ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 661 - (show annotations)
Sun Aug 21 09:00:54 2011 UTC (10 years, 1 month ago) by ph10
File size: 211607 byte(s)
Fix bug introduced by 8.13/37 concerning POSIX class recognition
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 in
10 a POSIX name.
13 Version 8.13 16-Aug-2011
14 ------------------------
16 1. The Unicode data tables have been updated to Unicode 6.0.0.
18 2. Two minor typos in pcre_internal.h have been fixed.
20 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
21 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
22 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
24 4. There were a number of related bugs in the code for matching backrefences
25 caselessly in UTF-8 mode when codes for the characters concerned were
26 different numbers of bytes. For example, U+023A and U+2C65 are an upper
27 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
28 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
29 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
30 2-byte code at the end of the subject (it thought there wasn't enough data
31 left).
33 5. Comprehensive information about what went wrong is now returned by
34 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
35 as the output vector has at least 2 elements. The offset of the start of
36 the failing character and a reason code are placed in the vector.
38 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
39 now returned is for the first byte of the failing character, instead of the
40 last byte inspected. This is an incompatible change, but I hope it is small
41 enough not to be a problem. It makes the returned offset consistent with
42 pcre_exec() and pcre_dfa_exec().
44 7. pcretest now gives a text phrase as well as the error number when
45 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
46 failure, the offset and reason code are output.
48 8. When \R was used with a maximizing quantifier it failed to skip backwards
49 over a \r\n pair if the subsequent match failed. Instead, it just skipped
50 back over a single character (\n). This seems wrong (because it treated the
51 two characters as a single entity when going forwards), conflicts with the
52 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
53 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
54 has been changed.
56 9. Some internal refactoring has changed the processing so that the handling
57 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
58 time (the PCRE_DOTALL option was changed this way some time ago: version
59 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
60 which was always a bit of a fudge. It also means that there is one less
61 argument for the match() function, which reduces its stack requirements
62 slightly. This change also fixes an incompatibility with Perl: the pattern
63 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
65 10. More internal refactoring has drastically reduced the number of recursive
66 calls to match() for possessively repeated groups such as (abc)++ when
67 using pcre_exec().
69 11. While implementing 10, a number of bugs in the handling of groups were
70 discovered and fixed:
72 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
73 (a|)*(?1) gave a compile-time internal error.
74 ((a|)+)+ did not notice that the outer group could match an empty string.
75 (^a|^)+ was not marked as anchored.
76 (.*a|.*)+ was not marked as matching at start or after a newline.
78 12. Yet more internal refactoring has removed another argument from the match()
79 function. Special calls to this function are now indicated by setting a
80 value in a variable in the "match data" data block.
82 13. Be more explicit in pcre_study() instead of relying on "default" for
83 opcodes that mean there is no starting character; this means that when new
84 ones are added and accidentally left out of pcre_study(), testing should
85 pick them up.
87 14. The -s option of pcretest has been documented for ages as being an old
88 synonym of -m (show memory usage). I have changed it to mean "force study
89 for every regex", that is, assume /S for every regex. This is similar to -i
90 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
91 using it. It makes it easier to run collections of tests with and without
92 study enabled, and thereby test pcre_study() more easily. All the standard
93 tests are now run with and without -s (but some patterns can be marked as
94 "never study" - see 20 below).
96 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
97 restoration of the capturing data to the outer values was not happening
98 correctly.
100 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
101 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
102 pattern had matched an empty string, and so incorrectly returned a no
103 match.
105 17. There was optimizing code for the last branch of non-capturing parentheses,
106 and also for the obeyed branch of a conditional subexpression, which used
107 tail recursion to cut down on stack usage. Unfortunately, now that there is
108 the possibility of (*THEN) occurring in these branches, tail recursion is
109 no longer possible because the return has to be checked for (*THEN). These
110 two optimizations have therefore been removed.
112 18. If a pattern containing \R was studied, it was assumed that \R always
113 matched two bytes, thus causing the minimum subject length to be
114 incorrectly computed because \R can also match just one byte.
116 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
117 was incorrectly computed.
119 20. If /S is present twice on a test pattern in pcretest input, it now
120 *disables* studying, thereby overriding the use of -s on the command line
121 (see 14 above). This is necessary for one or two tests to keep the output
122 identical in both cases.
124 21. When (*ACCEPT) was used in an assertion that matched an empty string and
125 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
127 22. When an atomic group that contained a capturing parenthesis was
128 successfully matched, but the branch in which it appeared failed, the
129 capturing was not being forgotten if a higher numbered group was later
130 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
131 group 1 to "a", when in fact it should be unset. This applied to multi-
132 branched capturing and non-capturing groups, repeated or not, and also to
133 positive assertions (capturing in negative assertions does not happen
134 in PCRE) and also to nested atomic groups.
136 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
137 subject after a captured substring, to make it easier to tell which of a
138 number of identical substrings has been captured.
140 24. The way atomic groups are processed by pcre_exec() has been changed so that
141 if they are repeated, backtracking one repetition now resets captured
142 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
143 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
144 "aaa". Previously, it would have been "a". As part of this code
145 refactoring, the way recursive calls are handled has also been changed.
147 25. If an assertion condition captured any substrings, they were not passed
148 back unless some other capturing happened later. For example, if
149 (?(?=(a))a) was matched against "a", no capturing was returned.
151 26. When studying a pattern that contained subroutine calls or assertions,
152 the code for finding the minimum length of a possible match was handling
153 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
154 group 1 called group 2 while simultaneously a separate group 2 called group
155 1). A stack overflow occurred in this case. I have fixed this by limiting
156 the recursion depth to 10.
158 27. Updated RunTest.bat in the distribution to the version supplied by Tom
159 Fortmann. This supports explicit test numbers on the command line, and has
160 argument validation and error reporting.
162 28. An instance of \X with an unlimited repeat could fail if at any point the
163 first character it looked at was a mark character.
165 29. Some minor code refactoring concerning Unicode properties and scripts
166 should reduce the stack requirement of match() slightly.
168 30. Added the '=' option to pcretest to check the setting of unused capturing
169 slots at the end of the pattern, which are documented as being -1, but are
170 not included in the return count.
172 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
173 compiled something random. Now it gives a compile-time error (as does
174 Perl).
176 32. A *MARK encountered during the processing of a positive assertion is now
177 recorded and passed back (compatible with Perl).
179 33. If --only-matching or --colour was set on a pcregrep call whose pattern
180 had alternative anchored branches, the search for a second match in a line
181 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
182 matched the line "0102" twice. The same bug affected patterns that started
183 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
184 twice.
186 34. Previously, PCRE did not allow quantification of assertions. However, Perl
187 does, and because of capturing effects, quantifying parenthesized
188 assertions may at times be useful. Quantifiers are now allowed for
189 parenthesized assertions.
191 35. A minor code tidy in pcre_compile() when checking options for \R usage.
193 36. \g was being checked for fancy things in a character class, when it should
194 just be a literal "g".
196 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
197 appearance of a nested POSIX class supersedes an apparent external class.
198 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
199 unescaped square brackets may also appear as part of class names. For
200 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
201 more like Perl. (But see 8.20/1 above.)
203 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
204 was because it thought it was \N{name}, which is not supported).
206 39. Add minix to OS list not supporting the -S option in pcretest.
208 40. PCRE tries to detect cases of infinite recursion at compile time, but it
209 cannot analyze patterns in sufficient detail to catch mutual recursions
210 such as ((?1))((?2)). There is now a runtime test that gives an error if a
211 subgroup is called recursively as a subpattern for a second time at the
212 same position in the subject string. In previous releases this might have
213 been caught by the recursion limit, or it might have run out of stack.
215 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
216 happen only once. PCRE was, however incorrectly giving a compile time error
217 "recursive call could loop indefinitely" because it cannot analyze the
218 pattern in sufficient detail. The compile time test no longer happens when
219 PCRE is compiling a conditional subpattern, but actual runaway loops are
220 now caught at runtime (see 40 above).
222 42. It seems that Perl allows any characters other than a closing parenthesis
223 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
224 has been changed to be the same.
226 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
227 as not to get warnings when autogen.sh is called. Also changed
228 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
230 44. To help people who use pcregrep to scan files containing exceedingly long
231 lines, the following changes have been made:
233 (a) The default value of the buffer size parameter has been increased from
234 8K to 20K. (The actual buffer used is three times this size.)
236 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
237 PCRE is built.
239 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
240 to be set at run time.
242 (d) Numerical values in pcregrep options can be followed by K or M, for
243 example --buffer-size=50K.
245 (e) If a line being scanned overflows pcregrep's buffer, an error is now
246 given and the return code is set to 2.
248 45. Add a pointer to the latest mark to the callout data block.
250 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
251 partial match of an empty string instead of no match. This was specific to
252 the use of ".".
254 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
255 complete match instead of a partial match. This bug was dependent on both
256 the PCRE_UTF8 and PCRE_DOTALL options being set.
258 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
259 starting byte set, because \b was not being ignored.
262 Version 8.12 15-Jan-2011
263 ------------------------
265 1. Fixed some typos in the markup of the man pages, and wrote a script that
266 checks for such things as part of the documentation building process.
268 2. On a big-endian 64-bit system, pcregrep did not correctly process the
269 --match-limit and --recursion-limit options (added for 8.11). In
270 particular, this made one of the standard tests fail. (The integer value
271 went into the wrong half of a long int.)
273 3. If the --colour option was given to pcregrep with -v (invert match), it
274 did strange things, either producing crazy output, or crashing. It should,
275 of course, ignore a request for colour when reporting lines that do not
276 match.
278 4. Another pcregrep bug caused similar problems if --colour was specified with
279 -M (multiline) and the pattern match finished with a line ending.
281 5. In pcregrep, when a pattern that ended with a literal newline sequence was
282 matched in multiline mode, the following line was shown as part of the
283 match. This seems wrong, so I have changed it.
285 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
286 the check for further matches in the same line (so they could be coloured)
287 to overrun the end of the current line. If another match was found, it was
288 incorrectly shown (and then shown again when found in the next line).
290 7. If pcregrep was compiled under Windows, there was a reference to the
291 function pcregrep_exit() before it was defined. I am assuming this was
292 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
293 reported by a user. I've moved the definition above the reference.
296 Version 8.11 10-Dec-2010
297 ------------------------
299 1. (*THEN) was not working properly if there were untried alternatives prior
300 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
301 backtracked to try for "b" instead of moving to the next alternative branch
302 at the same level (in this case, to look for "c"). The Perl documentation
303 is clear that when (*THEN) is backtracked onto, it goes to the "next
304 alternative in the innermost enclosing group".
306 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
307 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
308 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
309 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
310 (*THEN).
312 3. If \s appeared in a character class, it removed the VT character from
313 the class, even if it had been included by some previous item, for example
314 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
315 of \s, but is part of the POSIX "space" class.)
317 4. A partial match never returns an empty string (because you can always
318 match an empty string at the end of the subject); however the checking for
319 an empty string was starting at the "start of match" point. This has been
320 changed to the "earliest inspected character" point, because the returned
321 data for a partial match starts at this character. This means that, for
322 example, /(?<=abc)def/ gives a partial match for the subject "abc"
323 (previously it gave "no match").
325 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
326 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
327 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
328 has an implication that the given string is incomplete (because a partial
329 match is preferred over a full match). For this reason, these items now
330 give a partial match in this situation. [Aside: previously, the one case
331 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
332 match rather than a full match, which was wrong by the old rules, but is
333 now correct.]
335 6. There was a bug in the handling of #-introduced comments, recognized when
336 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
337 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
338 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
339 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
340 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
341 places in pcre_compile().
343 7. Related to (6) above, when pcre_compile() was skipping #-introduced
344 comments when looking ahead for named forward references to subpatterns,
345 the only newline sequence it recognized was NL. It now handles newlines
346 according to the set newline convention.
348 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
349 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
350 cater for a lack of strerror(). These oversights have been fixed.
352 9. Added --match-limit and --recursion-limit to pcregrep.
354 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
356 11. When the -o option was used, pcregrep was setting a return code of 1, even
357 when matches were found, and --line-buffered was not being honoured.
359 12. Added an optional parentheses number to the -o and --only-matching options
360 of pcregrep.
362 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
363 can match an empty string. The code to do it in pcretest and pcredemo
364 needed fixing:
366 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
367 only one byte after an empty string match just before CRLF (this case
368 just got forgotten; "any" and "anycrlf" were OK).
370 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
371 mode when an empty string match preceded an ASCII character followed by
372 a non-ASCII character. (The code for advancing by one character rather
373 than one byte was nonsense.)
375 (c) The pcredemo.c sample program did not have any code at all to handle
376 the cases when CRLF is a valid newline sequence.
378 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
379 as a starting offset was within the subject string. There is now a new
380 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
381 negative or greater than the length of the string. In order to test this,
382 pcretest is extended to allow the setting of negative starting offsets.
384 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
385 starting offset points to the beginning of a UTF-8 character was
386 unnecessarily clumsy. I tidied it up.
388 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
389 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
391 17. Nobody had reported that the --include_dir option, which was added in
392 release 7.7 should have been called --include-dir (hyphen, not underscore)
393 for compatibility with GNU grep. I have changed it to --include-dir, but
394 left --include_dir as an undocumented synonym, and the same for
395 --exclude-dir, though that is not available in GNU grep, at least as of
396 release 2.5.4.
398 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
399 characters from a string of bytes) have been redefined so as not to use
400 loops, in order to improve performance in some environments. At the same
401 time, I abstracted some of the common code into auxiliary macros to save
402 repetition (this should not affect the compiled code).
404 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
405 compile-time error is now given if \c is not followed by an ASCII
406 character, that is, a byte less than 128. (In EBCDIC mode, the code is
407 different, and any byte value is allowed.)
409 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
410 START_OPTIMIZE option, which is now allowed at compile time - but just
411 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
412 to pcregrep and other applications that have no direct access to PCRE
413 options. The new /Y option in pcretest sets this option when calling
414 pcre_compile().
416 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
417 back references. Groups containing recursive back references were forced to
418 be atomic by that change, but in the case of named groups, the amount of
419 memory required was incorrectly computed, leading to "Failed: internal
420 error: code overflow". This has been fixed.
422 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
423 pcretest.c, to avoid build problems in some Borland environments.
426 Version 8.10 25-Jun-2010
427 ------------------------
429 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
430 THEN.
432 2. (*ACCEPT) was not working when inside an atomic group.
434 3. Inside a character class, \B is treated as a literal by default, but
435 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
436 causes the error). The code is unchanged, but I tidied the documentation.
438 4. Inside a character class, PCRE always treated \R and \X as literals,
439 whereas Perl faults them if its -w option is set. I have changed PCRE so
440 that it faults them when PCRE_EXTRA is set.
442 5. Added support for \N, which always matches any character other than
443 newline. (It is the same as "." when PCRE_DOTALL is not set.)
445 6. When compiling pcregrep with newer versions of gcc which may have
446 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
447 declared with attribute warn_unused_result" were given. Just casting the
448 result to (void) does not stop the warnings; a more elaborate fudge is
449 needed. I've used a macro to implement this.
451 7. Minor change to pcretest.c to avoid a compiler warning.
453 8. Added four artifical Unicode properties to help with an option to make
454 \s etc use properties (see next item). The new properties are: Xan
455 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
457 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
458 use Unicode properties. (*UCP) at the start of a pattern can be used to set
459 this option. Modified pcretest to add /W to test this facility. Added
460 REG_UCP to make it available via the POSIX interface.
462 10. Added --line-buffered to pcregrep.
464 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
465 studied, and the match started with a letter with a code point greater than
466 127 whose first byte was different to the first byte of the other case of
467 the letter, the other case of this starting letter was not recognized
468 (#976).
470 12. If a pattern that was studied started with a repeated Unicode property
471 test, for example, \p{Nd}+, there was the theoretical possibility of
472 setting up an incorrect bitmap of starting bytes, but fortunately it could
473 not have actually happened in practice until change 8 above was made (it
474 added property types that matched character-matching opcodes).
476 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
477 possible starting bytes for non-anchored patterns.
479 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
480 \R, and also a number of cases that involve Unicode properties, both
481 explicit and implicit when PCRE_UCP is set.
483 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
484 input, it could crash or give wrong results if characters with values
485 greater than 0xc0 were present in the subject string. (Detail: it assumed
486 UTF-8 input when processing these items.)
488 16. Added a lot of (int) casts to avoid compiler warnings in systems where
489 size_t is 64-bit (#991).
491 17. Added a check for running out of memory when PCRE is compiled with
492 --disable-stack-for-recursion (#990).
494 18. If the last data line in a file for pcretest does not have a newline on
495 the end, a newline was missing in the output.
497 19. The default pcre_chartables.c file recognizes only ASCII characters (values
498 less than 128) in its various bitmaps. However, there is a facility for
499 generating tables according to the current locale when PCRE is compiled. It
500 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
501 characters, are recognized by isspace() and therefore were getting set in
502 these tables, and indeed these tables seem to approximate to ISO 8859. This
503 caused a problem in UTF-8 mode when pcre_study() was used to create a list
504 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
505 which of course cannot start UTF-8 characters. I have changed the code so
506 that only real ASCII characters (less than 128) and the correct starting
507 bytes for UTF-8 encodings are set for characters greater than 127 when in
508 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
509 altogether.)
511 20. Added the /T option to pcretest so as to be able to run tests with non-
512 standard character tables, thus making it possible to include the tests
513 used for 19 above in the standard set of tests.
515 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
516 reference to a subpattern the other side of a comment that contains an
517 opening parenthesis caused either an internal compiling error, or a
518 reference to the wrong subpattern.
521 Version 8.02 19-Mar-2010
522 ------------------------
524 1. The Unicode data tables have been updated to Unicode 5.2.0.
526 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
527 configured.
529 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
530 original author of that file, following a query about its status.
532 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
533 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
535 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
536 quantifier applied to a forward-referencing subroutine call, could compile
537 incorrect code or give the error "internal error: previously-checked
538 referenced subpattern not found".
540 6. Both MS Visual Studio and Symbian OS have problems with initializing
541 variables to point to external functions. For these systems, therefore,
542 pcre_malloc etc. are now initialized to local functions that call the
543 relevant global functions.
545 7. There were two entries missing in the vectors called coptable and poptable
546 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
547 I've fixed the data, and added a kludgy way of testing at compile time that
548 the lengths are correct (equal to the number of opcodes).
550 8. Following on from 7, I added a similar kludge to check the length of the
551 eint vector in pcreposix.c.
553 9. Error texts for pcre_compile() are held as one long string to avoid too
554 much relocation at load time. To find a text, the string is searched,
555 counting zeros. There was no check for running off the end of the string,
556 which could happen if a new error number was added without updating the
557 string.
559 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
561 11. \K was not working if it appeared in an atomic group or in a group that
562 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
563 \K is "not well defined" if used in an assertion. PCRE now accepts it if
564 the assertion is positive, but not if it is negative.
566 12. Change 11 fortuitously reduced the size of the stack frame used in the
567 "match()" function of pcre_exec.c by one pointer. Forthcoming
568 implementation of support for (*MARK) will need an extra pointer on the
569 stack; I have reserved it now, so that the stack frame size does not
570 decrease.
572 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
573 item in branch that calls a recursion is a subroutine call - as in the
574 second branch in the above example - was incorrectly given the compile-
575 time error "recursive call could loop indefinitely" because pcre_compile()
576 was not correctly checking the subroutine for matching a non-empty string.
578 14. The checks for overrunning compiling workspace could trigger after an
579 overrun had occurred. This is a "should never occur" error, but it can be
580 triggered by pathological patterns such as hundreds of nested parentheses.
581 The checks now trigger 100 bytes before the end of the workspace.
583 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
586 Version 8.01 19-Jan-2010
587 ------------------------
589 1. If a pattern contained a conditional subpattern with only one branch (in
590 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
591 computed the wrong minimum data length (which is of course zero for such
592 subpatterns). This could cause incorrect "no match" results.
594 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
595 the pattern is reset in the first branch, pcre_compile() failed with
596 "internal error: code overflow at offset...". This happened only when
597 the reset was to the original external option setting. (An optimization
598 abstracts leading options settings into an external setting, which was the
599 cause of this.)
601 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
602 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
603 assertion pattern did not match (meaning that the assertion was true), it
604 was incorrectly treated as false if the SKIP had been reached during the
605 matching. This also applied to assertions used as conditions.
607 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
608 assertion subpattern, including such a pattern used as a condition,
609 unpredictable results occurred, instead of the error return
612 5. The C++ GlobalReplace function was not working like Perl for the special
613 situation when an empty string is matched. It now does the fancy magic
614 stuff that is necessary.
616 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
617 removed. (These were left over from very, very early versions of PCRE.)
619 7. Some cosmetic changes to the code to make life easier when compiling it
620 as part of something else:
622 (a) Change DEBUG to PCRE_DEBUG.
624 (b) In pcre_compile(), rename the member of the "branch_chain" structure
625 called "current" as "current_branch", to prevent a collision with the
626 Linux macro when compiled as a kernel module.
628 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
629 prevent a collision with the Linux macro when compiled as a kernel
630 module.
632 8. In pcre_compile() there are some checks for integer overflows that used to
633 cast potentially large values to (double). This has been changed to that
634 when building, a check for int64_t is made, and if it is found, it is used
635 instead, thus avoiding the use of floating point arithmetic. (There is no
636 other use of FP in PCRE.) If int64_t is not found, the fallback is to
637 double.
639 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
640 2005 (difference between two addresses compared to an unsigned value).
642 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
643 custom one, because of the following reported problem in Windows:
645 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
646 under Win32.
647 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
648 therefore missing the function definition.
649 - The compiler thus generates a "C" signature for the test function.
650 - The linker fails to find the "C" function.
651 - PCRE fails to configure if asked to do so against libbz2.
653 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
654 messages were output:
656 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
657 rerunning libtoolize, to keep the correct libtool macros in-tree.
658 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
660 I have done both of these things.
662 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
663 most of the time, it *can* run out if it is given a pattern that contains a
664 runaway infinite recursion. I updated the discussion in the pcrestack man
665 page.
667 13. Now that we have gone to the x.xx style of version numbers, the minor
668 version may start with zero. Using 08 or 09 is a bad idea because users
669 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
670 interpreted as invalid octal numbers. I've updated the previous comment in
671 configure.ac, and also added a check that gives an error if 08 or 09 are
672 used.
674 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
675 causing partial matching to fail when the end of the subject matched \W
676 in a UTF-8 pattern where \W was quantified with a minimum of 3.
678 15. There were some discrepancies between the declarations in pcre_internal.h
679 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
680 their definitions. The declarations used "const uschar *" and the
681 definitions used USPTR. Even though USPTR is normally defined as "const
682 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
683 reported that: "This difference in casting confuses some C++ compilers, for
684 example, SunCC recognizes above declarations as different functions and
685 generates broken code for hbpcre." I have changed the declarations to use
686 USPTR.
688 16. GNU libtool is named differently on some systems. The autogen.sh script now
689 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
690 (FreeBSD).
692 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
693 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
694 comment: "Figure out how to create a longlong from a string: strtoll and
695 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
696 instance, but it only takes 2 args instead of 3!"
698 18. A subtle bug concerned with back references has been fixed by a change of
699 specification, with a corresponding code fix. A pattern such as
700 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
701 refers, was giving matches when it shouldn't. For example, xa=xaaa would
702 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
703 same bug. Such groups have to be quantified to be useful, or contained
704 inside another quantified group. (If there's no repetition, the reference
705 can never match.) The problem arises because, having left the group and
706 moved on to the rest of the pattern, a later failure that backtracks into
707 the group uses the captured value from the final iteration of the group
708 rather than the correct earlier one. I have fixed this in PCRE by forcing
709 any group that contains a reference to itself to be an atomic group; that
710 is, there cannot be any backtracking into it once it has completed. This is
711 similar to recursive and subroutine calls.
714 Version 8.00 19-Oct-09
715 ----------------------
717 1. The table for translating pcre_compile() error codes into POSIX error codes
718 was out-of-date, and there was no check on the pcre_compile() error code
719 being within the table. This could lead to an OK return being given in
720 error.
722 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
723 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
724 in a Windows environment.
726 3. The pcregrep --count option prints the count for each file even when it is
727 zero, as does GNU grep. However, pcregrep was also printing all files when
728 --files-with-matches was added. Now, when both options are given, it prints
729 counts only for those files that have at least one match. (GNU grep just
730 prints the file name in this circumstance, but including the count seems
731 more useful - otherwise, why use --count?) Also ensured that the
732 combination -clh just lists non-zero counts, with no names.
734 4. The long form of the pcregrep -F option was incorrectly implemented as
735 --fixed_strings instead of --fixed-strings. This is an incompatible change,
736 but it seems right to fix it, and I didn't think it was worth preserving
737 the old behaviour.
739 5. The command line items --regex=pattern and --regexp=pattern were not
740 recognized by pcregrep, which required --regex pattern or --regexp pattern
741 (with a space rather than an '='). The man page documented the '=' forms,
742 which are compatible with GNU grep; these now work.
744 6. No libpcreposix.pc file was created for pkg-config; there was just
745 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
747 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
748 when UCP support is not needed, by modifying the Python script that
749 generates it from Unicode data files. This should not matter if the module
750 is correctly used as a library, but I received one complaint about 50K of
751 unwanted data. My guess is that the person linked everything into his
752 program rather than using a library. Anyway, it does no harm.
754 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
755 was a minimum greater than 1 for a wide character in a possessive
756 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
757 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
758 character. Chaos in the form of incorrect output or a compiling loop could
759 result.
761 9. The restrictions on what a pattern can contain when partial matching is
762 requested for pcre_exec() have been removed. All patterns can now be
763 partially matched by this function. In addition, if there are at least two
764 slots in the offset vector, the offset of the earliest inspected character
765 for the match and the offset of the end of the subject are set in them when
766 PCRE_ERROR_PARTIAL is returned.
768 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
769 synonymous with PCRE_PARTIAL, for backwards compatibility, and
770 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
771 and may be more useful for multi-segment matching.
773 11. Partial matching with pcre_exec() is now more intuitive. A partial match
774 used to be given if ever the end of the subject was reached; now it is
775 given only if matching could not proceed because another character was
776 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
777 string "Z", which now yields "no match" instead of "partial match". In the
778 case of pcre_dfa_exec(), "no match" is given if every matching path for the
779 final character ended with (*FAIL).
781 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
782 if the pattern had a "must contain" character that was already found in the
783 earlier partial match, unless partial matching was again requested. For
784 example, with the pattern /dog.(body)?/, the "must contain" character is
785 "g". If the first part-match was for the string "dog", restarting with
786 "sbody" failed. This bug has been fixed.
788 13. The string returned by pcre_dfa_exec() after a partial match has been
789 changed so that it starts at the first inspected character rather than the
790 first character of the match. This makes a difference only if the pattern
791 starts with a lookbehind assertion or \b or \B (\K is not supported by
792 pcre_dfa_exec()). It's an incompatible change, but it makes the two
793 matching functions compatible, and I think it's the right thing to do.
795 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
796 so that the demonstration program is easily available in environments where
797 PCRE has not been installed from source.
799 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
800 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
801 library.
803 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
804 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
805 is not the first non-POSIX option to be added. Clearly some people find
806 these options useful.
808 17. If a caller to the POSIX matching function regexec() passes a non-zero
809 value for nmatch with a NULL value for pmatch, the value of
810 nmatch is forced to zero.
812 18. RunGrepTest did not have a test for the availability of the -u option of
813 the diff command, as RunTest does. It now checks in the same way as
814 RunTest, and also checks for the -b option.
816 19. If an odd number of negated classes containing just a single character
817 interposed, within parentheses, between a forward reference to a named
818 subpattern and the definition of the subpattern, compilation crashed with
819 an internal error, complaining that it could not find the referenced
820 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
821 [The bug was that it was starting one character too far in when skipping
822 over the character class, thus treating the ] as data rather than
823 terminating the class. This meant it could skip too much.]
825 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
826 /g option in pcretest when the pattern contains \K, which makes it possible
827 to have an empty string match not at the start, even when the pattern is
828 anchored. Updated pcretest and pcredemo to use this option.
830 21. If the maximum number of capturing subpatterns in a recursion was greater
831 than the maximum at the outer level, the higher number was returned, but
832 with unset values at the outer level. The correct (outer level) value is
833 now given.
835 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
836 PCRE did not set those parentheses (unlike Perl). I have now found a way to
837 make it do so. The string so far is captured, making this feature
838 compatible with Perl.
840 23. The tests have been re-organized, adding tests 11 and 12, to make it
841 possible to check the Perl 5.10 features against Perl 5.10.
843 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
844 pattern matches a fixed length string. PCRE did not allow this; now it
845 does. Neither allows recursion.
847 25. I finally figured out how to implement a request to provide the minimum
848 length of subject string that was needed in order to match a given pattern.
849 (It was back references and recursion that I had previously got hung up
850 on.) This code has now been added to pcre_study(); it finds a lower bound
851 to the length of subject needed. It is not necessarily the greatest lower
852 bound, but using it to avoid searching strings that are too short does give
853 some useful speed-ups. The value is available to calling programs via
854 pcre_fullinfo().
856 26. While implementing 25, I discovered to my embarrassment that pcretest had
857 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
858 study optimizations had never been tested with that matching function.
859 Oops. What is worse, even when it was passed study data, there was a bug in
860 pcre_dfa_exec() that meant it never actually used it. Double oops. There
861 were also very few tests of studied patterns with pcre_dfa_exec().
863 27. If (?| is used to create subpatterns with duplicate numbers, they are now
864 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
865 on the other side of the coin, they are no longer allowed to have different
866 names, because these cannot be distinguished in PCRE, and this has caused
867 confusion. (This is a difference from Perl.)
869 28. When duplicate subpattern names are present (necessarily with different
870 numbers, as required by 27 above), and a test is made by name in a
871 conditional pattern, either for a subpattern having been matched, or for
872 recursion in such a pattern, all the associated numbered subpatterns are
873 tested, and the overall condition is true if the condition is true for any
874 one of them. This is the way Perl works, and is also more like the way
875 testing by number works.
878 Version 7.9 11-Apr-09
879 ---------------------
881 1. When building with support for bzlib/zlib (pcregrep) and/or readline
882 (pcretest), all targets were linked against these libraries. This included
883 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
884 libraries. This caused unwanted dependencies to be created. This problem
885 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
886 pcretest is linked with readline.
888 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
889 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
890 moved outside it again, because FALSE and TRUE are already defined in AIX,
891 but BOOL is not.
893 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
894 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
896 4. The pcregrep documentation said spaces were inserted as well as colons (or
897 hyphens) following file names and line numbers when outputting matching
898 lines. This is not true; no spaces are inserted. I have also clarified the
899 wording for the --colour (or --color) option.
901 5. In pcregrep, when --colour was used with -o, the list of matching strings
902 was not coloured; this is different to GNU grep, so I have changed it to be
903 the same.
905 6. When --colo(u)r was used in pcregrep, only the first matching substring in
906 each matching line was coloured. Now it goes on to look for further matches
907 of any of the test patterns, which is the same behaviour as GNU grep.
909 7. A pattern that could match an empty string could cause pcregrep to loop; it
910 doesn't make sense to accept an empty string match in pcregrep, so I have
911 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
912 seems to be how GNU grep behaves.
914 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
915 start or after a newline", because the conditional assertion was not being
916 correctly handled. The rule now is that both the assertion and what follows
917 in the first alternative must satisfy the test.
919 9. If auto-callout was enabled in a pattern with a conditional group whose
920 condition was an assertion, PCRE could crash during matching, both with
921 pcre_exec() and pcre_dfa_exec().
923 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
924 used for matching.
926 11. Unicode property support in character classes was not working for
927 characters (bytes) greater than 127 when not in UTF-8 mode.
929 12. Added the -M command line option to pcretest.
931 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
933 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
935 16. Added comments and documentation about mis-use of no_arg in the C++
936 wrapper.
938 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
939 from Martin Jerabek that uses macro names for all relevant character and
940 string constants.
942 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
943 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
944 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
945 these, but not everybody uses configure.
947 19. A conditional group that had only one branch was not being correctly
948 recognized as an item that could match an empty string. This meant that an
949 enclosing group might also not be so recognized, causing infinite looping
950 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
951 with the subject "ab", where knowledge that the repeated group can match
952 nothing is needed in order to break the loop.
954 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
955 exec(), but without supplying a callout function, matching went wrong.
957 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
958 leak if the size of the offset vector was greater than 30. When the vector
959 is smaller, the saved offsets during recursion go onto a local stack
960 vector, but for larger vectors malloc() is used. It was failing to free
961 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
962 error, in fact).
964 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
965 heapframe that is used only when UTF-8 support is enabled. This caused no
966 problem, but was untidy.
968 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
969 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
970 included within another project.
972 24. Steven Van Ingelgem's patches to add more options to the CMake support,
973 slightly modified by me:
975 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
976 not building pcregrep.
978 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
979 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
981 25. Forward references, both numeric and by name, in patterns that made use of
982 duplicate group numbers, could behave incorrectly or give incorrect errors,
983 because when scanning forward to find the reference group, PCRE was not
984 taking into account the duplicate group numbers. A pattern such as
985 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
987 26. Changed a few more instances of "const unsigned char *" to USPTR, making
988 the feature of a custom pointer more persuasive (as requested by a user).
990 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
991 pcretest.c, inside #ifndefs, because it seems they are sometimes already
992 pre-defined.
994 28. Added support for (*UTF8) at the start of a pattern.
996 29. Arrange for flags added by the "release type" setting in CMake to be shown
997 in the configuration summary.
1000 Version 7.8 05-Sep-08
1001 ---------------------
1003 1. Replaced UCP searching code with optimized version as implemented for Ad
1004 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1005 stage table and inline lookup instead of a function, giving speed ups of 2
1006 to 5 times on some simple patterns that I tested. Permission was given to
1007 distribute the MultiStage2.py script that generates the tables (it's not in
1008 the tarball, but is in the Subversion repository).
1010 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1011 scripts.
1013 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1014 a group with a zero qualifier. The result of the study could be incorrect,
1015 or the function might crash, depending on the pattern.
1017 4. Caseless matching was not working for non-ASCII characters in back
1018 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1019 It now works when Unicode Property Support is available.
1021 5. In pcretest, an escape such as \x{de} in the data was always generating
1022 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1023 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1024 truncation.
1026 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1028 7. Added two (int) casts to pcregrep when printing the difference of two
1029 pointers, in case they are 64-bit values.
1031 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1032 test 2 if it fails.
1034 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1035 and a #define of that name to empty if it is not externally set. This is to
1036 allow users of MSVC to set it if necessary.
1038 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1039 the convenience functions in the pcre_get.c source file.
1041 11. An option change at the start of a pattern that had top-level alternatives
1042 could cause overwriting and/or a crash. This command provoked a crash in
1043 some environments:
1045 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1047 This potential security problem was recorded as CVE-2008-2371.
1049 12. For a pattern where the match had to start at the beginning or immediately
1050 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1051 pcre_dfa_exec() could read past the end of the passed subject if there was
1052 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1053 pcretest so that it places the subject at the end of its malloc-ed buffer.
1055 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1056 exec() might read past the end of the data buffer in UTF-8 mode.
1058 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1059 the data contained the byte 0x85 as part of a UTF-8 character within its
1060 first line. This applied both to normal and DFA matching.
1062 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1063 /^[^d]*?$/8 failed to match "abc".
1065 16. Added a missing copyright notice to pcrecpp_internal.h.
1067 17. Make it more clear in the documentation that values returned from
1068 pcre_exec() in ovector are byte offsets, not character counts.
1070 18. Tidied a few places to stop certain compilers from issuing warnings.
1072 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1073 supplied by Stefan Weber. I made a further small update for 7.8 because
1074 there is a change of source arrangements: the pcre_searchfuncs.c module is
1075 replaced by pcre_ucd.c.
1078 Version 7.7 07-May-08
1079 ---------------------
1081 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1082 a string to a long long, pretend we don't even have a long long." This is
1083 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1085 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1086 pre-7.6 versions, which defined a global no_arg variable instead of putting
1087 it in the RE class. (See also #8 below.)
1089 3. Remove a line of dead code, identified by coverity and reported by Nuno
1090 Lopes.
1092 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1094 (1) The include/exclude patterns were being applied to the whole pathnames
1095 of files, instead of just to the final components.
1097 (2) If there was more than one level of directory, the subdirectories were
1098 skipped unless they satisfied the include/exclude conditions. This is
1099 inconsistent with GNU grep (and could even be seen as contrary to the
1100 pcregrep specification - which I improved to make it absolutely clear).
1101 The action now is always to scan all levels of directory, and just
1102 apply the include/exclude patterns to regular files.
1104 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1105 --exclude_dir in the tests to avoid scanning .svn directories.
1107 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1108 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1109 doesn't support NULs in patterns.
1111 7. Added some missing "const"s to declarations of static tables in
1112 pcre_compile.c and pcre_dfa_exec.c.
1114 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1115 caused by fix #2 above. (Subsequently also a second patch to fix the
1116 first patch. And a third patch - this was a messy problem.)
1118 9. Applied Craig's patch to remove the use of push_back().
1120 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1121 matching function regexec().
1123 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1124 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1125 references. PCRE supports relative numbers with this syntax (I don't think
1126 Oniguruma does).
1128 12. Previously, a group with a zero repeat such as (...){0} was completely
1129 omitted from the compiled regex. However, this means that if the group
1130 was called as a subroutine from elsewhere in the pattern, things went wrong
1131 (an internal error was given). Such groups are now left in the compiled
1132 pattern, with a new opcode that causes them to be skipped at execution
1133 time.
1135 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1136 to the way PCRE behaves:
1138 (a) A lone ] character is dis-allowed (Perl treats it as data).
1140 (b) A back reference to an unmatched subpattern matches an empty string
1141 (Perl fails the current match path).
1143 (c) A data ] in a character class must be notated as \] because if the
1144 first data character in a class is ], it defines an empty class. (In
1145 Perl it is not possible to have an empty class.) The empty class []
1146 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1147 The negative empty class [^] matches any one character, independently
1148 of the DOTALL setting.
1150 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1151 non-existent subpattern following a character class starting with ']' and
1152 containing () gave an internal compiling error instead of "reference to
1153 non-existent subpattern". Fortunately, when the pattern did exist, the
1154 compiled code was correct. (When scanning forwards to check for the
1155 existencd of the subpattern, it was treating the data ']' as terminating
1156 the class, so got the count wrong. When actually compiling, the reference
1157 was subsequently set up correctly.)
1159 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1160 it was being rejected as not supported by pcre_dfa_exec(), even though
1161 other assertions are supported. I have made pcre_dfa_exec() support
1162 (*FAIL).
1164 16. The implementation of 13c above involved the invention of a new opcode,
1165 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1166 cannot be changed at match time, I realized I could make a small
1167 improvement to matching performance by compiling OP_ALLANY instead of
1168 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1169 on the OP_ANY path.
1171 17. Compiling pcretest on Windows with readline support failed without the
1172 following two fixes: (1) Make the unistd.h include conditional on
1173 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1175 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1176 ncurses library to be included for pcretest when ReadLine support is
1177 requested, but also to allow for it to be overridden. This patch came from
1178 Daniel Bergström.
1180 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1181 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1182 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1183 spotting this.
1186 Version 7.6 28-Jan-08
1187 ---------------------
1189 1. A character class containing a very large number of characters with
1190 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1191 overflow.
1193 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1194 HAVE_LONG_LONG is not defined.
1196 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1197 bring them up to date and include new features. This patch includes:
1199 - Fixed PH's badly added libz and libbz2 support.
1200 - Fixed a problem with static linking.
1201 - Added pcredemo. [But later removed - see 7 below.]
1202 - Fixed dftables problem and added an option.
1203 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1205 - Added readline support for pcretest.
1206 - Added an listing of the option settings after cmake has run.
1208 4. A user submitted a patch to Makefile that makes it easy to create
1209 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1210 Makefile.am that cause it to include this special target, without
1211 affecting anything else. Note that the same mingw target plus all
1212 the other distribution libraries and programs are now supported
1213 when configuring with CMake (see 6 below) instead of with
1214 Configure/Make.
1216 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1217 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1218 exported in the Windows port". It has not yet been confirmed that the patch
1219 solves the problem, but it does no harm.
1221 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1222 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1223 with CMake, and also correct the comment about stack recursion.
1225 7. Remove the automatic building of pcredemo from the ./configure system and
1226 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1227 of a program that users should build themselves after PCRE is installed, so
1228 building it automatically is not really right. What is more, it gave
1229 trouble in some build environments.
1231 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1234 Version 7.5 10-Jan-08
1235 ---------------------
1237 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1238 values in parens when parsing an RE using the C++ wrapper."
1240 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1241 Characters greater than 255 were excluded from the class instead of being
1242 included.
1244 3. The same bug as (2) above applied to negated POSIX classes such as
1245 [:^space:].
1247 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1248 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1249 I have changed it.
1251 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1252 first named subpattern) and a construct such as (?&a) would reference the
1253 first named subpattern whose name started with "a" (in other words, the
1254 length check was missing). Both these problems are fixed. "Subpattern name
1255 expected" is now given for (?&) (a zero-length name), and this patch also
1256 makes it give the same error for \k'' (previously it complained that that
1257 was a reference to a non-existent subpattern).
1259 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1260 this is right because (?- can be followed by option settings as well as by
1261 digits. I have, however, made the messages clearer.
1263 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1264 than the number used in the conditional) now cause a compile-time error.
1265 This is actually not compatible with Perl, which accepts such patterns, but
1266 treats the conditional as always being FALSE (as PCRE used to), but it
1267 seems to me that giving a diagnostic is better.
1269 8. Change "alphameric" to the more common word "alphanumeric" in comments
1270 and messages.
1272 9. Fix two occurrences of "backslash" in comments that should have been
1273 "backspace".
1275 10. Remove two redundant lines of code that can never be obeyed (their function
1276 was moved elsewhere).
1278 11. The program that makes PCRE's Unicode character property table had a bug
1279 which caused it to generate incorrect table entries for sequences of
1280 characters that have the same character type, but are in different scripts.
1281 It amalgamated them into a single range, with the script of the first of
1282 them. In other words, some characters were in the wrong script. There were
1283 thirteen such cases, affecting characters in the following ranges:
1285 U+002b0 - U+002c1
1286 U+0060c - U+0060d
1287 U+0061e - U+00612
1288 U+0064b - U+0065e
1289 U+0074d - U+0076d
1290 U+01800 - U+01805
1291 U+01d00 - U+01d77
1292 U+01d9b - U+01dbf
1293 U+0200b - U+0200f
1294 U+030fc - U+030fe
1295 U+03260 - U+0327f
1296 U+0fb46 - U+0fbb1
1297 U+10450 - U+1049d
1299 12. The -o option (show only the matching part of a line) for pcregrep was not
1300 compatible with GNU grep in that, if there was more than one match in a
1301 line, it showed only the first of them. It now behaves in the same way as
1302 GNU grep.
1304 13. If the -o and -v options were combined for pcregrep, it printed a blank
1305 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1306 does the same. The return code can be used to tell if there were any
1307 non-matching lines.
1309 14. Added --file-offsets and --line-offsets to pcregrep.
1311 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1312 infinitely looping recursion. The bug was that positive lookaheads were not
1313 being skipped when checking for a possible empty match (negative lookaheads
1314 and both kinds of lookbehind were skipped).
1316 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1317 inclusion of <windows.h> to before rather than after the definition of
1318 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1320 17. Specifying a possessive quantifier with a specific limit for a Unicode
1321 character property caused pcre_compile() to compile bad code, which led at
1322 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1323 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1324 caused the error; without that there was no problem.
1326 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1328 19. Added --enable-pcretest-libreadline.
1330 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1331 RE::GlobalReplace(). As a result, the number of replacements returned was
1332 double what it should be. I removed one of the increments, but Craig sent a
1333 later patch that removed the other one (the right fix) and added unit tests
1334 that check the return values (which was not done before).
1336 21. Several CMake things:
1338 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1339 the names libpcre and libpcreposix, not just pcre and pcreposix.
1341 (2) The above change means that pcretest and pcregrep are now correctly
1342 linked with the newly-built libraries, not previously installed ones.
1346 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1347 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1348 UTF-8 newline character). The key issue is that the pattern starts .*;
1349 this means that the match must be either at the beginning, or after a
1350 newline. The bug was in the code for advancing after a failed match and
1351 checking that the new position followed a newline. It was not taking
1352 account of UTF-8 characters correctly.
1354 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1355 character classes. PCRE was not treating the sequence [:...:] as a
1356 character class unless the ... were all letters. Perl, however, seems to
1357 allow any characters between [: and :], though of course it rejects as
1358 unknown any "names" that contain non-letters, because all the known class
1359 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1360 for example, whereas PCRE did not - it did not recognize a POSIX character
1361 class. This seemed a bit dangerous, so the code has been changed to be
1362 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1363 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1364 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1365 Perl does, and where it didn't before.
1367 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1368 Windows environments %n is disabled by default.
1371 Version 7.4 21-Sep-07
1372 ---------------------
1374 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1375 means that a class such as [\s] counted as "explicit reference to CR or
1376 LF". That isn't really right - the whole point of the change was to try to
1377 help when there was an actual mention of one of the two characters. So now
1378 the change happens only if \r or \n (or a literal CR or LF) character is
1379 encountered.
1381 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1382 of both had grown to the point where there were only 3 bits left.
1383 Fortunately, there was spare space in the data structure, and so I have
1384 moved the internal flags into a new 16-bit field to free up more option
1385 bits.
1387 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1388 but did not set the internal JCHANGED flag - either of these is enough to
1389 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1390 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1391 start sets both bits.
1393 4. Added options (at build time, compile time, exec time) to change \R from
1394 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1396 5. doc/pcresyntax.html was missing from the distribution.
1398 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1399 compatibility, even though it is no longer used.
1401 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1402 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1403 windows.h file is present (where different names are used). [This was
1404 reversed later after testing - see 16 below.]
1406 8. Changed all #include <config.h> to #include "config.h". There were also
1407 some further <pcre.h> cases that I changed to "pcre.h".
1409 9. When pcregrep was used with the --colour option, it missed the line ending
1410 sequence off the lines that it output.
1412 10. It was pointed out to me that arrays of string pointers cause lots of
1413 relocations when a shared library is dynamically loaded. A technique of
1414 using a single long string with a table of offsets can drastically reduce
1415 these. I have refactored PCRE in four places to do this. The result is
1416 dramatic:
1418 Originally: 290
1419 After changing UCP table: 187
1420 After changing error message table: 43
1421 After changing table of "verbs" 36
1422 After changing table of Posix names 22
1424 Thanks to the folks working on Gregex for glib for this insight.
1426 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1427 unicode-properties was also set.
1429 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1431 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1432 checked only for CRLF.
1434 14. Added casts to pcretest.c to avoid compiler warnings.
1436 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1438 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1439 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1440 entirely. This removes changes made in 7 above.
1442 17. The CMake files have been updated, and there is now more information about
1443 building with CMake in the NON-UNIX-USE document.
1446 Version 7.3 28-Aug-07
1447 ---------------------
1449 1. In the rejigging of the build system that eventually resulted in 7.1, the
1450 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1451 brackets there is not right, since it causes compilers to look for an
1452 installed pcre.h, not the version that is in the source that is being
1453 compiled (which of course may be different). I have changed it back to:
1455 #include "pcre.h"
1457 I have a vague recollection that the change was concerned with compiling in
1458 different directories, but in the new build system, that is taken care of
1459 by the VPATH setting the Makefile.
1461 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1462 when the subject happened to end in the byte 0x85 (e.g. if the last
1463 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1464 characters but of course it shouldn't be taken as a newline when it is part
1465 of another character. The bug was that, for an unlimited repeat of . in
1466 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1467 characters when looking for a newline.
1469 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1471 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1472 in debug output.
1474 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1475 long printing in the pcrecpp unittest when running under MinGW.
1477 6. ESC_K was left out of the EBCDIC table.
1479 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1480 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1481 limit also applies to "virtual nesting" when a pattern is recursive, and in
1482 this case 1000 isn't so big. I have been able to remove this limit at the
1483 expense of backing off one optimization in certain circumstances. Normally,
1484 when pcre_exec() would call its internal match() function recursively and
1485 immediately return the result unconditionally, it uses a "tail recursion"
1486 feature to save stack. However, when a subpattern that can match an empty
1487 string has an unlimited repetition quantifier, it no longer makes this
1488 optimization. That gives it a stack frame in which to save the data for
1489 checking that an empty string has been matched. Previously this was taken
1490 from the 1000-entry workspace that had been reserved. So now there is no
1491 explicit limit, but more stack is used.
1493 8. Applied Daniel's patches to solve problems with the import/export magic
1494 syntax that is required for Windows, and which was going wrong for the
1495 pcreposix and pcrecpp parts of the library. These were overlooked when this
1496 problem was solved for the main library.
1498 9. There were some crude static tests to avoid integer overflow when computing
1499 the size of patterns that contain repeated groups with explicit upper
1500 limits. As the maximum quantifier is 65535, the maximum group length was
1501 set at 30,000 so that the product of these two numbers did not overflow a
1502 32-bit integer. However, it turns out that people want to use groups that
1503 are longer than 30,000 bytes (though not repeat them that many times).
1504 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1505 made it possible to implement the integer overflow checks in a much more
1506 dynamic way, which I have now done. The artificial limitation on group
1507 length has been removed - we now have only the limit on the total length of
1508 the compiled pattern, which depends on the LINK_SIZE setting.
1510 10. Fixed a bug in the documentation for get/copy named substring when
1511 duplicate names are permitted. If none of the named substrings are set, the
1512 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1513 empty string.
1515 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1516 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1517 because the ] is interpreted as the first data character and the
1518 terminating ] is not found. PCRE has been made compatible with Perl in this
1519 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1520 cause memory overwriting.
1522 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1523 string has been matched (to stop an infinite loop). It was not recognizing
1524 a conditional subpattern that could match an empty string if that
1525 subpattern was within another subpattern. For example, it looped when
1526 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1527 condition was not nested. This bug has been fixed.
1529 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1530 past the start of the subject in the presence of bytes with the top bit
1531 set, for example "\x8aBCD".
1533 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1534 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1536 14. Optimized (?!) to (*FAIL).
1538 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1539 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1540 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1541 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1542 does: it's just the validity check that is more restrictive.
1544 16. Inserted checks for integer overflows during escape sequence (backslash)
1545 processing, and also fixed erroneous offset values for syntax errors during
1546 backslash processing.
1548 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1549 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1551 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1552 caused an overrun.
1554 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1555 something other than just ASCII characters) inside a group that had an
1556 unlimited repeat caused a loop at compile time (while checking to see
1557 whether the group could match an empty string).
1559 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1560 [\P{Any}] did so. (Error in the code for printing property names.)
1562 21. An orphan \E inside a character class could cause a crash.
1564 22. A repeated capturing bracket such as (A)? could cause a wild memory
1565 reference during compilation.
1567 23. There are several functions in pcre_compile() that scan along a compiled
1568 expression for various reasons (e.g. to see if it's fixed length for look
1569 behind). There were bugs in these functions when a repeated \p or \P was
1570 present in the pattern. These operators have additional parameters compared
1571 with \d, etc, and these were not being taken into account when moving along
1572 the compiled data. Specifically:
1574 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1575 length.
1577 (b) An item such as \pL+ within a repeated group could cause crashes or
1578 loops.
1580 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1581 "reference to non-existent subpattern" error.
1583 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1585 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1586 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1588 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1590 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1591 character were causing crashes (broken optimization).
1593 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1594 \p or \P) caused a compile-time loop.
1596 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1597 break. For example, the unstudied pattern [\r\n]A does not match the string
1598 "\r\nA" because change 7.0/46 below moves the current point on by two
1599 characters after failing to match at the start. However, the pattern \nA
1600 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1601 the same is true. There doesn't seem any very clean way out of this, but
1602 what I have chosen to do makes the common cases work: PCRE now takes note
1603 of whether there can be an explicit match for \r or \n anywhere in the
1604 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1605 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1606 pattern has explicit CR or LF references.
1608 29. Added (*CR) etc for changing newline setting at start of pattern.
1611 Version 7.2 19-Jun-07
1612 ---------------------
1614 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1615 which is apparently normally available under Windows.
1617 2. Re-jig the pcregrep tests with different newline settings in an attempt
1618 to make them independent of the local environment's newline setting.
1620 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1622 4. Some of the "internals" tests were previously cut out when the link size
1623 was not 2, because the output contained actual offsets. The recent new
1624 "Z" feature of pcretest means that these can be cut out, making the tests
1625 usable with all link sizes.
1627 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1628 stack recursion. This gives a massive performance boost under BSD, but just
1629 a small improvement under Linux. However, it saves one field in the frame
1630 in all cases.
1632 6. Added more features from the forthcoming Perl 5.10:
1634 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1635 recursion call. It refers to the nth most recently opened parentheses.
1637 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1638 to be opened parentheses.
1640 (c) Conditions that refer to capturing parentheses can be specified
1641 relatively, for example, (?(-2)... or (?(+3)...
1643 (d) \K resets the start of the current match so that everything before
1644 is not part of it.
1646 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1648 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1649 reference syntax.
1651 (g) (?| introduces a group in which the numbering of parentheses in each
1652 alternative starts with the same number.
1654 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1656 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1659 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1660 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1661 for detecting groups that can match an empty string.
1663 9. A pattern with a very large number of alternatives (more than several
1664 hundred) was running out of internal workspace during the pre-compile
1665 phase, where pcre_compile() figures out how much memory will be needed. A
1666 bit of new cunning has reduced the workspace needed for groups with
1667 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1668 workspace instead of running out of the 4096 that are available.
1670 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1672 11. Applied patch from Google to remove an optimization that didn't quite work.
1673 The report of the bug said:
1675 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1676 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1677 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1679 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1680 it matched the wrong number of bytes.
1683 Version 7.1 24-Apr-07
1684 ---------------------
1686 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1687 that is more "standard", making use of automake and other Autotools. There
1688 is some re-arrangement of the files and adjustment of comments consequent
1689 on this.
1691 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1692 for recursive directory scanning broke on some systems because the files
1693 are not scanned in any specific order and on different systems the order
1694 was different. A call to "sort" has been inserted into RunGrepTest for the
1695 approprate test as a short-term fix. In the longer term there may be an
1696 alternative.
1698 3. I had an email from Eric Raymond about problems translating some of PCRE's
1699 man pages to HTML (despite the fact that I distribute HTML pages, some
1700 people do their own conversions for various reasons). The problems
1701 concerned the use of low-level troff macros .br and .in. I have therefore
1702 removed all such uses from the man pages (some were redundant, some could
1703 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1704 HTML has been updated to handle .nf/.fi and to complain if it encounters
1705 .br or .in.
1707 4. Updated comments in configure.ac that get placed in config.h.in and also
1708 arranged for config.h to be included in the distribution, with the name
1709 config.h.generic, for the benefit of those who have to compile without
1710 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1712 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1713 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1714 makevp.bat for the current PCRE, using the additional files
1715 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1717 6. A Windows user reported a minor discrepancy with test 2, which turned out
1718 to be caused by a trailing space on an input line that had got lost in his
1719 copy. The trailing space was an accident, so I've just removed it.
1721 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1722 that is needed.
1724 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1725 as "const" (a) because they are and (b) because it helps the PHP
1726 maintainers who have recently made a script to detect big data structures
1727 in the php code that should be moved to the .rodata section. I remembered
1728 to update Builducptable as well, so it won't revert if ucptable.h is ever
1729 re-created.
1731 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1732 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1733 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1734 support is not required. This saves 1.5-2K of code, which is important in
1735 some applications.
1737 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1738 so as not to refer to the tables, even though these functions will never be
1739 called when UTF-8 support is disabled. Otherwise there are problems with a
1740 shared library.
1742 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1744 (a) It was defining its arguments as char * instead of void *.
1746 (b) It was assuming that all moves were upwards in memory; this was true
1747 a long time ago when I wrote it, but is no longer the case.
1749 The emulated memove() is provided for those environments that have neither
1750 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1751 is clearly not the case, as these two bugs were recently reported.
1753 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1754 and Detrail to create the HTML documentation, the .txt form of the man
1755 pages, and it removes trailing spaces from listed files. It also creates
1756 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1757 case, it wraps all the #defines with #ifndefs. This script should be run
1758 before "make dist".
1760 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1761 with Unicode property support.
1763 (a) For a maximizing quantifier, if the two different cases of the
1764 character were of different lengths in their UTF-8 codings (there are
1765 some cases like this - I found 11), and the matching function had to
1766 back up over a mixture of the two cases, it incorrectly assumed they
1767 were both the same length.
1769 (b) When PCRE was configured to use the heap rather than the stack for
1770 recursion during matching, it was not correctly preserving the data for
1771 the other case of a UTF-8 character when checking ahead for a match
1772 while processing a minimizing repeat. If the check also involved
1773 matching a wide character, but failed, corruption could cause an
1774 erroneous result when trying to check for a repeat of the original
1775 character.
1777 13. Some tidying changes to the testing mechanism:
1779 (a) The RunTest script now detects the internal link size and whether there
1780 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1781 values substituted by "configure". (The RunGrepTest script already did
1782 this for UTF-8.) The configure.ac script no longer substitutes the
1783 relevant variables.
1785 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1786 with length and offset values. This means that the output is different
1787 for different internal link sizes. Test 2 is skipped for link sizes
1788 other than 2 because of this, bypassing the problem. Unfortunately,
1789 there was also a test in test 3 (the locale tests) that used /B and
1790 failed for link sizes other than 2. Rather than cut the whole test out,
1791 I have added a new /Z option to pcretest that replaces the length and
1792 offset values with spaces. This is now used to make test 3 independent
1793 of link size. (Test 2 will be tidied up later.)
1795 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1796 segmentation fault instead of returning the appropriate error message.
1798 15. In multiline mode when the newline sequence was set to "any", the pattern
1799 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1800 This doesn't seem right; it now treats the CRLF combination as the line
1801 ending, and so does not match in that case. It's only a pattern such as ^$
1802 that would hit this one: something like ^ABC$ would have failed after \r
1803 and then tried again after \r\n.
1805 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1806 in an attempt to make files that differ only in their line terminators
1807 compare equal. This works on Linux.
1809 17. Under certain error circumstances pcregrep might try to free random memory
1810 as it exited. This is now fixed, thanks to valgrind.
1812 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1813 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1814 was because its rules for how to advance for /g after matching an empty
1815 string at the end of a line did not allow for this case. They now check for
1816 it specially.
1818 20. pcretest is supposed to handle patterns and data of any length, by
1819 extending its buffers when necessary. It was getting this wrong when the
1820 buffer for a data line had to be extended.
1822 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1823 CRLF as a newline sequence.
1825 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1826 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1827 I have nevertheless tidied it up.
1829 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1831 24. Added a man page for pcre-config.
1834 Version 7.0 19-Dec-06
1835 ---------------------
1837 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1838 moving to gcc 4.1.1.
1840 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1841 sys/time.h, which is documented as needed for this function. It doesn't
1842 seem to matter on Linux, but it showed up on some releases of OS X.
1844 3. It seems that there are systems where bytes whose values are greater than
1845 127 match isprint() in the "C" locale. The "C" locale should be the
1846 default when a C program starts up. In most systems, only ASCII printing
1847 characters match isprint(). This difference caused the output from pcretest
1848 to vary, making some of the tests fail. I have changed pcretest so that:
1850 (a) When it is outputting text in the compiled version of a pattern, bytes
1851 other than 32-126 are always shown as hex escapes.
1853 (b) When it is outputting text that is a matched part of a subject string,
1854 it does the same, unless a different locale has been set for the match
1855 (using the /L modifier). In this case, it uses isprint() to decide.
1857 4. Fixed a major bug that caused incorrect computation of the amount of memory
1858 required for a compiled pattern when options that changed within the
1859 pattern affected the logic of the preliminary scan that determines the
1860 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1861 that the computed length was too small. The symptoms of this bug were
1862 either the PCRE error "internal error: code overflow" from pcre_compile(),
1863 or a glibc crash with a message such as "pcretest: free(): invalid next
1864 size (fast)". Examples of patterns that provoked this bug (shown in
1865 pcretest format) are:
1867 /(?-x: )/x
1868 /(?x)(?-x: \s*#\s*)/
1869 /((?i)[\x{c0}])/8
1870 /(?i:[\x{c0}])/8
1872 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1873 is now done differently.
1875 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1876 wrapper classes; (b) implement a new function in the C++ scanner that is
1877 more efficient than the old way of doing things because it avoids levels of
1878 recursion in the regex matching; (c) add a paragraph to the documentation
1879 for the FullMatch() function.
1881 6. The escape sequence \n was being treated as whatever was defined as
1882 "newline". Not only was this contrary to the documentation, which states
1883 that \n is character 10 (hex 0A), but it also went horribly wrong when
1884 "newline" was defined as CRLF. This has been fixed.
1886 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1887 was being set to -1 for the "end of line" case (supposedly a value that no
1888 character can have). Though this value is never used (the check for end of
1889 line is "zero bytes in current character"), it caused compiler complaints.
1890 I've changed it to 0xffffffff.
1892 8. In pcre_version.c, the version string was being built by a sequence of
1893 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1894 string (as it is for production releases) called a macro with an empty
1895 argument. The C standard says the result of this is undefined. The gcc
1896 compiler treats it as an empty string (which was what was wanted) but it is
1897 reported that Visual C gives an error. The source has been hacked around to
1898 avoid this problem.
1900 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1901 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1902 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1903 of them did).
1905 10. Originally, pcretest opened its input and output without "b"; then I was
1906 told that "b" was needed in some environments, so it was added for release
1907 5.0 to both the input and output. (It makes no difference on Unix-like
1908 systems.) Later I was told that it is wrong for the input on Windows. I've
1909 now abstracted the modes into two macros, to make it easier to fiddle with
1910 them, and removed "b" from the input mode under Windows.
1912 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1914 12. Added -help and --help to pcretest as an official way of being reminded
1915 of the options.
1917 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1918 and pcrecpp.cc because they annoy compilers at high warning levels.
1920 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1922 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1923 scripts are not C programs :-) and which was not noticed because it works
1924 on Linux.
1926 16. pcretest is supposed to handle any length of pattern and data line (as one
1927 line or as a continued sequence of lines) by extending its input buffer if
1928 necessary. This feature was broken for very long pattern lines, leading to
1929 a string of junk being passed to pcre_compile() if the pattern was longer
1930 than about 50K.
1932 17. I have done a major re-factoring of the way pcre_compile() computes the
1933 amount of memory needed for a compiled pattern. Previously, there was code
1934 that made a preliminary scan of the pattern in order to do this. That was
1935 OK when PCRE was new, but as the facilities have expanded, it has become
1936 harder and harder to keep it in step with the real compile phase, and there
1937 have been a number of bugs (see for example, 4 above). I have now found a
1938 cunning way of running the real compile function in a "fake" mode that
1939 enables it to compute how much memory it would need, while actually only
1940 ever using a few hundred bytes of working memory and without too many
1941 tests of the mode. This should make future maintenance and development
1942 easier. A side effect of this work is that the limit of 200 on the nesting
1943 depth of parentheses has been removed (though this was never a serious
1944 limitation, I suspect). However, there is a downside: pcre_compile() now
1945 runs more slowly than before (30% or more, depending on the pattern). I
1946 hope this isn't a big issue. There is no effect on runtime performance.
1948 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1949 newline (only possible for the last line of a file) and it was a
1950 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1952 19. Added additional timing features to pcretest. (1) The -tm option now times
1953 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1954 separate command line item, by a number that specifies the number of
1955 repeats to use when timing. The default is 50000; this gives better
1956 precision, but takes uncomfortably long for very large patterns.
1958 20. Extended pcre_study() to be more clever in cases where a branch of a
1959 subpattern has no definite first character. For example, (a*|b*)[cd] would
1960 previously give no result from pcre_study(). Now it recognizes that the
1961 first character must be a, b, c, or d.
1963 21. There was an incorrect error "recursive call could loop indefinitely" if
1964 a subpattern (or the entire pattern) that was being tested for matching an
1965 empty string contained only one non-empty item after a nested subpattern.
1966 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1967 incorrectly, because the \d was being skipped in the check.
1969 22. The pcretest program now has a new pattern option /B and a command line
1970 option -b, which is equivalent to adding /B to every pattern. This causes
1971 it to show the compiled bytecode, without the additional information that
1972 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1973 is the same as /B/I).
1975 23. A new optimization is now able automatically to treat some sequences such
1976 as a*b as a*+b. More specifically, if something simple (such as a character
1977 or a simple class like \d) has an unlimited quantifier, and is followed by
1978 something that cannot possibly match the quantified thing, the quantifier
1979 is automatically "possessified".
1981 24. A recursive reference to a subpattern whose number was greater than 39
1982 went wrong under certain circumstances in UTF-8 mode. This bug could also
1983 have affected the operation of pcre_study().
1985 25. Realized that a little bit of performance could be had by replacing
1986 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1988 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1990 27. Possessive quantifiers such as a++ were previously implemented by turning
1991 them into atomic groups such as ($>a+). Now they have their own opcodes,
1992 which improves performance. This includes the automatically created ones
1993 from 23 above.
1995 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1996 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1997 the first matched character to be a colon. This applied both to named and
1998 numbered groups.
2000 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2002 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2003 building PCRE on Windows easier, so I have included it in the distribution.
2005 31. There is now a check in pcretest against a ridiculously large number being
2006 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2007 loop, the loop is abandoned.
2009 32. Forward references to subpatterns in conditions such as (?(2)...) where
2010 subpattern 2 is defined later cause pcre_compile() to search forwards in
2011 the pattern for the relevant set of parentheses. This search went wrong
2012 when there were unescaped parentheses in a character class, parentheses
2013 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2015 33. "Subroutine" calls and backreferences were previously restricted to
2016 referencing subpatterns earlier in the regex. This restriction has now
2017 been removed.
2019 34. Added a number of extra features that are going to be in Perl 5.10. On the
2020 whole, these are just syntactic alternatives for features that PCRE had
2021 previously implemented using the Python syntax or my own invention. The
2022 other formats are all retained for compatibility.
2024 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2025 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2026 also .NET compatible.
2028 (b) A recursion or subroutine call to a named group can now be defined as
2029 (?&name) as well as (?P>name).
2031 (c) A backreference to a named group can now be defined as \k<name> or
2032 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2033 5.10, are also .NET compatible.
2035 (d) A conditional reference to a named group can now use the syntax
2036 (?(<name>) or (?('name') as well as (?(name).
2038 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2039 groups (named and numbered) that are never evaluated inline, but can be
2040 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2041 is always false. There may be only one alternative in such a group.
2043 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2044 as the simple (?(R). The condition is true only if the most recent
2045 recursion is that of the given number or name. It does not search out
2046 through the entire recursion stack.
2048 (g) The escape \gN or \g{N} has been added, where N is a positive or
2049 negative number, specifying an absolute or relative reference.
2051 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2052 some "unreachable code" warnings.
2054 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2055 things, this adds five new scripts.
2057 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2058 There were also incompatibilities regarding the handling of \Q..\E inside
2059 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2060 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2062 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2063 matches an empty string, and forcibly breaks the loop. There were bugs in
2064 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2065 against aaaa the result was just "a" rather than "aaaa", for example. Two
2066 separate and independent bugs (that affected different cases) have been
2067 fixed.
2069 39. Refactored the code to abolish the use of different opcodes for small
2070 capturing bracket numbers. This is a tidy that I avoided doing when I
2071 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2072 The new approach is not only tidier, it makes it possible to reduce the
2073 memory needed to fix the previous bug (38).
2075 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2076 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2077 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2078 mode.
2080 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2081 report.
2083 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2084 copy construction and assignment in the C++ wrapper.
2086 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2087 couple of bugs that could have given wrong results in the "--newline=crlf"
2088 case.
2090 44. Added a number of casts and did some reorganization of signed/unsigned int
2091 variables following suggestions from Dair Grant. Also renamed the variable
2092 "this" as "item" because it is a C++ keyword.
2094 45. Arranged for dftables to add
2096 #include "pcre_internal.h"
2098 to pcre_chartables.c because without it, gcc 4.x may remove the array
2099 definition from the final binary if PCRE is built into a static library and
2100 dead code stripping is activated.
2102 46. For an unanchored pattern, if a match attempt fails at the start of a
2103 newline sequence, and the newline setting is CRLF or ANY, and the next two
2104 characters are CRLF, advance by two characters instead of one.
2107 Version 6.7 04-Jul-06
2108 ---------------------
2110 1. In order to handle tests when input lines are enormously long, pcretest has
2111 been re-factored so that it automatically extends its buffers when
2112 necessary. The code is crude, but this _is_ just a test program. The
2113 default size has been increased from 32K to 50K.
2115 2. The code in pcre_study() was using the value of the re argument before
2116 testing it for NULL. (Of course, in any sensible call of the function, it
2117 won't be NULL.)
2119 3. The memmove() emulation function in pcre_internal.h, which is used on
2120 systems that lack both memmove() and bcopy() - that is, hardly ever -
2121 was missing a "static" storage class specifier.
2123 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2124 containing an extended class (one that cannot be represented by a bitmap
2125 because it contains high-valued characters or Unicode property items, e.g.
2126 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2127 pattern, but PCRE should not loop if you do not (it no longer does).
2128 [Detail: two cases were found: (a) a repeated subpattern containing an
2129 extended class; (b) a recursive reference to a subpattern that followed a
2130 previous extended class. It wasn't skipping over the extended class
2131 correctly when UTF-8 mode was not set.]
2133 5. A negated single-character class was not being recognized as fixed-length
2134 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2135 compile error "lookbehind assertion is not fixed length".
2137 6. The RunPerlTest auxiliary script was showing an unexpected difference
2138 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2139 write a Perl script that can interpret lines of an input file either as
2140 byte characters or as UTF-8, which is what "perltest" was being required to
2141 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2142 can't do is switch easily at run time between having the "use utf8;" pragma
2143 or not. In the end, I fudged it by using the RunPerlTest script to insert
2144 "use utf8;" explicitly for the UTF-8 tests.
2146 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2147 the end of the subject string, contrary to the documentation and to what
2148 Perl does. This was true of both matching functions. Now it matches only at
2149 the start of the subject and immediately after *internal* newlines.
2151 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2152 a pointer to an int instead of a pointer to an unsigned long int. This
2153 caused problems on 64-bit systems.
2155 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2156 instance of the 'standard' template library not being so standard".
2158 10. There was no check on the number of named subpatterns nor the maximum
2159 length of a subpattern name. The product of these values is used to compute
2160 the size of the memory block for a compiled pattern. By supplying a very
2161 long subpattern name and a large number of named subpatterns, the size
2162 computation could be caused to overflow. This is now prevented by limiting
2163 the length of names to 32 characters, and the number of named subpatterns
2164 to 10,000.
2166 11. Subpatterns that are repeated with specific counts have to be replicated in
2167 the compiled pattern. The size of memory for this was computed from the
2168 length of the subpattern and the repeat count. The latter is limited to
2169 65535, but there was no limit on the former, meaning that integer overflow
2170 could in principle occur. The compiled length of a repeated subpattern is
2171 now limited to 30,000 bytes in order to prevent this.
2173 12. Added the optional facility to have named substrings with the same name.
2175 13. Added the ability to use a named substring as a condition, using the
2176 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2177 are numbers (not recommended). Forward references are permitted.
2179 14. Added forward references in named backreferences (if you see what I mean).
2181 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2182 pattern could run off the end of the subject. For example, the pattern
2183 "(?s)(.{1,5})"8 did this with the subject "ab".
2185 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2186 PCRE_CASELESS was set when matching characters that were quantified with ?
2187 or *.
2189 17. A character class other than a single negated character that had a minimum
2190 but no maximum quantifier - for example [ab]{6,} - was not handled
2191 correctly by pce_dfa_exec(). It would match only one character.
2193 18. A valid (though odd) pattern that looked like a POSIX character
2194 class but used an invalid character after [ (for example [[,abc,]]) caused
2195 pcre_compile() to give the error "Failed: internal error: code overflow" or
2196 in some cases to crash with a glibc free() error. This could even happen if
2197 the pattern terminated after [[ but there just happened to be a sequence of
2198 letters, a binary zero, and a closing ] in the memory that followed.
2200 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2201 over the years. Originally (before any Unicode support), just the bottom 8
2202 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2203 output from "man perlunicode" includes this:
2205 The regular expression compiler produces polymorphic opcodes. That
2206 is, the pattern adapts to the data and automatically switches to
2207 the Unicode character scheme when presented with Unicode data--or
2208 instead uses a traditional byte scheme when presented with byte
2209 data.
2211 Sadly, a wide octal escape does not cause a switch, and in a string with
2212 no other multibyte characters, these octal escapes are treated as before.
2213 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2214 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2215 Unicode string.
2217 I have not perpetrated such confusion in PCRE. Up till now, it took just
2218 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2219 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2220 translate to the appropriate multibyte character.
2222 29. Applied some refactoring to reduce the number of warnings from Microsoft
2223 and Borland compilers. This has included removing the fudge introduced
2224 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2225 a warning about an unused variable.
2227 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2228 characters since release 4.0, because Perl (from release 5.004) does not.
2229 [Or at least, is documented not to: some releases seem to be in conflict
2230 with the documentation.] However, when a pattern was studied with
2231 pcre_study() and all its branches started with \s, PCRE still included VT
2232 as a possible starting character. Of course, this did no harm; it just
2233 caused an unnecessary match attempt.
2235 22. Removed a now-redundant internal flag bit that recorded the fact that case
2236 dependency changed within the pattern. This was once needed for "required
2237 byte" processing, but is no longer used. This recovers a now-scarce options
2238 bit. Also moved the least significant internal flag bit to the most-
2239 significant bit of the word, which was not previously used (hangover from
2240 the days when it was an int rather than a uint) to free up another bit for
2241 the future.
2243 23. Added support for CRLF line endings as well as CR and LF. As well as the
2244 default being selectable at build time, it can now be changed at runtime
2245 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2246 specify that it is scanning data with non-default line endings.
2248 24. Changed the definition of CXXLINK to make it agree with the definition of
2249 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2251 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2252 recursions. This makes a big different to stack usage for some patterns.
2254 26. If a subpattern containing a named recursion or subroutine reference such
2255 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2256 the space required for the compiled pattern went wrong and gave too small a
2257 value. Depending on the environment, this could lead to "Failed: internal
2258 error: code overflow at offset 49" or "glibc detected double free or
2259 corruption" errors.
2261 27. Applied patches from Google (a) to support the new newline modes and (b) to
2262 advance over multibyte UTF-8 characters in GlobalReplace.
2264 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2265 difference for some implementation of PCRE in some Windows version.
2267 29. Added some extra testing facilities to pcretest:
2269 \q<number> in a data line sets the "match limit" value
2270 \Q<number> in a data line sets the "match recursion limt" value
2271 -S <number> sets the stack size, where <number> is in megabytes
2273 The -S option isn't available for Windows.
2276 Version 6.6 06-Feb-06
2277 ---------------------
2279 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2280 in pcreposix.h. I have copied the definition from pcre.h.
2282 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2283 because pcre.h is no longer a built file.
2285 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2286 not normally included in the compiled code.
2289 Version 6.5 01-Feb-06
2290 ---------------------
2292 1. When using the partial match feature with pcre_dfa_exec(), it was not
2293 anchoring the second and subsequent partial matches at the new starting
2294 point. This could lead to incorrect results. For example, with the pattern
2295 /1234/, partially matching against "123" and then "a4" gave a match.
2297 2. Changes to pcregrep:
2299 (a) All non-match returns from pcre_exec() were being treated as failures
2300 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2301 error message is output. Some extra information is given for the
2303 probably the only errors that are likely to be caused by users (by
2304 specifying a regex that has nested indefinite repeats, for instance).
2305 If there are more than 20 of these errors, pcregrep is abandoned.
2307 (b) A binary zero was treated as data while matching, but terminated the
2308 output line if it was written out. This has been fixed: binary zeroes
2309 are now no different to any other data bytes.
2311 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2312 used to set a locale for matching. The --locale=xxxx long option has
2313 been added (no short equivalent) to specify a locale explicitly on the
2314 pcregrep command, overriding the environment variables.
2316 (d) When -B was used with -n, some line numbers in the output were one less
2317 than they should have been.
2319 (e) Added the -o (--only-matching) option.
2321 (f) If -A or -C was used with -c (count only), some lines of context were
2322 accidentally printed for the final match.
2324 (g) Added the -H (--with-filename) option.
2326 (h) The combination of options -rh failed to suppress file names for files
2327 that were found from directory arguments.
2329 (i) Added the -D (--devices) and -d (--directories) options.
2331 (j) Added the -F (--fixed-strings) option.
2333 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2335 (l) Added the --colo(u)r option.
2337 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2338 is not present by default.
2340 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2341 items such as (?R) or (?1), when the recursion could match a number of
2342 alternatives. If it matched one of the alternatives, but subsequently,
2343 outside the recursion, there was a failure, the code tried to back up into
2344 the recursion. However, because of the way PCRE is implemented, this is not
2345 possible, and the result was an incorrect result from the match.
2347 In order to prevent this happening, the specification of recursion has
2348 been changed so that all such subpatterns are automatically treated as
2349 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2351 4. I had overlooked the fact that, in some locales, there are characters for
2352 which isalpha() is true but neither isupper() nor islower() are true. In
2353 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2354 and ordfeminine) are like this. This affected the treatment of \w and \W
2355 when they appeared in character classes, but not when they appeared outside
2356 a character class. The bit map for "word" characters is now created
2357 separately from the results of isalnum() instead of just taking it from the
2358 upper, lower, and digit maps. (Plus the underscore character, of course.)
2360 5. The above bug also affected the handling of POSIX character classes such as
2361 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2362 permanent tables. Instead, the bit maps for such a class were previously
2363 created as the appropriate unions of the upper, lower, and digit bitmaps.
2364 Now they are created by subtraction from the [[:word:]] class, which has
2365 its own bitmap.
2367 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2368 It is created by subtracting the vertical space characters (\x09, \x0a,
2369 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2370 subtraction was done in the overall bitmap for a character class, meaning
2371 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2372 be recognized. This bug has been fixed.
2374 7. Patches from the folks at Google:
2376 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2377 real life, but is still worth protecting against".
2379 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2380 regular expressions".
2382 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2383 have it.
2385 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2386 "configure" and the latter not, in order to fix a problem somebody had
2387 with compiling the Arg class on HP-UX.
2389 (e) Improve the error-handling of the C++ wrapper a little bit.
2391 (f) New tests for checking recursion limiting.
2393 8. The pcre_memmove() function, which is used only if the environment does not
2394 have a standard memmove() function (and is therefore rarely compiled),
2395 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2396 returning a result (though PCRE never actually uses the result).
2398 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2399 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2400 returned instead of calling malloc() with an overflowing number that would
2401 most likely cause subsequent chaos.
2403 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2405 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2406 with this option is matched, the nmatch and pmatch options of regexec() are
2407 ignored.
2409 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2410 provided in case anyone wants to the the POSIX interface with UTF-8
2411 strings.
2413 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2414 C++ linking (needed for some HP-UX environments).
2416 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2417 (unused parameter) and in the pcre_printint() function (omitted "default"
2418 switch label when the default is to do nothing).
2420 15. Added some code to make it possible, when PCRE is compiled as a C++
2421 library, to replace subject pointers for pcre_exec() with a smart pointer
2422 class, thus making it possible to process discontinuous strings.
2424 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2425 much the same function. They were added by different people who were trying
2426 to make PCRE easy to compile on non-Unix systems. It has been suggested
2427 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2428 for compiling on Windows systems. I have therefore replaced it with
2429 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2430 defaults to "extern" for C or "extern C" for C++, which works fine on
2431 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2432 SCOPE with something explicit in config.h. In addition:
2434 (a) pcreposix.h still had just "extern" instead of either of these macros;
2435 I have replaced it with PCRE_DATA_SCOPE.
2437 (b) Functions such as _pcre_xclass(), which are internal to the library,
2438 but external in the C sense, all had PCRE_EXPORT in their definitions.
2439 This is apparently wrong for the Windows case, so I have removed it.
2440 (It makes no difference on Unix-like systems.)
2442 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2443 of recursive calls to match(). This is different to MATCH_LIMIT because
2444 that limits the total number of calls to match(), not all of which increase
2445 the depth of recursion. Limiting the recursion depth limits the amount of
2446 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2447 when PCRE is compiled, and changed at run time. A patch from Google adds
2448 this functionality to the C++ interface.
2450 18. Changes to the handling of Unicode character properties:
2452 (a) Updated the table to Unicode 4.1.0.
2454 (b) Recognize characters that are not in the table as "Cn" (undefined).
2456 (c) I revised the way the table is implemented to a much improved format
2457 which includes recognition of ranges. It now supports the ranges that
2458 are defined in UnicodeData.txt, and it also amalgamates other
2459 characters into ranges. This has reduced the number of entries in the
2460 table from around 16,000 to around 3,000, thus reducing its size
2461 considerably. I realized I did not need to use a tree structure after
2462 all - a binary chop search is just as efficient. Having reduced the
2463 number of entries, I extended their size from 6 bytes to 8 bytes to
2464 allow for more data.
2466 (d) Added support for Unicode script names via properties such as \p{Han}.
2468 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2469 matching that character.
2471 20. When matching a repeated Unicode property with a minimum greater than zero,
2472 (for example \pL{2,}), PCRE could look past the end of the subject if it
2473 reached it while seeking the minimum number of characters. This could
2474 happen only if some of the characters were more than one byte long, because
2475 there is a check for at least the minimum number of bytes.
2477 21. Refactored the implementation of \p and \P so as to be more general, to
2478 allow for more different types of property in future. This has changed the
2479 compiled form incompatibly. Anybody with saved compiled patterns that use
2480 \p or \P will have to recompile them.
2482 22. Added "Any" and "L&" to the supported property types.
2484 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2485 but give a compile time error if the value is greater than 0xff.
2487 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2488 accidentally not being installed or uninstalled.
2490 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2491 made were to insert the current release number. This seemed silly, because
2492 it made things harder for people building PCRE on systems that don't run
2493 "configure". I have turned pcre.h into a distributed file, no longer built
2494 by "configure", with the version identification directly included. There is
2495 no longer a pcre.h.in file.
2497 However, this change necessitated a change to the pcre-config script as
2498 well. It is built from pcre-config.in, and one of the substitutions was the
2499 release number. I have updated configure.ac so that ./configure now finds
2500 the release number by grepping pcre.h.
2502 26. Added the ability to run the tests under valgrind.
2505 Version 6.4 05-Sep-05
2506 ---------------------
2508 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2509 "--" to be printed when multiple files were scanned, even when none of the
2510 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2511 consider it to be a bug, and have restored the previous behaviour.
2513 2. A couple of code tidies to get rid of compiler warnings.
2515 3. The pcretest program used to cheat by referring to symbols in the library
2516 whose names begin with _pcre_. These are internal symbols that are not
2517 really supposed to be visible externally, and in some environments it is
2518 possible to suppress them. The cheating is now confined to including
2519 certain files from the library's source, which is a bit cleaner.
2521 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2522 file's purpose clearer.
2524 5. Reorganized pcre_ucp_findchar().
2527 Version 6.3 15-Aug-05
2528 ---------------------
2530 1. The file libpcre.pc.in did not have general read permission in the tarball.
2532 2. There were some problems when building without C++ support:
2534 (a) If C++ support was not built, "make install" and "make test" still
2535 tried to test it.
2537 (b) There were problems when the value of CXX was explicitly set. Some
2538 changes have been made to try to fix these, and ...
2540 (c) --disable-cpp can now be used to explicitly disable C++ support.
2542 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2543 backslash in a target when C++ was disabled. This confuses some
2544 versions of "make", apparently. Using an intermediate variable solves
2545 this. (Same for CPP_LOBJ.)
2547 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2548 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2549 necessary on certain architectures.
2551 4. Added a setting of -export-symbols-regex to the link command to remove
2552 those symbols that are exported in the C sense, but actually are local
2553 within the library, and not documented. Their names all begin with
2554 "_pcre_". This is not a perfect job, because (a) we have to except some
2555 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2556 available (and never for static libraries). I have made a note to try to
2557 find a way round (a) in the future.
2560 Version 6.2 01-Aug-05
2561 ---------------------
2563 1. There was no test for integer overflow of quantifier values. A construction
2564 such as {1111111111111111} would give undefined results. What is worse, if
2565 a minimum quantifier for a parenthesized subpattern overflowed and became
2566 negative, the calculation of the memory size went wrong. This could have
2567 led to memory overwriting.
2569 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2571 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2572 operating environments where this matters.
2574 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2575 PCRE options from within the C++ wrapper.
2577 5. Named capturing subpatterns were not being correctly counted when a pattern
2578 was compiled. This caused two problems: (a) If there were more than 100
2579 such subpatterns, the calculation of the memory needed for the whole
2580 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2581 back references of the form \12, where the number was greater than 9, were
2582 not recognized as back references, even though there were sufficient
2583 previous subpatterns.
2585 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2586 versions of gcc, e.g. 2.95.4.
2589 Version 6.1 21-Jun-05
2590 ---------------------
2592 1. There was one reference to the variable "posix" in pcretest.c that was not
2593 surrounded by "#if !defined NOPOSIX".
2595 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2596 the cross-check on the old pcre_info() function, for the benefit of the
2597 cut-down version of PCRE that is currently imported into Exim.
2599 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2600 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2601 patterns that start (?i) but I don't think that matters. The use of (?i) is
2602 just an example; this all applies to the other options as well.
2604 4. Since libtool seems to echo the compile commands it is issuing, the output
2605 from "make" can be reduced a bit by putting "@" in front of each libtool
2606 compile command.
2608 5. Patch from the folks at Google for configure.in to be a bit more thorough
2609 in checking for a suitable C++ installation before trying to compile the
2610 C++ stuff. This should fix a reported problem when a compiler was present,
2611 but no suitable headers.
2613 6. The man pages all had just "PCRE" as their title. I have changed them to
2614 be the relevant file name. I have also arranged that these names are
2615 retained in the file doc/pcre.txt, which is a concatenation in text format
2616 of all the man pages except the little individual ones for each function.
2618 7. The NON-UNIX-USE file had not been updated for the different set of source
2619 files that come with release 6. I also added a few comments about the C++
2620 wrapper.
2623 Version 6.0 07-Jun-05
2624 ---------------------
2626 1. Some minor internal re-organization to help with my DFA experiments.
2628 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2629 didn't matter for the library itself when fully configured, but did matter
2630 when compiling without UCP support, or within Exim, where the ucp files are
2631 not imported.
2633 3. Refactoring of the library code to split up the various functions into
2634 different source modules. The addition of the new DFA matching code (see
2635 below) to a single monolithic source would have made it really too
2636 unwieldy, quite apart from causing all the code to be include in a
2637 statically linked application, when only some functions are used. This is
2638 relevant even without the DFA addition now that patterns can be compiled in
2639 one application and matched in another.
2641 The downside of splitting up is that there have to be some external
2642 functions and data tables that are used internally in different modules of
2643 the library but which are not part of the API. These have all had their
2644 names changed to start with "_pcre_" so that they are unlikely to clash
2645 with other external names.
2647 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2648 a different (DFA) algorithm. Although it is slower than the original
2649 function, it does have some advantages for certain types of matching
2650 problem.
2652 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2653 including restarting after a partial match.
2655 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2656 defined when compiling for Windows was sent to me. I have put it into the
2657 code, though I have no means of testing or verifying it.
2659 7. Added the pcre_refcount() auxiliary function.
2661 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2662 match before or at the first newline in the subject string. In pcretest,
2663 the /f option on a pattern can be used to set this.
2665 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2666 would behave wrongly. This has been present in PCRE since release 4.0.
2668 10. A number of changes to the pcregrep command:
2670 (a) Refactored how -x works; insert ^(...)$ instead of setting
2671 PCRE_ANCHORED and checking the length, in preparation for adding
2672 something similar for -w.
2674 (b) Added the -w (match as a word) option.
2676 (c) Refactored the way lines are read and buffered so as to have more
2677 than one at a time available.
2679 (d) Implemented a pcregrep test script.
2681 (e) Added the -M (multiline match) option. This allows patterns to match
2682 over several lines of the subject. The buffering ensures that at least
2683 8K, or the rest of the document (whichever is the shorter) is available
2684 for matching (and similarly the previous 8K for lookbehind assertions).
2686 (f) Changed the --help output so that it now says
2688 -w, --word-regex(p)
2690 instead of two lines, one with "regex" and the other with "regexp"
2691 because that confused at least one person since the short forms are the
2692 same. (This required a bit of code, as the output is generated
2693 automatically from a table. It wasn't just a text change.)
2695 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2696 option but starts with a hyphen. Could be a pattern or a path name
2697 starting with a hyphen, for instance.
2699 (h) "-" can be given as a file name to represent stdin.
2701 (i) When file names are being printed, "(standard input)" is used for
2702 the standard input, for compatibility with GNU grep. Previously
2703 "<stdin>" was used.
2705 (j) The option --label=xxx can be used to supply a name to be used for
2706 stdin when file names are being printed. There is no short form.
2708 (k) Re-factored the options decoding logic because we are going to add
2709 two more options that take data. Such options can now be given in four
2710 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2712 (l) Added the -A, -B, and -C options for requesting that lines of context
2713 around matches be printed.
2715 (m) Added the -L option to print the names of files that do not contain
2716 any matching lines, that is, the complement of -l.
2718 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2719 continue to scan other files.
2721 (o) The -s option was incorrectly implemented. For compatibility with other
2722 greps, it now suppresses the error message for a non-existent or non-
2723 accessible file (but not the return code). There is a new option called
2724 -q that suppresses the output of matching lines, which was what -s was
2725 previously doing.
2727 (p) Added --include and --exclude options to specify files for inclusion
2728 and exclusion when recursing.
2730 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2731 Hopefully, it now does.
2733 12. Missing cast in pcre_study().
2735 13. Added an "uninstall" target to the makefile.
2737 14. Replaced "extern" in the function prototypes in Makefile.in with
2738 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2739 world, but is set differently for Windows.
2741 15. Added a second compiling function called pcre_compile2(). The only
2742 difference is that it has an extra argument, which is a pointer to an
2743 integer error code. When there is a compile-time failure, this is set
2744 non-zero, in addition to the error test pointer being set to point to an
2745 error message. The new argument may be NULL if no error number is required
2746 (but then you may as well call pcre_compile(), which is now just a
2747 wrapper). This facility is provided because some applications need a
2748 numeric error indication, but it has also enabled me to tidy up the way
2749 compile-time errors are handled in the POSIX wrapper.
2751 16. Added VPATH=.libs to the makefile; this should help when building with one
2752 prefix path and installing with another. (Or so I'm told by someone who
2753 knows more about this stuff than I do.)
2755 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2756 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2757 match everything, including newlines. This is not POSIX-compatible, but
2758 somebody wanted the feature. From pcretest it can be activated by using
2759 both the P and the s flags.
2761 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2763 19. libpcre.pc was being incorrectly installed as executable.
2765 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2766 it now also looks for '\r' so that it will work unmodified on Windows.
2768 21. Added Google's contributed C++ wrapper to the distribution.
2770 22. Added some untidy missing memory free() calls in pcretest, to keep
2771 Electric Fence happy when testing.
2775 Version 5.0 13-Sep-04
2776 ---------------------
2778 1. Internal change: literal characters are no longer packed up into items
2779 containing multiple characters in a single byte-string. Each character
2780 is now matched using a separate opcode. However, there may be more than one
2781 byte in the character in UTF-8 mode.
2783 2. The pcre_callout_block structure has two new fields: pattern_position and
2784 next_item_length. These contain the offset in the pattern to the next match
2785 item, and its length, respectively.
2787 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2788 insertion of callouts before each pattern item. Added the /C option to
2789 pcretest to make use of this.
2791 4. On the advice of a Windows user, the lines
2793 #if defined(_WIN32) || defined(WIN32)
2794 _setmode( _fileno( stdout ), 0x8000 );
2795 #endif /* defined(_WIN32) || defined(WIN32) */
2797 have been added to the source of pcretest. This apparently does useful
2798 magic in relation to line terminators.
2800 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2801 for the benefit of those environments where the "b" makes a difference.
2803 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2804 to know about it. I have put a hack into configure.in that adds in code
2805 to set GCC=yes if CC=icc. This seems to end up at a point in the
2806 generated configure script that is early enough to affect the setting of
2807 compiler options, which is what is needed, but I have no means of testing
2808 whether it really works. (The user who reported this had patched the
2809 generated configure script, which of course I cannot do.)
2811 LATER: After change 22 below (new libtool files), the configure script
2812 seems to know about icc (and also ecc). Therefore, I have commented out
2813 this hack in configure.in.
2815 7. Added support for pkg-config (2 patches were sent in).
2817 8. Negated POSIX character classes that used a combination of internal tables
2818 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2819 [[:^ascii]]. Typically, they would match almost any characters. The other
2820 POSIX classes were not broken in this way.
2822 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2823 to find the match, as PCRE was deluded into thinking that the match had to
2824 start at the start point or following a newline. The same bug applied to
2825 patterns with negative forward assertions or any backward assertions
2826 preceding ".*" at the start, unless the pattern required a fixed first
2827 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2829 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2830 starting at the last subject character, bytes beyond the end of the subject
2831 string were read.
2833 11. Renamed the variable "class" as "classbits" to make life easier for C++
2834 users. (Previously there was a macro definition, but it apparently wasn't
2835 enough.)
2837 12. Added the new field "tables" to the extra data so that tables can be passed
2838 in at exec time, or the internal tables can be re-selected. This allows
2839 a compiled regex to be saved and re-used at a later time by a different
2840 program that might have everything at different addresses.
2842 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2843 -R library as well as a -L library.
2845 14. The debugging options of pcretest (-d on the command line or D on a
2846 pattern) showed incorrect output for anything following an extended class
2847 that contained multibyte characters and which was followed by a quantifier.
2849 15. Added optional support for general category Unicode character properties
2850 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2851 support. It adds about 90K to the size of the library. The meanings of the
2852 inbuilt class escapes such as \d and \s have NOT been changed.
2854 16. Updated pcredemo.c to include calls to free() to release the memory for the
2855 compiled pattern.
2857 17. The generated file chartables.c was being created in the source directory
2858 instead of in the building directory. This caused the build to fail if the
2859 source directory was different from the building directory, and was
2860 read-only.
2862 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2863 file. No doubt somebody will tell me if they don't make sense... Also added
2864 Dan Mooney's comments about building on OpenVMS.
2866 19. Added support for partial matching via the PCRE_PARTIAL option for
2867 pcre_exec() and the \P data escape in pcretest.
2869 20. Extended pcretest with 3 new pattern features:
2871 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2872 write the compiled pattern to the file whose name is "rest-of-line".
2873 This is a straight binary dump of the data, with the saved pointer to
2874 the character tables forced to be NULL. The study data, if any, is
2875 written too. After writing, pcretest reads a new pattern.
2877 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2878 compiled pattern from the given file. There must not be any
2879 occurrences of "<" in the file name (pretty unlikely); if there are,
2880 pcretest will instead treat the initial "<" as a pattern delimiter.
2881 After reading in the pattern, pcretest goes on to read data lines as
2882 usual.
2884 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2885 and 16-bit fields in a compiled pattern, to simulate a pattern that
2886 was compiled on a host of opposite endianness.
2888 21. The pcre-exec() function can now cope with patterns that were compiled on
2889 hosts of opposite endianness, with this restriction:
2891 As for any compiled expression that is saved and used later, the tables
2892 pointer field cannot be preserved; the extra_data field in the arguments
2893 to pcre_exec() should be used to pass in a tables address if a value
2894 other than the default internal tables were used at compile time.
2896 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2897 now diagnosed as an error. Previously, most of the time, a negative number
2898 would have been treated as zero, but if in addition "ovector" was passed as
2899 NULL, a crash could occur.
2901 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2902 new versions from the libtool 1.5 distribution (the last one is a copy of
2903 a file called libtool.m4). This seems to have fixed the need to patch
2904 "configure" to support Darwin 1.3 (which I used to do). However, I still
2905 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2906 workstation).
2908 24. Changed the PCRE licence to be the more standard "BSD" licence.
2911 Version 4.5 01-Dec-03
2912 ---------------------
2914 1. There has been some re-arrangement of the code for the match() function so
2915 that it can be compiled in a version that does not call itself recursively.
2916 Instead, it keeps those local variables that need separate instances for
2917 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2918 needs to "recurse". Keeping track of where control must go is done by means
2919 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2920 hide most of the details from the main code, and operates only if
2921 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2922 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2923 operating.
2925 To make it easier for callers to provide specially tailored get/free
2926 functions for this usage, two new functions, pcre_stack_malloc, and
2927 pcre_stack_free, are used. They are always called in strict stacking order,
2928 and the size of block requested is always the same.
2930 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2931 PCRE has been compiled to use the stack or the heap for recursion. The
2932 -C option of pcretest uses this to show which version is compiled.
2934 A new data escape \S, is added to pcretest; it causes the amounts of store
2935 obtained and freed by both kinds of malloc/free at match time to be added
2936 to the output.
2938 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2939 what's available on my current Linux desktop machine.
2941 3. When matching a UTF-8 string, the test for a valid string at the start has
2942 been extended. If start_offset is not zero, PCRE now checks that it points
2943 to a byte that is the start of a UTF-8 character. If not, it returns
2944 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2945 this is necessary because there may be backward assertions in the pattern.
2946 When matching the same subject several times, it may save resources to use
2947 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2949 4. The code for checking the validity of UTF-8 strings has been tightened so
2950 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2951 containing "overlong sequences".
2953 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2954 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2955 should have been "&", but it just so happened that all the cases this let
2956 through by mistake were picked up later in the function.
2958 6. I had used a variable called "isblank" - this is a C99 function, causing
2959 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2961 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2962 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2963 so the version is shown; (c) stop "make test" echoing "./RunTest".
2965 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2967 9. The prototype for memmove() for systems that don't have it was using
2968 size_t, but the inclusion of the header that defines size_t was later. I've
2969 moved the #includes for the C headers earlier to avoid this.
2971 10. Added some adjustments to the code to make it easier to compiler on certain
2972 special systems:
2974 (a) Some "const" qualifiers were missing.
2975 (b) Added the macro EXPORT before all exported functions; by default this
2976 is defined to be empty.
2977 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2978 that it reads its output file name as an argument instead of writing
2979 to the standard output and assuming this can be redirected.
2981 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2982 class containing characters with values greater than 255, PCRE compilation
2983 went into a loop.
2985 12. A recursive reference to a subpattern that was within another subpattern
2986 that had a minimum quantifier of zero caused PCRE to crash. For example,
2987 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2988 recursion. If the recursively-called subpattern itself had a zero repeat,
2989 that was OK.
2991 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2992 buffer into which it was copied (for escape processing) was still set at
2993 1024, so long lines caused crashes.
2995 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2996 "internal error: code overflow...". This applied to any character class
2997 that was followed by a possessive quantifier.
2999 15. Modified the Makefile to add libpcre.la as a prerequisite for
3000 libpcreposix.la because I was told this is needed for a parallel build to
3001 work.
3003 16. If a pattern that contained .* following optional items at the start was
3004 studied, the wrong optimizing data was generated, leading to matching
3005 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3006 matching string must start with a or b or c. The correct conclusion for
3007 this pattern is that a match can start with any character.
3010 Version 4.4 13-Aug-03
3011 ---------------------
3013 1. In UTF-8 mode, a character class containing characters with values between
3014 127 and 255 was not handled correctly if the compiled pattern was studied.
3015 In fixing this, I have also improved the studying algorithm for such
3016 classes (slightly).
3018 2. Three internal functions had redundant arguments passed to them. Removal
3019 might give a very teeny performance improvement.
3021 3. Documentation bug: the value of the capture_top field in a callout is *one
3022 more than* the number of the hightest numbered captured substring.
3024 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3025 in incorrectly linking with a previously installed version. They now link
3026 explicitly with libpcre.la.
3028 5. configure.in no longer needs to recognize Cygwin specially.
3030 6. A problem in pcre.in for Windows platforms is fixed.
3032 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3033 pcretest, it used to include the size of the study block as part of its
3034 output. Unfortunately, the structure contains a field that has a different
3035 size on different hardware architectures. This meant that the tests that
3036 showed this size failed. As the block is currently always of a fixed size,
3037 this information isn't actually particularly useful in pcretest output, so
3038 I have just removed it.
3040 8. Three pre-processor statements accidentally did not start in column 1.
3041 Sadly, there are *still* compilers around that complain, even though
3042 standard C has not required this for well over a decade. Sigh.
3044 9. In pcretest, the code for checking callouts passed small integers in the
3045 callout_data field, which is a void * field. However, some picky compilers
3046 complained about the casts involved for this on 64-bit systems. Now
3047 pcretest passes the address of the small integer instead, which should get
3048 rid of the warnings.
3050 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3051 both compile and run time, and gives an error if an invalid UTF-8 sequence
3052 is found. There is a option for disabling this check in cases where the
3053 string is known to be correct and/or the maximum performance is wanted.
3055 11. In response to a bug report, I changed one line in Makefile.in from
3057 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3058 to
3059 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3061 to look similar to other lines, but I have no way of telling whether this
3062 is the right thing to do, as I do not use Windows. No doubt I'll get told
3063 if it's wrong...
3066 Version 4.3 21-May-03
3067 ---------------------
3069 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3070 Makefile.
3072 2. Some refactoring to improve the quality of the code:
3074 (i) The utf8_table... variables are now declared "const".
3076 (ii) The code for \cx, which used the "case flipping" table to upper case
3077 lower case letters, now just substracts 32. This is ASCII-specific,
3078 but the whole concept of \cx is ASCII-specific, so it seems
3079 reasonable.
3081 (iii) PCRE was using its character types table to recognize decimal and
3082 hexadecimal digits in the pattern. This is silly, because it handles
3083 only 0-9, a-f, and A-F, but the character types table is locale-
3084 specific, which means strange things might happen. A private
3085 table is now used for this - though it costs 256 bytes, a table is
3086 much faster than multiple explicit tests. Of course, the standard
3087 character types table is still used for matching digits in subject
3088 strings against \d.
3090 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3091 ending in _t are). So I've renamed it as ESC_tee.
3093 3. The first argument for regexec() in the POSIX wrapper should have been
3094 defined as "const".
3096 4. Changed pcretest to use malloc() for its buffers so that they can be
3097 Electric Fenced for debugging.
3099 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3100 to read one or more bytes before the start of the subject string. Often this
3101 had no effect on PCRE's behaviour, but in some circumstances it could
3102 provoke a segmentation fault.
3104 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3105 to try to read one or more bytes before the start of the subject string.
3107 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3108 UTF-8 support could misbehave in various ways if the subject string
3109 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3110 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3111 back over UTF-8 characters.)
3114 Version 4.2 14-Apr-03
3115 ---------------------
3117 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3119 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3120 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3121 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3122 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3123 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3125 Note: automatic setting of the BUILD variables is not yet working
3126 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3127 compile-time but not at link-time
3128 [LINK]: use for linking executables only
3129 make different versions for Windows and non-Windows
3130 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3131 libraries
3132 [LINK_FOR_BUILD]: new variable
3133 [OBJEXT]: use throughout
3134 [EXEEXT]: use throughout
3135 <winshared>: new target
3136 <wininstall>: new target
3137 <dftables.o>: use native compiler
3138 <dftables>: use native linker
3139 <install>: handle Windows platform correctly
3140 <clean>: ditto
3141 <check>: ditto
3142 copy DLL to top builddir before testing
3144 As part of these changes, -no-undefined was removed again. This was reported
3145 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3146 in any case.
3148 3. Some tidies to get rid of compiler warnings:
3150 . In the match_data structure, match_limit was an unsigned long int, whereas
3151 match_call_count was an int. I've made them both unsigned long ints.
3153 . In pcretest the fact that a const uschar * doesn't automatically cast to
3154 a void * provoked a warning.
3156 . Turning on some more compiler warnings threw up some "shadow" variables
3157 and a few more missing casts.
3159 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3160 option, a class that contained a single character with a value between 128
3161 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3163 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3164 option, a class that contained several characters, but with at least one
3165 whose value was between 128 and 255 caused PCRE to crash.
3168 Version 4.1 12-Mar-03
3169 ---------------------
3171 1. Compiling with gcc -pedantic found a couple of places where casts were
3172 needed, and a string in dftables.c that was longer than standard compilers are
3173 required to support.
3175 2. Compiling with Sun's compiler found a few more places where the code could
3176 be tidied up in order to avoid warnings.
3178 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3179 first of these names is deprecated in the latest Autoconf in favour of the name
3180 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3181 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3182 analogy I have changed it to CFLAGS_FOR_BUILD.
3184 4. Added -no-undefined to the linking command in the Makefile, because this is
3185 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3186 linking step for the pcreposix library.
3188 5. PCRE was failing to diagnose the case of two named groups with the same
3189 name.
3191 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3192 literal character that is needed in the subject for a match, and scans along to
3193 ensure that it is present before embarking on the full matching process. This
3194 saves time in cases of nested unlimited repeats that are never going to match.
3195 Problem: the scan can take a lot of time if the subject is very long (e.g.
3196 megabytes), thus penalizing straightforward matches. It is now done only if the
3197 amount of subject to be scanned is less than 1000 bytes.
3199 7. A lesser problem with the same optimization is that it was recording the
3200 first character of an anchored pattern as "needed", thus provoking a search
3201 right along the subject, even when the first match of the pattern was going to
3202 fail. The "needed" character is now not set for anchored patterns, unless it
3203 follows something in the pattern that is of non-fixed length. Thus, it still
3204 fulfils its original purpose of finding quick non-matches in cases of nested
3205 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3208 Version 4.0 17-Feb-03
3209 ---------------------
3211 1. If a comment in an extended regex that started immediately after a meta-item
3212 extended to the end of string, PCRE compiled incorrect data. This could lead to
3213 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3215 2. Moved to autoconf 2.53 and libtool 1.4.2.
3217 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3218 the special perltest8 script is no longer needed - all the tests can be run
3219 from a single perltest script.
3221 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3222 by \s. It has now been removed in PCRE. This means it isn't recognized as
3223 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3224 class [:space:] *does* include VT, thereby creating a mess.
3226 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3227 space and tab.
3229 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3230 its new features into the main test script, reducing the number of scripts.
3232 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3233 were backward compatible, and made the (?i) apply to the whole pattern, as if
3234 /i were given. Now it behaves more logically, and applies the option setting
3235 only to what follows. PCRE has been changed to follow suit. However, if it
3236 finds options settings right at the start of the pattern, it extracts them into
3237 the global options, as before. Thus, they show up in the info data.
3239 8. Added support for the \Q...\E escape sequence. Characters in between are
3240 treated as literals. This is slightly different from Perl in that $ and @ are
3241 also handled as literals inside the quotes. In Perl, they will cause variable
3242 interpolation. Note the following examples:
3244 Pattern PCRE matches Perl matches
3246 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3247 \Qabc\$xyz\E abc\$xyz abc\$xyz
3248 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3250 For compatibility with Perl, \Q...\E sequences are recognized inside character
3251 classes as well as outside them.
3253 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3254 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3255 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3256 signed/unsigned warnings.
3258 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3259 option for pcretest, so I've replaced it by a simple function that does just
3260 that job.
3262 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3263 "pcregrep -".
3265 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3266 Java package. This provides some syntactic sugar for simple cases of what my
3267 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3268 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3269 item, you can use this simplified notation. Note that only makes sense with
3270 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3271 greediness, whatever the setting of the PCRE_UNGREEDY option.
3273 13. A change of greediness default within a pattern was not taking effect at
3274 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3275 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3276 was abstracted outside.
3278 14. PCRE now supports the \G assertion. It is true when the current matching
3279 position is at the start point of the match. This differs from \A when the
3280 starting offset is non-zero. Used with the /g option of pcretest (or similar
3281 code), it works in the same way as it does for Perl's /g option. If all
3282 alternatives of a regex begin with \G, the expression is anchored to the start
3283 match position, and the "anchored" flag is set in the compiled expression.
3285 15. Some bugs concerning the handling of certain option changes within patterns
3286 have been fixed. These applied to options other than (?ims). For example,
3287 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3288 been the other way round. Some of this was related to change 7 above.
3290 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3291 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3292 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3293 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3295 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3296 mode. Unlike ".", it always matches newline, whatever the setting of
3297 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3298 assertions. Perl allows it, but it doesn't (in general) work because it can't
3299 calculate the length of the lookbehind. At least, that's the case for Perl
3300 5.8.0 - I've been told they are going to document that it doesn't work in
3301 future.
3303 18. Added an error diagnosis for escapes that PCRE does not support: these are
3304 \L, \l, \N, \P, \p, \U, \u, and \X.
3306 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3307 reading past the end of the pattern in cases such as /[abcd/.
3309 20. PCRE was getting more memory than necessary for patterns with classes that
3310 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3312 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3313 compiling PCRE for use with Virtual Pascal.
3315 22. Small fix to the Makefile to make it work properly if the build is done
3316 outside the source tree.
3318 23. Added a new extension: a condition to go with recursion. If a conditional
3319 subpattern starts with (?(R) the "true" branch is used if recursion has
3320 happened, whereas the "false" branch is used only at the top level.
3322 24. When there was a very long string of literal characters (over 255 bytes
3323 without UTF support, over 250 bytes with UTF support), the computation of how
3324 much memory was required could be incorrect, leading to segfaults or other
3325 strange effects.
3327 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3328 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3329 there was a subsequent back reference to those brackets. This meant that, for
3330 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3331 possible to check for precisely this case. All we can do is abandon the
3332 optimization if .* occurs inside capturing brackets when there are any back
3333 references whatsoever. (See below for a better fix that came later.)
3335 26. The handling of the optimization for finding the first character of a
3336 non-anchored pattern, and for finding a character that is required later in the
3337 match were failing in some cases. This didn't break the matching; it just
3338 failed to optimize when it could. The way this is done has been re-implemented.
3340 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3342 28. Added a new feature that provides some of the functionality that Perl
3343 provides with (?{...}). The facility is termed a "callout". The way it is done
3344 in PCRE is for the caller to provide an optional function, by setting
3345 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3346 global variable. By default it is unset, which disables all calling out. To get
3347 the function called, the regex must include (?C) at appropriate points. This
3348 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3349 This provides a means of identifying different callout points. When PCRE
3350 reaches such a point in the regex, if pcre_callout has been set, the external
3351 function is called. It is provided with data in a structure called
3352 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3353 matching continues; if it returns a non-zero value, the match at the current
3354 point fails. However, backtracking will occur if possible. [This was changed
3355 later and other features added - see item 49 below.]
3357 29. pcretest is upgraded to test the callout functionality. It provides a
3358 callout function that displays information. By default, it shows the start of
3359 the match and the current position in the text. There are some new data escapes
3360 to vary what happens:
3362 \C+ in addition, show current contents of captured substrings
3363 \C- do not supply a callout function
3364 \C!n return 1 when callout number n is reached
3365 \C!n!m return 1 when callout number n is reached for the mth time
3367 30. If pcregrep was called with the -l option and just a single file name, it
3368 output "<stdin>" if a match was found, instead of the file name.
3370 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3371 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3372 pcre_exec(). This saves a malloc/free per call. The default value of
3373 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3374 when configuring.
3376 32. The default maximum size of a compiled pattern is 64K. There have been a
3377 few cases of people hitting this limit. The code now uses macros to handle the
3378 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3379 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3380 configuring. Tests 2 and 5 work only with 2-byte links because they output
3381 debugging information about compiled patterns.
3383 33. Internal code re-arrangements:
3385 (a) Moved the debugging function for printing out a compiled regex into
3386 its own source file (printint.c) and used #include to pull it into
3387 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3388 separate copies.
3390 (b) Defined the list of op-code names for debugging as a macro in
3391 internal.h so that it is next to the definition of the opcodes.
3393 (c) Defined a table of op-code lengths for simpler skipping along compiled
3394 code. This is again a macro in internal.h so that it is next to the
3395 definition of the opcodes.
3397 34. Added support for recursive calls to individual subpatterns, along the
3398 lines of Robin Houston's patch (but implemented somewhat differently).
3400 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3401 allow it to read and process whole directories in Win32. This code was
3402 contributed by Lionel Fourquaux; it has not been tested by me.
3404 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3405 used to name a group. Names consist of alphanumerics and underscores, and must
3406 be unique. Back references use the syntax (?P=name) and recursive calls use
3407 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3408 numbers. The function pcre_fullinfo() can be used after compilation to extract
3409 a name/number map. There are three relevant calls:
3411 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3412 PCRE_INFO_NAMECOUNT yields the number of entries
3413 PCRE_INFO_NAMETABLE yields a pointer to the map.
3415 The map is a vector of fixed-size entries. The size of each entry depends on
3416 the length of the longest name used. The first two bytes of each entry are the
3417 group number, most significant byte first. There follows the corresponding
3418 name, zero terminated. The names are in alphabetical order.
3420 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3421 case instead of 255. Making it the same both with and without UTF-8 support
3422 means that the same test output works with both.
3424 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3425 calling malloc() with a zero argument.
3427 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3428 optimization. I've improved things by keeping a bitmap of backreferences with
3429 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3430 fact referenced, the optimization can be applied. It is unlikely that a
3431 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3432 the match to follow \n) will appear inside brackets with a number greater than
3433 31, but if it does, any back reference > 31 suppresses the optimization.
3435 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3436 of disabling numbered capturing parentheses. Any opening parenthesis that is
3437 not followed by ? behaves as if it were followed by ?: but named parentheses
3438 can still be used for capturing (and they will acquire numbers in the usual
3439 way).
3441 41. Redesigned the return codes from the match() function into yes/no/error so
3442 that errors can be passed back from deep inside the nested calls. A malloc
3443 failure while inside a recursive subpattern call now causes the
3444 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3446 42. It is now possible to set a limit on the number of times the match()
3447 function is called in a call to pcre_exec(). This facility makes it possible to
3448 limit the amount of recursion and backtracking, though not in a directly
3449 obvious way, because the match() function is used in a number of different
3450 circumstances. The count starts from zero for each position in the subject
3451 string (for non-anchored patterns). The default limit is, for compatibility, a
3452 large number, namely 10 000 000. You can change this in two ways:
3454 (a) When configuring PCRE before making, you can use --with-match-limit=n
3455 to set a default value for the compiled library.
3457 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3458 a different value is set. See 45 below.
3460 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3462 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3463 of things that can be changed at compile time. The first argument specifies
3464 what is wanted and the second points to where the information is to be placed.
3465 The current list of available information is:
3469 The output is an integer that is set to one if UTF-8 support is available;
3470 otherwise it is set to zero.
3474 The output is an integer that it set to the value of the code that is used for
3475 newline. It is either LF (10) or CR (13).
3479 The output is an integer that contains the number of bytes used for internal
3480 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3484 The output is an integer that contains the threshold above which the POSIX
3485 interface uses malloc() for output vectors. See item 31 above.
3489 The output is an unsigned integer that contains the default limit of the number
3490 of match() calls in a pcre_exec() execution. See 42 above.
3492 44. pcretest has been upgraded by the addition of the -C option. This causes it
3493 to extract all the available output from the new pcre_config() function, and to
3494 output it. The program then exits immediately.
3496 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3497 order to support additional features. One way would have been to define
3498 pcre_exec2() (for example) with extra arguments, but this would not have been
3499 extensible, and would also have required all calls to the original function to
3500 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3501 is used for passing in "extra" data from pcre_study().
3503 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3504 contains the following fields:
3506 flags a bitmap indicating which of the following fields are set
3507 study_data opaque data from pcre_study()
3508 match_limit a way of specifying a limit on match() calls for a specific
3509 call to pcre_exec()
3510 callout_data data for callouts (see 49 below)
3512 The flag bits are also defined in pcre.h, and are
3518 The pcre_study() function now returns one of these new pcre_extra blocks, with
3519 the actual study data pointed to by the study_data field, and the
3520 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3521 before. That is, this change is entirely upwards-compatible and requires no
3522 change to existing code.
3524 If you want to pass in additional data to pcre_exec(), you can either place it
3525 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3526 block.
3528 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3529 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3530 times with different match limits, until it finds the minimum value needed for
3531 pcre_exec() to complete. The value is then output. This can be instructive; for
3532 most simple matches the number is quite small, but for pathological cases it
3533 gets very large very quickly.
3535 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3536 returns the size of the data block pointed to by the study_data field in a
3537 pcre_extra block, that is, the value that was passed as the argument to
3538 pcre_malloc() when PCRE was getting memory in which to place the information
3539 created by pcre_study(). The fourth argument should point to a size_t variable.
3540 pcretest has been extended so that this information is shown after a successful
3541 pcre_study() call when information about the compiled regex is being displayed.
3543 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3544 because what follows is always an absolute path. (Later: it turns out that this
3545 is more than cosmetic for MinGW, because it doesn't like empty path
3546 components.)
3548 49. Some changes have been made to the callout feature (see 28 above):
3550 (i) A callout function now has three choices for what it returns:
3552 0 => success, carry on matching
3553 > 0 => failure at this point, but backtrack if possible
3554 < 0 => serious error, return this value from pcre_exec()
3556 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3557 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3558 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3559 use by callout functions. It will never be used by PCRE itself.
3561 (ii) The pcre_extra structure (see 45 above) has a void * field called
3562 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3563 pcre_callout_block structure has a field of the same name. The contents of
3564 the field passed in the pcre_extra structure are passed to the callout
3565 function in the corresponding field in the callout block. This makes it
3566 easier to use the same callout-containing regex from multiple threads. For
3567 testing, the pcretest program has a new data escape
3569 \C*n pass the number n (may be negative) as callout_data
3571 If the callout function in pcretest receives a non-zero value as
3572 callout_data, it returns that value.
3574 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3575 there were some redundant $(CFLAGS) in commands that are now specified as
3576 $(LINK), which already includes $(CFLAGS).
3578 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3579 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3580 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3581 one-byte characters throughout. Note that case-insensitive matching applies
3582 only to characters whose values are less than 256. PCRE doesn't support the
3583 notion of cases for higher-valued characters.
3585 (i) A character class whose characters are all within 0-255 is handled as
3586 a bit map, and the map is inverted for negative classes. Previously, a
3587 character > 255 always failed to match such a class; however it should
3588 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3590 (ii) A negated character class with a single character < 255 is coded as
3591 "not this character" (OP_NOT). This wasn't working properly when the test
3592 character was multibyte, either singly or repeated.
3594 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3595 mode, for example: \x{100}{2,3}.
3597 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3598 singly or repeated) now correctly test multibyte characters. However,
3599 PCRE doesn't recognize any characters with values greater than 255 as
3600 digits, spaces, or word characters. Such characters always match \D, \S,
3601 and \W, and never match \d, \s, or \w.
3603 (v) Classes may now contain characters and character ranges with values
3604 greater than 255. For example: [ab\x{100}-\x{400}].
3606 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3607 PCRE in UTF-8 mode.
3609 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3610 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3611 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3612 value.)
3614 53. The single man page has become too large. I have therefore split it up into
3615 a number of separate man pages. These also give rise to individual HTML pages;
3616 these are now put in a separate directory, and there is an index.html page that
3617 lists them all. Some hyperlinking between the pages has been installed.
3619 54. Added convenience functions for handling named capturing parentheses.
3621 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3622 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3623 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3624 are faulted.
3626 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3627 calling configure. These values are used when compiling the dftables.c program
3628 which is run to generate the source of the default character tables. They
3629 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3630 you will need to set these values.
3632 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3635 Version 3.9 02-Jan-02
3636 ---------------------
3638 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3640 2. If --disable-static was given, the building process failed when trying to
3641 build pcretest and pcregrep. (For some reason it was using libtool to compile
3642 them, which is not right, as they aren't part of the library.)
3645 Version 3.8 18-Dec-01
3646 ---------------------
3648 1. The experimental UTF-8 code was completely screwed up. It was packing the
3649 bytes in the wrong order. How dumb can you get?
3652 Version 3.7 29-Oct-01
3653 ---------------------
3655 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3656 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3657 this didn't happen under Solaris 8, where I normally test things.
3659 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3660 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3661 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3662 this fix an hour or so after the initial 3.7 release.)
3665 Version 3.6 23-Oct-01
3666 ---------------------
3668 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3669 offsets passed as NULL with zero offset count.
3671 2. The config.guess and config.sub files had not been updated when I moved to
3672 the latest autoconf.
3675 Version 3.5 15-Aug-01
3676 ---------------------
3678 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3679 had been forgotten.
3681 2. By using declared but undefined structures, we can avoid using "void"
3682 definitions in pcre.h while keeping the internal definitions of the structures
3683 private.
3685 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3686 user point of view, this means that both static and shared libraries are built
3687 by default, but this can be individually controlled. More of the work of
3688 handling this static/shared cases is now inside libtool instead of PCRE's make
3689 file.
3691 4. The pcretest utility is now installed along with pcregrep because it is
3692 useful for users (to test regexs) and by doing this, it automatically gets
3693 relinked by libtool. The documentation has been turned into a man page, so
3694 there are now .1, .txt, and .html versions in /doc.
3696 5. Upgrades to pcregrep:
3697 (i) Added long-form option names like gnu grep.
3698 (ii) Added --help to list all options with an explanatory phrase.
3699 (iii) Added -r, --recursive to recurse into sub-directories.
3700 (iv) Added -f, --file to read patterns from a file.
3702 6. pcre_exec() was referring to its "code" argument before testing that
3703 argument for NULL (and giving an error if it was NULL).
3705 7. Upgraded Makefile.in to allow for compiling in a different directory from
3706 the source directory.
3708 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3709 options bits, the pointer it was passed was to an int instead of to an unsigned
3710 long int. This mattered only on 64-bit systems.
3712 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3713 generated) instead of pcre.in, which it its source. Also made the same change
3714 in several of the .c files.
3716 10. A new release of gcc defines printf() as a macro, which broke pcretest
3717 because it had an ifdef in the middle of a string argument for printf(). Fixed
3718 by using separate calls to printf().
3720 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3721 script, to force use of CR or LF instead of \n in the source. On non-Unix
3722 systems, the value can be set in config.h.
3724 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3725 absolute limit. Changed the text of the error message to make this clear, and
3726 likewise updated the man page.
3728 13. The limit of 99 on the number of capturing subpatterns has been removed.
3729 The new limit is 65535, which I hope will not be a "real" limit.
3732 Version 3.4 22-Aug-00
3733 ---------------------
3735 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3737 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3740 Version 3.3 01-Aug-00
3741 ---------------------
3743 1. If an octal character was given, but the value was greater than \377, it
3744 was not getting masked to the least significant bits, as documented. This could
3745 lead to crashes in some systems.
3747 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3748 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3750 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3751 These just pass their arguments on to (pcre_free)(), but they are provided
3752 because some uses of PCRE bind it to non-C systems that can call its functions,
3753 but cannot call free() or pcre_free() directly.
3755 4. Add "make test" as a synonym for "make check". Corrected some comments in
3756 the Makefile.
3758 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3759 Makefile.
3761 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3762 command called pgrep for grepping around the active processes.
3764 7. Added the beginnings of support for UTF-8 character strings.
3766 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3767 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3768 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3769 out for the ar command.)
3772 Version 3.2 12-May-00
3773 ---------------------
3775 This is purely a bug fixing release.
3777 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3778 of ZA. This was just one example of several cases that could provoke this bug,
3779 which was introduced by change 9 of version 2.00. The code for breaking
3780 infinite loops after an iteration that matches an empty string was't working
3781 correctly.
3783 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3784 when matched against abbab (for example). After matching an empty string, it
3785 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3786 caused it to match further down the string than it should.
3788 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3789 was there because it doesn't seem to be needed, and it causes trouble on some
3790 systems, as it is not a Standard C header. It has been removed.
3792 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3793 were reported on the Macintosh. The changes were from
3795 while ((c = *(++ptr)) != 0 && c != '\n');
3796 to
3797 while ((c = *(++ptr)) != 0 && c != '\n') ;
3799 Totally extraordinary, but if that's what it takes...
3801 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3802 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3803 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3804 assumes the way PCRE uses memmove() (always moving upwards).
3806 6. PCRE is being used in one environment where strchr() is not available. There
3807 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3808 faster code anyway.
3811 Version 3.1 09-Feb-00
3812 ---------------------
3814 The only change in this release is the fixing of some bugs in Makefile.in for
3815 the "install" target:
3817 (1) It was failing to install pcreposix.h.
3819 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3822 Version 3.0 01-Feb-00
3823 ---------------------
3825 1. Add support for the /+ modifier to perltest (to output $` like it does in
3826 pcretest).
3828 2. Add support for the /g modifier to perltest.
3830 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3831 matches null strings.
3833 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3834 pattern. Perl treats empty patterns specially - it reuses the most recent
3835 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3836 effect.
3838 5. The POSIX interface was broken in that it was just handing over the POSIX
3839 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3840 required a bigger vector, with some working space on the end. This means that
3841 the POSIX wrapper now has to get and free some memory, and copy the results.
3843 6. Added some simple autoconf support, placing the test data and the
3844 documentation in separate directories, re-organizing some of the
3845 information files, and making it build pcre-config (a GNU standard). Also added
3846 libtool support for building PCRE as a shared library, which is now the
3847 default.
3849 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3850 09 are not valid octal constants. Single digits will be used for minor values
3851 less than 10.
3853 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3854 existing programs that set these in the POSIX interface can use PCRE without
3855 modification.
3857 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3858 return all that pcre_info() returns, plus additional data. The pcre_info()
3859 function is retained for compatibility, but is considered to be obsolete.
3861 10. Added experimental recursion feature (?R) to handle one common case that
3862 Perl 5.6 will be able to do with (?p{...}).
3864 11. Added support for POSIX character classes like [:alpha:], which Perl is
3865 adopting.
3868 Version 2.08 31-Aug-99
3869 ----------------------
3871 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3872 trying to match at the startoffset position, but instead was moving forward to
3873 the next newline as if a previous match had failed.
3875 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3876 and could get into a loop if a null string was matched other than at the start
3877 of the subject.
3879 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3880 be distinguished at compile time, and for completeness also added PCRE_DATE.
3882 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3883 in GnuWin32 environments.
3886 Version 2.07 29-Jul-99
3887 ----------------------
3889 1. The documentation is now supplied in plain text form and HTML as well as in
3890 the form of man page sources.
3892 2. C++ compilers don't like assigning (void *) values to other pointer types.
3893 In particular this affects malloc(). Although there is no problem in Standard
3894 C, I've put in casts to keep C++ compilers happy.
3896 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3897 should be (const char *).
3899 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3900 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3901 However, I haven't made this a standard facility. The documentation doesn't
3902 mention it, and the Makefile doesn't support it.
3904 5. The Makefile now contains an "install" target, with editable destinations at
3905 the top of the file. The pcretest program is not installed.
3907 6. pgrep -V now gives the PCRE version number and date.
3909 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3910 causing the entire string to be ignored, instead of just the last character.
3912 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3913 non-matching string, it can take a very, very long time, even for strings of
3914 quite modest length, because of the nested recursion. PCRE now does better in
3915 some of these cases. It does this by remembering the last required literal
3916 character in the pattern, and pre-searching the subject to ensure it is present
3917 before running the real match. In other words, it applies a heuristic to detect
3918 some types of certain failure quickly, and in the above example, if presented
3919 with a string that has no trailing " it gives "no match" very quickly.
3921 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3922 other alternatives are tried instead.
3925 Version 2.06 09-Jun-99
3926 ----------------------
3928 1. Change pcretest's output for amount of store used to show just the code
3929 space, because the remainder (the data block) varies in size between 32-bit and
3930 64-bit systems.
3932 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3933 start matching at. This allows lookbehinds to work when searching for multiple
3934 occurrences in a string.
3936 3. Added additional options to pcretest for testing multiple occurrences:
3938 /+ outputs the rest of the string that follows a match
3939 /g loops for multiple occurrences, using the new startoffset argument
3940 /G loops for multiple occurrences by passing an incremented pointer
3942 4. PCRE wasn't doing the "first character" optimization for patterns starting
3943 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3944 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3945 the letter 'x'. On long subject strings, this gives a significant speed-up.
3948 Version 2.05 21-Apr-99
3949 ----------------------
3951 1. Changed the type of magic_number from int to long int so that it works
3952 properly on 16-bit systems.
3954 2. Fixed a bug which caused patterns starting with .* not to work correctly
3955 when the subject string contained newline characters. PCRE was assuming
3956 anchoring for such patterns in all cases, which is not correct because .* will
3957 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3958 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3959 must be retried after every newline in the subject.
3962 Version 2.04 18-Feb-99
3963 ----------------------
3965 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3966 computation of the store needed to hold the pattern was incorrect (too large).
3967 If such patterns were nested a few deep, this could multiply and become a real
3968 problem.
3970 2. Added /M option to pcretest to show the memory requirement of a specific
3971 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3973 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3974 compiled in such a way that the backtracking after subsequent failure was
3975 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3976 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3979 Version 2.03 02-Feb-99
3980 ----------------------
3982 1. Fixed typo and small mistake in man page.
3984 2. Added 4th condition (GPL supersedes if conflict) and created separate
3985 LICENCE file containing the conditions.
3987 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3988 Perl, that is the internal \ allows the delimiter to be included in the
3989 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3990 the final delimiter, add \ to the end of the pattern (to test the error).
3992 4. Added the convenience functions for extracting substrings after a successful
3993 match. Updated pcretest to make it able to test these functions.
3996 Version 2.02 14-Jan-99
3997 ----------------------
3999 1. Initialized the working variables associated with each extraction so that
4000 their saving and restoring doesn't refer to uninitialized store.
4002 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4003 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4004 fix the problem.
4006 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4007 calls, and wasn't printing the correct value for compiling calls. Increased the
4008 default value of LOOPREPEAT, and the number of significant figures in the
4009 times.
4011 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4013 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4014 a building problem on Windows NT with a FAT file system.
4017 Version 2.01 21-Oct-98
4018 ----------------------
4020 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4021 to character tables built by pcre_maketables() in the current locale. If NULL
4022 is passed, the default tables are used.
4025 Version 2.00 24-Sep-98
4026 ----------------------
4028 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4029 it any more.
4031 2. Allow quantification of (?>) groups, and make it work correctly.
4033 3. The first character computation wasn't working for (?>) groups.
4035 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4036 end of the subject) and add 5.005's \z, which really does match only at the
4037 very end of the subject.
4039 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4041 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4042 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4043 localized options. All options to pcre_study() were also removed.
4045 7. Add other new features from 5.005:
4047 $(?<= positive lookbehind
4048 $(?<! negative lookbehind
4049 (?imsx-imsx) added the unsetting capability
4050 such a setting is global if at outer level; local otherwise
4051 (?imsx-imsx:) non-capturing groups with option setting
4052 (?(cond)re|re) conditional pattern matching
4054 A backreference to itself in a repeated group matches the previous
4055 captured string.
4057 8. General tidying up of studying (both automatic and via "study")
4058 consequential on the addition of new assertions.
4060 9. As in 5.005, unlimited repeated groups that could match an empty substring
4061 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4062 runtime if any iteration does actually match an empty substring.
4064 10. Include the RunTest script in the distribution.
4066 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4067 discrepancies, some of which were old and were also with respect to 5.004. They
4068 have now been fixed.
4071 Version 1.09 28-Apr-98
4072 ----------------------
4074 1. A negated single character class followed by a quantifier with a minimum
4075 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4076 program crashes, or just wrong answers. This did not apply to negated classes
4077 containing more than one character, or to minima other than one.
4080 Version 1.08 27-Mar-98
4081 ----------------------
4083 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4085 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4086 latter must appear before anything that relies on it in the pattern.
4089 Version 1.07 16-Feb-98
4090 ----------------------
4092 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4093 repeat of a potentially empty string).
4096 Version 1.06 23-Jan-98
4097 ----------------------
4099 1. Added Markus Oberhumer's little patches for C++.
4101 2. Literal strings longer than 255 characters were broken.
4104 Version 1.05 23-Dec-97
4105 ----------------------
4107 1. Negated character classes containing more than one character were failing if
4108 PCRE_CASELESS was set at run time.
4111 Version 1.04 19-Dec-97
4112 ----------------------
4114 1. Corrected the man page, where some "const" qualifiers had been omitted.
4116 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4117 input syntax.
4119 3. Fixed memory leak which occurred when a regex with back references was
4120 matched with an offsets vector that wasn't big enough. The temporary memory
4121 that is used in this case wasn't being freed if the match failed.
4123 4. Tidied pcretest to ensure it frees memory that it gets.
4125 5. Temporary memory was being obtained in the case where the passed offsets
4126 vector was exactly big enough.
4128 6. Corrected definition of offsetof() from change 5 below.
4130 7. I had screwed up change 6 below and broken the rules for the use of
4131 setjmp(). Now fixed.
4134 Version 1.03 18-Dec-97
4135 ----------------------
4137 1. A erroneous regex with a missing opening parenthesis was correctly
4138 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4139 on some systems.
4141 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4142 it was reported that one broken compiler failed on the former because "code" is
4143 also an independent variable.
4145 3. The erroneous regex a[]b caused an array overrun reference.
4147 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4148 fail on data ending with that character. (It was going on too far, and checking
4149 the next character, typically a binary zero.) This was specific to the
4150 optimized code for single-character negative classes.
4152 5. Added a contributed patch from the TIN world which does the following:
4154 + Add an undef for memmove, in case the the system defines a macro for it.
4156 + Add a definition of offsetof(), in case there isn't one. (I don't know
4157 the reason behind this - offsetof() is part of the ANSI standard - but
4158 it does no harm).
4160 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4161 most of the places where whitespace preceded '#'. I have given up and
4162 allowed the remaining 2 cases to be at the margin.
4164 + Rename some variables in pcre to eliminate shadowing. This seems very
4165 pedantic, but does no harm, of course.
4167 6. Moved the call to setjmp() into its own function, to get rid of warnings
4168 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4170 7. Constructs such as \d{8,} were compiling into the equivalent of
4171 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4172 outcome, but in this particular case used more store than had been allocated,
4173 which caused the bug to be discovered because it threw up an internal error.
4175 8. The debugging code in both pcre and pcretest for outputting the compiled
4176 form of a regex was going wrong in the case of back references followed by
4177 curly-bracketed repeats.
4180 Version 1.02 12-Dec-97
4181 ----------------------
4183 1. Typos in pcre.3 and comments in the source fixed.
4185 2. Applied a contributed patch to get rid of places where it used to remove
4186 'const' from variables, and fixed some signed/unsigned and uninitialized
4187 variable warnings.
4189 3. Added the "runtest" target to Makefile.
4191 4. Set default compiler flag to -O2 rather than just -O.
4194 Version 1.01 19-Nov-97
4195 ----------------------
4197 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4198 like /([ab]*)*/, that is, for classes with more than one character in them.
4200 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4201 as /((?>a*))*/ (a PCRE_EXTRA facility).
4204 Version 1.00 18-Nov-97
4205 ----------------------
4207 1. Added compile-time macros to support systems such as SunOS4 which don't have
4208 memmove() or strerror() but have other things that can be used instead.
4210 2. Arranged that "make clean" removes the executables.
4213 Version 0.99 27-Oct-97
4214 ----------------------
4216 1. Fixed bug in code for optimizing classes with only one character. It was
4217 initializing a 32-byte map regardless, which could cause it to run off the end
4218 of the memory it had got.
4220 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4223 Version 0.98 22-Oct-97
4224 ----------------------
4226 1. Fixed bug in code for handling temporary memory usage when there are more
4227 back references than supplied space in the ovector. This could cause segfaults.
4230 Version 0.97 21-Oct-97
4231 ----------------------
4233 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4235 2. Optimized negated single characters not to use a bit map.
4237 3. Brought error texts together as macro definitions; clarified some of them;
4238 fixed one that was wrong - it said "range out of order" when it meant "invalid
4239 escape sequence".
4241 4. Changed some char * arguments to const char *.
4243 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4245 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4246 pcretest.
4249 Version 0.96 16-Oct-97
4250 ----------------------
4252 1. Added a simple "pgrep" utility to the distribution.
4254 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4255 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4256 where "ddd" means "one or more decimal digits".
4258 3. Fixed serious bug. If a pattern had a back reference, but the call to
4259 pcre_exec() didn't supply a large enough ovector to record the related
4260 identifying subpattern, the match always failed. PCRE now remembers the number
4261 of the largest back reference, and gets some temporary memory in which to save
4262 the offsets during matching if necessary, in order to ensure that
4263 backreferences always work.
4265 4. Increased the compatibility with Perl in a number of ways:
4267 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4268 to request this handling. The option can be set at compile or exec time.
4270 (b) $ matches before a terminating newline by default; an option
4271 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4272 mode). The option can be set at compile or exec time.
4274 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4275 the same as Perl's. If the decimal number it represents is less than 10
4276 or there aren't that many previous left capturing parentheses, an octal
4277 escape is read. Inside a character class, it's always an octal escape,
4278 even if it is a single digit.
4280 (d) An escaped but undefined alphabetic character is taken as a literal,
4281 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4282 escapes.
4284 (e) {0} is now permitted. (The previous item is removed from the compiled
4285 pattern).
4287 5. Changed all the names of code files so that the basic parts are no longer
4288 than 10 characters, and abolished the teeny "globals.c" file.
4290 6. Changed the handling of character classes; they are now done with a 32-byte
4291 bit map always.
4293 7. Added the -d and /D options to pcretest to make it possible to look at the
4294 internals of compilation without having to recompile pcre.
4297 Version 0.95 23-Sep-97
4298 ----------------------
4300 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4301 \x20 at the start of a run of normal characters. These were being treated as
4302 real characters, instead of the source characters being re-checked.
4305 Version 0.94 18-Sep-97
4306 ----------------------
4308 1. The functions are now thread-safe, with the caveat that the global variables
4309 containing pointers to malloc() and free() or alternative functions are the
4310 same for all threads.
4312 2. Get pcre_study() to generate a bitmap of initial characters for non-
4313 anchored patterns when this is possible, and use it if passed to pcre_exec().
4316 Version 0.93 15-Sep-97
4317 ----------------------
4319 1. /(b)|(:+)/ was computing an incorrect first character.
4321 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4322 but not actually doing anything yet.
4324 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4325 as Perl does (e.g. [-az] or [az-]).
4327 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4328 all possible positions.
4330 5. Split up into different modules to avoid including unneeded functions in a
4331 compiled binary. However, compile and exec are still in one module. The "study"
4332 function is split off.
4334 6. The character tables are now in a separate module whose source is generated
4335 by an auxiliary program - but can then be edited by hand if required. There are
4336 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4337 toupper() in the code.
4339 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4340 make them global. Abolish the function for setting them, as the caller can now
4341 set them directly.
4344 Version 0.92 11-Sep-97
4345 ----------------------
4347 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4348 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4350 2. Caseless matching was not working in character classes if the characters in
4351 the pattern were in upper case.
4353 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4355 4. Make PCRE_ANCHORED public and accept as a compile option.
4357 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4358 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4359 pass them.
4361 6. Give an error if bad option bits passed at compile or run time.
4363 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4364 pcretest to cause it to pass that flag.
4366 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4367 options, and the first character, if set.
4369 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4372 Version 0.91 10-Sep-97
4373 ----------------------
4375 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4376 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4378 2. PCRE was looping on encountering an indefinitely repeated back reference to
4379 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4380 Perl does - treats the match as successful.
4382 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5