02-Jan-02
3656 ---------------------
3658 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3660 2. If --disable-static was given, the building process failed when trying to
3661 build pcretest and pcregrep. (For some reason it was using libtool to compile
3662 them, which is not right, as they aren't part of the library.)
3665 Version 3.8 18-Dec-01
3666 ---------------------
3668 1. The experimental UTF-8 code was completely screwed up. It was packing the
3669 bytes in the wrong order. How dumb can you get?
3672 Version 3.7 29-Oct-01
3673 ---------------------
3675 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3676 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3677 this didn't happen under Solaris 8, where I normally test things.
3679 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3680 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3681 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3682 this fix an hour or so after the initial 3.7 release.)
3685 Version 3.6 23-Oct-01
3686 ---------------------
3688 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3689 offsets passed as NULL with zero offset count.
3691 2. The config.guess and config.sub files had not been updated when I moved to
3692 the latest autoconf.
3695 Version 3.5 15-Aug-01
3696 ---------------------
3698 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3699 had been forgotten.
3701 2. By using declared but undefined structures, we can avoid using "void"
3702 definitions in pcre.h while keeping the internal definitions of the structures
3703 private.
3705 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3706 user point of view, this means that both static and shared libraries are built
3707 by default, but this can be individually controlled. More of the work of
3708 handling this static/shared cases is now inside libtool instead of PCRE's make
3709 file.
3711 4. The pcretest utility is now installed along with pcregrep because it is
3712 useful for users (to test regexs) and by doing this, it automatically gets
3713 relinked by libtool. The documentation has been turned into a man page, so
3714 there are now .1, .txt, and .html versions in /doc.
3716 5. Upgrades to pcregrep:
3717 (i) Added long-form option names like gnu grep.
3718 (ii) Added --help to list all options with an explanatory phrase.
3719 (iii) Added -r, --recursive to recurse into sub-directories.
3720 (iv) Added -f, --file to read patterns from a file.
3722 6. pcre_exec() was referring to its "code" argument before testing that
3723 argument for NULL (and giving an error if it was NULL).
3725 7. Upgraded Makefile.in to allow for compiling in a different directory from
3726 the source directory.
3728 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3729 options bits, the pointer it was passed was to an int instead of to an unsigned
3730 long int. This mattered only on 64-bit systems.
3732 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3733 generated) instead of pcre.in, which it its source. Also made the same change
3734 in several of the .c files.
3736 10. A new release of gcc defines printf() as a macro, which broke pcretest
3737 because it had an ifdef in the middle of a string argument for printf(). Fixed
3738 by using separate calls to printf().
3740 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3741 script, to force use of CR or LF instead of \n in the source. On non-Unix
3742 systems, the value can be set in config.h.
3744 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3745 absolute limit. Changed the text of the error message to make this clear, and
3746 likewise updated the man page.
3748 13. The limit of 99 on the number of capturing subpatterns has been removed.
3749 The new limit is 65535, which I hope will not be a "real" limit.
3752 Version 3.4 22-Aug-00
3753 ---------------------
3755 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3757 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3760 Version 3.3 01-Aug-00
3761 ---------------------
3763 1. If an octal character was given, but the value was greater than \377, it
3764 was not getting masked to the least significant bits, as documented. This could
3765 lead to crashes in some systems.
3767 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3768 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3770 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3771 These just pass their arguments on to (pcre_free)(), but they are provided
3772 because some uses of PCRE bind it to non-C systems that can call its functions,
3773 but cannot call free() or pcre_free() directly.
3775 4. Add "make test" as a synonym for "make check". Corrected some comments in
3776 the Makefile.
3778 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3779 Makefile.
3781 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3782 command called pgrep for grepping around the active processes.
3784 7. Added the beginnings of support for UTF-8 character strings.
3786 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3787 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3788 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3789 out for the ar command.)
3792 Version 3.2 12-May-00
3793 ---------------------
3795 This is purely a bug fixing release.
3797 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3798 of ZA. This was just one example of several cases that could provoke this bug,
3799 which was introduced by change 9 of version 2.00. The code for breaking
3800 infinite loops after an iteration that matches an empty string was't working
3801 correctly.
3803 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3804 when matched against abbab (for example). After matching an empty string, it
3805 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3806 caused it to match further down the string than it should.
3808 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3809 was there because it doesn't seem to be needed, and it causes trouble on some
3810 systems, as it is not a Standard C header. It has been removed.
3812 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3813 were reported on the Macintosh. The changes were from
3815 while ((c = *(++ptr)) != 0 && c != '\n');
3816 to
3817 while ((c = *(++ptr)) != 0 && c != '\n') ;
3819 Totally extraordinary, but if that's what it takes...
3821 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3822 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3823 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3824 assumes the way PCRE uses memmove() (always moving upwards).
3826 6. PCRE is being used in one environment where strchr() is not available. There
3827 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3828 faster code anyway.
3831 Version 3.1 09-Feb-00
3832 ---------------------
3834 The only change in this release is the fixing of some bugs in Makefile.in for
3835 the "install" target:
3837 (1) It was failing to install pcreposix.h.
3839 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3842 Version 3.0 01-Feb-00
3843 ---------------------
3845 1. Add support for the /+ modifier to perltest (to output $` like it does in
3846 pcretest).
3848 2. Add support for the /g modifier to perltest.
3850 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3851 matches null strings.
3853 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3854 pattern. Perl treats empty patterns specially - it reuses the most recent
3855 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3856 effect.
3858 5. The POSIX interface was broken in that it was just handing over the POSIX
3859 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3860 required a bigger vector, with some working space on the end. This means that
3861 the POSIX wrapper now has to get and free some memory, and copy the results.
3863 6. Added some simple autoconf support, placing the test data and the
3864 documentation in separate directories, re-organizing some of the
3865 information files, and making it build pcre-config (a GNU standard). Also added
3866 libtool support for building PCRE as a shared library, which is now the
3867 default.
3869 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3870 09 are not valid octal constants. Single digits will be used for minor values
3871 less than 10.
3873 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3874 existing programs that set these in the POSIX interface can use PCRE without
3875 modification.
3877 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3878 return all that pcre_info() returns, plus additional data. The pcre_info()
3879 function is retained for compatibility, but is considered to be obsolete.
3881 10. Added experimental recursion feature (?R) to handle one common case that
3882 Perl 5.6 will be able to do with (?p{...}).
3884 11. Added support for POSIX character classes like [:alpha:], which Perl is
3885 adopting.
3888 Version 2.08 31-Aug-99
3889 ----------------------
3891 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3892 trying to match at the startoffset position, but instead was moving forward to
3893 the next newline as if a previous match had failed.
3895 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3896 and could get into a loop if a null string was matched other than at the start
3897 of the subject.
3899 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3900 be distinguished at compile time, and for completeness also added PCRE_DATE.
3902 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3903 in GnuWin32 environments.
3906 Version 2.07 29-Jul-99
3907 ----------------------
3909 1. The documentation is now supplied in plain text form and HTML as well as in
3910 the form of man page sources.
3912 2. C++ compilers don't like assigning (void *) values to other pointer types.
3913 In particular this affects malloc(). Although there is no problem in Standard
3914 C, I've put in casts to keep C++ compilers happy.
3916 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3917 should be (const char *).
3919 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3920 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3921 However, I haven't made this a standard facility. The documentation doesn't
3922 mention it, and the Makefile doesn't support it.
3924 5. The Makefile now contains an "install" target, with editable destinations at
3925 the top of the file. The pcretest program is not installed.
3927 6. pgrep -V now gives the PCRE version number and date.
3929 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3930 causing the entire string to be ignored, instead of just the last character.
3932 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3933 non-matching string, it can take a very, very long time, even for strings of
3934 quite modest length, because of the nested recursion. PCRE now does better in
3935 some of these cases. It does this by remembering the last required literal
3936 character in the pattern, and pre-searching the subject to ensure it is present
3937 before running the real match. In other words, it applies a heuristic to detect
3938 some types of certain failure quickly, and in the above example, if presented
3939 with a string that has no trailing " it gives "no match" very quickly.
3941 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3942 other alternatives are tried instead.
3945 Version 2.06 09-Jun-99
3946 ----------------------
3948 1. Change pcretest's output for amount of store used to show just the code
3949 space, because the remainder (the data block) varies in size between 32-bit and
3950 64-bit systems.
3952 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3953 start matching at. This allows lookbehinds to work when searching for multiple
3954 occurrences in a string.
3956 3. Added additional options to pcretest for testing multiple occurrences:
3958 /+ outputs the rest of the string that follows a match
3959 /g loops for multiple occurrences, using the new startoffset argument
3960 /G loops for multiple occurrences by passing an incremented pointer
3962 4. PCRE wasn't doing the "first character" optimization for patterns starting
3963 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3964 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3965 the letter 'x'. On long subject strings, this gives a significant speed-up.
3968 Version 2.05 21-Apr-99
3969 ----------------------
3971 1. Changed the type of magic_number from int to long int so that it works
3972 properly on 16-bit systems.
3974 2. Fixed a bug which caused patterns starting with .* not to work correctly
3975 when the subject string contained newline characters. PCRE was assuming
3976 anchoring for such patterns in all cases, which is not correct because .* will
3977 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3978 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3979 must be retried after every newline in the subject.
3982 Version 2.04 18-Feb-99
3983 ----------------------
3985 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3986 computation of the store needed to hold the pattern was incorrect (too large).
3987 If such patterns were nested a few deep, this could multiply and become a real
3988 problem.
3990 2. Added /M option to pcretest to show the memory requirement of a specific
3991 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3993 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3994 compiled in such a way that the backtracking after subsequent failure was
3995 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3996 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3999 Version 2.03 02-Feb-99
4000 ----------------------
4002 1. Fixed typo and small mistake in man page.
4004 2. Added 4th condition (GPL supersedes if conflict) and created separate
4005 LICENCE file containing the conditions.
4007 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4008 Perl, that is the internal \ allows the delimiter to be included in the
4009 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4010 the final delimiter, add \ to the end of the pattern (to test the error).
4012 4. Added the convenience functions for extracting substrings after a successful
4013 match. Updated pcretest to make it able to test these functions.
4016 Version 2.02 14-Jan-99
4017 ----------------------
4019 1. Initialized the working variables associated with each extraction so that
4020 their saving and restoring doesn't refer to uninitialized store.
4022 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4023 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4024 fix the problem.
4026 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4027 calls, and wasn't printing the correct value for compiling calls. Increased the
4028 default value of LOOPREPEAT, and the number of significant figures in the
4029 times.
4031 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4033 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4034 a building problem on Windows NT with a FAT file system.
4037 Version 2.01 21-Oct-98
4038 ----------------------
4040 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4041 to character tables built by pcre_maketables() in the current locale. If NULL
4042 is passed, the default tables are used.
4045 Version 2.00 24-Sep-98
4046 ----------------------
4048 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4049 it any more.
4051 2. Allow quantification of (?>) groups, and make it work correctly.
4053 3. The first character computation wasn't working for (?>) groups.
4055 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4056 end of the subject) and add 5.005's \z, which really does match only at the
4057 very end of the subject.
4059 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4061 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4062 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4063 localized options. All options to pcre_study() were also removed.
4065 7. Add other new features from 5.005:
4067 $(?<= positive lookbehind
4068 $(?<! negative lookbehind
4069 (?imsx-imsx) added the unsetting capability
4070 such a setting is global if at outer level; local otherwise
4071 (?imsx-imsx:) non-capturing groups with option setting
4072 (?(cond)re|re) conditional pattern matching
4074 A backreference to itself in a repeated group matches the previous
4075 captured string.
4077 8. General tidying up of studying (both automatic and via "study")
4078 consequential on the addition of new assertions.
4080 9. As in 5.005, unlimited repeated groups that could match an empty substring
4081 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4082 runtime if any iteration does actually match an empty substring.
4084 10. Include the RunTest script in the distribution.
4086 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4087 discrepancies, some of which were old and were also with respect to 5.004. They
4088 have now been fixed.
4091 Version 1.09 28-Apr-98
4092 ----------------------
4094 1. A negated single character class followed by a quantifier with a minimum
4095 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4096 program crashes, or just wrong answers. This did not apply to negated classes
4097 containing more than one character, or to minima other than one.
4100 Version 1.08 27-Mar-98
4101 ----------------------
4103 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4105 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4106 latter must appear before anything that relies on it in the pattern.
4109 Version 1.07 16-Feb-98
4110 ----------------------
4112 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4113 repeat of a potentially empty string).
4116 Version 1.06 23-Jan-98
4117 ----------------------
4119 1. Added Markus Oberhumer's little patches for C++.
4121 2. Literal strings longer than 255 characters were broken.
4124 Version 1.05 23-Dec-97
4125 ----------------------
4127 1. Negated character classes containing more than one character were failing if
4128 PCRE_CASELESS was set at run time.
4131 Version 1.04 19-Dec-97
4132 ----------------------
4134 1. Corrected the man page, where some "const" qualifiers had been omitted.
4136 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4137 input syntax.
4139 3. Fixed memory leak which occurred when a regex with back references was
4140 matched with an offsets vector that wasn't big enough. The temporary memory
4141 that is used in this case wasn't being freed if the match failed.
4143 4. Tidied pcretest to ensure it frees memory that it gets.
4145 5. Temporary memory was being obtained in the case where the passed offsets
4146 vector was exactly big enough.
4148 6. Corrected definition of offsetof() from change 5 below.
4150 7. I had screwed up change 6 below and broken the rules for the use of
4151 setjmp(). Now fixed.
4154 Version 1.03 18-Dec-97
4155 ----------------------
4157 1. A erroneous regex with a missing opening parenthesis was correctly
4158 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4159 on some systems.
4161 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4162 it was reported that one broken compiler failed on the former because "code" is
4163 also an independent variable.
4165 3. The erroneous regex a[]b caused an array overrun reference.
4167 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4168 fail on data ending with that character. (It was going on too far, and checking
4169 the next character, typically a binary zero.) This was specific to the
4170 optimized code for single-character negative classes.
4172 5. Added a contributed patch from the TIN world which does the following:
4174 + Add an undef for memmove, in case the the system defines a macro for it.
4176 + Add a definition of offsetof(), in case there isn't one. (I don't know
4177 the reason behind this - offsetof() is part of the ANSI standard - but
4178 it does no harm).
4180 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4181 most of the places where whitespace preceded '#'. I have given up and
4182 allowed the remaining 2 cases to be at the margin.
4184 + Rename some variables in pcre to eliminate shadowing. This seems very
4185 pedantic, but does no harm, of course.
4187 6. Moved the call to setjmp() into its own function, to get rid of warnings
4188 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4190 7. Constructs such as \d{8,} were compiling into the equivalent of
4191 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4192 outcome, but in this particular case used more store than had been allocated,
4193 which caused the bug to be discovered because it threw up an internal error.
4195 8. The debugging code in both pcre and pcretest for outputting the compiled
4196 form of a regex was going wrong in the case of back references followed by
4197 curly-bracketed repeats.
4200 Version 1.02 12-Dec-97
4201 ----------------------
4203 1. Typos in pcre.3 and comments in the source fixed.
4205 2. Applied a contributed patch to get rid of places where it used to remove
4206 'const' from variables, and fixed some signed/unsigned and uninitialized
4207 variable warnings.
4209 3. Added the "runtest" target to Makefile.
4211 4. Set default compiler flag to -O2 rather than just -O.
4214 Version 1.01 19-Nov-97
4215 ----------------------
4217 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4218 like /([ab]*)*/, that is, for classes with more than one character in them.
4220 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4221 as /((?>a*))*/ (a PCRE_EXTRA facility).
4224 Version 1.00 18-Nov-97
4225 ----------------------
4227 1. Added compile-time macros to support systems such as SunOS4 which don't have
4228 memmove() or strerror() but have other things that can be used instead.
4230 2. Arranged that "make clean" removes the executables.
4233 Version 0.99 27-Oct-97
4234 ----------------------
4236 1. Fixed bug in code for optimizing classes with only one character. It was
4237 initializing a 32-byte map regardless, which could cause it to run off the end
4238 of the memory it had got.
4240 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4243 Version 0.98 22-Oct-97
4244 ----------------------
4246 1. Fixed bug in code for handling temporary memory usage when there are more
4247 back references than supplied space in the ovector. This could cause segfaults.
4250 Version 0.97 21-Oct-97
4251 ----------------------
4253 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4255 2. Optimized negated single characters not to use a bit map.
4257 3. Brought error texts together as macro definitions; clarified some of them;
4258 fixed one that was wrong - it said "range out of order" when it meant "invalid
4259 escape sequence".
4261 4. Changed some char * arguments to const char *.
4263 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4265 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4266 pcretest.
4269 Version 0.96 16-Oct-97
4270 ----------------------
4272 1. Added a simple "pgrep" utility to the distribution.
4274 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4275 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4276 where "ddd" means "one or more decimal digits".
4278 3. Fixed serious bug. If a pattern had a back reference, but the call to
4279 pcre_exec() didn't supply a large enough ovector to record the related
4280 identifying subpattern, the match always failed. PCRE now remembers the number
4281 of the largest back reference, and gets some temporary memory in which to save
4282 the offsets during matching if necessary, in order to ensure that
4283 backreferences always work.
4285 4. Increased the compatibility with Perl in a number of ways:
4287 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4288 to request this handling. The option can be set at compile or exec time.
4290 (b) $ matches before a terminating newline by default; an option
4291 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4292 mode). The option can be set at compile or exec time.
4294 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4295 the same as Perl's. If the decimal number it represents is less than 10
4296 or there aren't that many previous left capturing parentheses, an octal
4297 escape is read. Inside a character class, it's always an octal escape,
4298 even if it is a single digit.
4300 (d) An escaped but undefined alphabetic character is taken as a literal,
4301 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4302 escapes.
4304 (e) {0} is now permitted. (The previous item is removed from the compiled
4305 pattern).
4307 5. Changed all the names of code files so that the basic parts are no longer
4308 than 10 characters, and abolished the teeny "globals.c" file.
4310 6. Changed the handling of character classes; they are now done with a 32-byte
4311 bit map always.
4313 7. Added the -d and /D options to pcretest to make it possible to look at the
4314 internals of compilation without having to recompile pcre.
4317 Version 0.95 23-Sep-97
4318 ----------------------
4320 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4321 \x20 at the start of a run of normal characters. These were being treated as
4322 real characters, instead of the source characters being re-checked.
4325 Version 0.94 18-Sep-97
4326 ----------------------
4328 1. The functions are now thread-safe, with the caveat that the global variables
4329 containing pointers to malloc() and free() or alternative functions are the
4330 same for all threads.
4332 2. Get pcre_study() to generate a bitmap of initial characters for non-
4333 anchored patterns when this is possible, and use it if passed to pcre_exec().
4336 Version 0.93 15-Sep-97
4337 ----------------------
4339 1. /(b)|(:+)/ was computing an incorrect first character.
4341 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4342 but not actually doing anything yet.
4344 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4345 as Perl does (e.g. [-az] or [az-]).
4347 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4348 all possible positions.
4350 5. Split up into different modules to avoid including unneeded functions in a
4351 compiled binary. However, compile and exec are still in one module. The "study"
4352 function is split off.
4354 6. The character tables are now in a separate module whose source is generated
4355 by an auxiliary program - but can then be edited by hand if required. There are
4356 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4357 toupper() in the code.
4359 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4360 make them global. Abolish the function for setting them, as the caller can now
4361 set them directly.
4364 Version 0.92 11-Sep-97
4365 ----------------------
4367 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4368 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4370 2. Caseless matching was not working in character classes if the characters in
4371 the pattern were in upper case.
4373 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4375 4. Make PCRE_ANCHORED public and accept as a compile option.
4377 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4378 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4379 pass them.
4381 6. Give an error if bad option bits passed at compile or run time.
4383 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4384 pcretest to cause it to pass that flag.
4386 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4387 options, and the first character, if set.
4389 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4392 Version 0.91 10-Sep-97
4393 ----------------------
4395 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4396 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4398 2. PCRE was looping on encountering an indefinitely repeated back reference to
4399 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4400 Perl does - treats the match as successful.
4402 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5