/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 620 - (show annotations) (download)
Sun Jul 17 13:53:14 2011 UTC (2 years, 9 months ago) by ph10
File size: 206399 byte(s)
Fix mutual recursion minimum calculation in study bug.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.13 30-Apr-2011
5 ------------------------
6
7 1. The Unicode data tables have been updated to Unicode 6.0.0.
8
9 2. Two minor typos in pcre_internal.h have been fixed.
10
11 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
14
15 4. There were a number of related bugs in the code for matching backrefences
16 caselessly in UTF-8 mode when codes for the characters concerned were
17 different numbers of bytes. For example, U+023A and U+2C65 are an upper
18 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21 2-byte code at the end of the subject (it thought there wasn't enough data
22 left).
23
24 5. Comprehensive information about what went wrong is now returned by
25 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26 as the output vector has at least 2 elements. The offset of the start of
27 the failing character and a reason code are placed in the vector.
28
29 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
30 now returned is for the first byte of the failing character, instead of the
31 last byte inspected. This is an incompatible change, but I hope it is small
32 enough not to be a problem. It makes the returned offset consistent with
33 pcre_exec() and pcre_dfa_exec().
34
35 7. pcretest now gives a text phrase as well as the error number when
36 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37 failure, the offset and reason code are output.
38
39 8. When \R was used with a maximizing quantifier it failed to skip backwards
40 over a \r\n pair if the subsequent match failed. Instead, it just skipped
41 back over a single character (\n). This seems wrong (because it treated the
42 two characters as a single entity when going forwards), conflicts with the
43 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
44 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
45 has been changed.
46
47 9. Some internal refactoring has changed the processing so that the handling
48 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
49 time (the PCRE_DOTALL option was changed this way some time ago: version
50 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
51 which was always a bit of a fudge. It also means that there is one less
52 argument for the match() function, which reduces its stack requirements
53 slightly. This change also fixes an incompatibility with Perl: the pattern
54 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
55
56 10. More internal refactoring has drastically reduced the number of recursive
57 calls to match() for possessively repeated groups such as (abc)++ when
58 using pcre_exec().
59
60 11. While implementing 10, a number of bugs in the handling of groups were
61 discovered and fixed:
62
63 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
64 (a|)*(?1) gave a compile-time internal error.
65 ((a|)+)+ did not notice that the outer group could match an empty string.
66 (^a|^)+ was not marked as anchored.
67 (.*a|.*)+ was not marked as matching at start or after a newline.
68
69 12. Yet more internal refactoring has removed another argument from the match()
70 function. Special calls to this function are now indicated by setting a
71 value in a variable in the "match data" data block.
72
73 13. Be more explicit in pcre_study() instead of relying on "default" for
74 opcodes that mean there is no starting character; this means that when new
75 ones are added and accidentally left out of pcre_study(), testing should
76 pick them up.
77
78 14. The -s option of pcretest has been documented for ages as being an old
79 synonym of -m (show memory usage). I have changed it to mean "force study
80 for every regex", that is, assume /S for every regex. This is similar to -i
81 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
82 using it. It makes it easier to run collections of tests with and without
83 study enabled, and thereby test pcre_study() more easily. All the standard
84 tests are now run with and without -s (but some patterns can be marked as
85 "never study" - see 20 below).
86
87 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
88 restoration of the capturing data to the outer values was not happening
89 correctly.
90
91 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
92 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
93 pattern had matched an empty string, and so incorrectly returned a no
94 match.
95
96 17. There was optimizing code for the last branch of non-capturing parentheses,
97 and also for the obeyed branch of a conditional subexpression, which used
98 tail recursion to cut down on stack usage. Unfortunately, not that there is
99 the possibility of (*THEN) occurring in these branches, tail recursion is
100 no longer possible because the return has to be checked for (*THEN). These
101 two optimizations have therefore been removed.
102
103 18. If a pattern containing \R was studied, it was assumed that \R always
104 matched two bytes, thus causing the minimum subject length to be
105 incorrectly computed because \R can also match just one byte.
106
107 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
108 was incorrectly computed.
109
110 20. If /S is present twice on a test pattern in pcretest input, it *disables*
111 studying, thereby overriding the use of -s on the command line. This is
112 necessary for one or two tests to keep the output identical in both cases.
113
114 21. When (*ACCEPT) was used in an assertion that matched an empty string and
115 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
116
117 22. When an atomic group that contained a capturing parenthesis was
118 successfully matched, but the branch in which it appeared failed, the
119 capturing was not being forgotten if a higher numbered group was later
120 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
121 group 1 to "a", when in fact it should be unset. This applied to multi-
122 branched capturing and non-capturing groups, repeated or not, and also to
123 positive assertions (capturing in negative assertions is not well defined
124 in PCRE) and also to nested atomic groups.
125
126 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
127 subject after a captured substring (to make it easier to tell which of a
128 number of identical substrings has been captured).
129
130 24. The way atomic groups are processed by pcre_exec() has been changed so that
131 if they are repeated, backtracking one repetition now resets captured
132 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
133 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
134 "aaa". Previously, it would have been "a". As part of this code
135 refactoring, the way recursive calls are handled has also been changed.
136
137 24. If an assertion condition captured any substrings, they were not passed
138 back unless some other capturing happened later. For example, if
139 (?(?=(a))a) was matched against "a", no capturing was returned.
140
141 25. When studying a pattern that contained subroutine calls or assertions,
142 the code for finding the minimum length of a possible match was handling
143 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
144 group 1 called group 2 while simultaneously a separate group 2 called group
145 1). A stack overflow occurred in this case. I have fixed this by limiting
146 the recursion depth to 10.
147
148
149 Version 8.12 15-Jan-2011
150 ------------------------
151
152 1. Fixed some typos in the markup of the man pages, and wrote a script that
153 checks for such things as part of the documentation building process.
154
155 2. On a big-endian 64-bit system, pcregrep did not correctly process the
156 --match-limit and --recursion-limit options (added for 8.11). In
157 particular, this made one of the standard tests fail. (The integer value
158 went into the wrong half of a long int.)
159
160 3. If the --colour option was given to pcregrep with -v (invert match), it
161 did strange things, either producing crazy output, or crashing. It should,
162 of course, ignore a request for colour when reporting lines that do not
163 match.
164
165 4. Another pcregrep bug caused similar problems if --colour was specified with
166 -M (multiline) and the pattern match finished with a line ending.
167
168 5. In pcregrep, when a pattern that ended with a literal newline sequence was
169 matched in multiline mode, the following line was shown as part of the
170 match. This seems wrong, so I have changed it.
171
172 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
173 the check for further matches in the same line (so they could be coloured)
174 to overrun the end of the current line. If another match was found, it was
175 incorrectly shown (and then shown again when found in the next line).
176
177 7. If pcregrep was compiled under Windows, there was a reference to the
178 function pcregrep_exit() before it was defined. I am assuming this was
179 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
180 reported by a user. I've moved the definition above the reference.
181
182
183 Version 8.11 10-Dec-2010
184 ------------------------
185
186 1. (*THEN) was not working properly if there were untried alternatives prior
187 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
188 backtracked to try for "b" instead of moving to the next alternative branch
189 at the same level (in this case, to look for "c"). The Perl documentation
190 is clear that when (*THEN) is backtracked onto, it goes to the "next
191 alternative in the innermost enclosing group".
192
193 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
194 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
195 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
196 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
197 (*THEN).
198
199 3. If \s appeared in a character class, it removed the VT character from
200 the class, even if it had been included by some previous item, for example
201 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
202 of \s, but is part of the POSIX "space" class.)
203
204 4. A partial match never returns an empty string (because you can always
205 match an empty string at the end of the subject); however the checking for
206 an empty string was starting at the "start of match" point. This has been
207 changed to the "earliest inspected character" point, because the returned
208 data for a partial match starts at this character. This means that, for
209 example, /(?<=abc)def/ gives a partial match for the subject "abc"
210 (previously it gave "no match").
211
212 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
213 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
214 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
215 has an implication that the given string is incomplete (because a partial
216 match is preferred over a full match). For this reason, these items now
217 give a partial match in this situation. [Aside: previously, the one case
218 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
219 match rather than a full match, which was wrong by the old rules, but is
220 now correct.]
221
222 6. There was a bug in the handling of #-introduced comments, recognized when
223 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
224 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
225 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
226 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
227 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
228 places in pcre_compile().
229
230 7. Related to (6) above, when pcre_compile() was skipping #-introduced
231 comments when looking ahead for named forward references to subpatterns,
232 the only newline sequence it recognized was NL. It now handles newlines
233 according to the set newline convention.
234
235 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
236 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
237 cater for a lack of strerror(). These oversights have been fixed.
238
239 9. Added --match-limit and --recursion-limit to pcregrep.
240
241 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
242
243 11. When the -o option was used, pcregrep was setting a return code of 1, even
244 when matches were found, and --line-buffered was not being honoured.
245
246 12. Added an optional parentheses number to the -o and --only-matching options
247 of pcregrep.
248
249 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
250 can match an empty string. The code to do it in pcretest and pcredemo
251 needed fixing:
252
253 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
254 only one byte after an empty string match just before CRLF (this case
255 just got forgotten; "any" and "anycrlf" were OK).
256
257 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
258 mode when an empty string match preceded an ASCII character followed by
259 a non-ASCII character. (The code for advancing by one character rather
260 than one byte was nonsense.)
261
262 (c) The pcredemo.c sample program did not have any code at all to handle
263 the cases when CRLF is a valid newline sequence.
264
265 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
266 as a starting offset was within the subject string. There is now a new
267 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
268 negative or greater than the length of the string. In order to test this,
269 pcretest is extended to allow the setting of negative starting offsets.
270
271 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
272 starting offset points to the beginning of a UTF-8 character was
273 unnecessarily clumsy. I tidied it up.
274
275 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
276 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
277
278 17. Nobody had reported that the --include_dir option, which was added in
279 release 7.7 should have been called --include-dir (hyphen, not underscore)
280 for compatibility with GNU grep. I have changed it to --include-dir, but
281 left --include_dir as an undocumented synonym, and the same for
282 --exclude-dir, though that is not available in GNU grep, at least as of
283 release 2.5.4.
284
285 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
286 characters from a string of bytes) have been redefined so as not to use
287 loops, in order to improve performance in some environments. At the same
288 time, I abstracted some of the common code into auxiliary macros to save
289 repetition (this should not affect the compiled code).
290
291 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
292 compile-time error is now given if \c is not followed by an ASCII
293 character, that is, a byte less than 128. (In EBCDIC mode, the code is
294 different, and any byte value is allowed.)
295
296 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
297 START_OPTIMIZE option, which is now allowed at compile time - but just
298 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
299 to pcregrep and other applications that have no direct access to PCRE
300 options. The new /Y option in pcretest sets this option when calling
301 pcre_compile().
302
303 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
304 back references. Groups containing recursive back references were forced to
305 be atomic by that change, but in the case of named groups, the amount of
306 memory required was incorrectly computed, leading to "Failed: internal
307 error: code overflow". This has been fixed.
308
309 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
310 pcretest.c, to avoid build problems in some Borland environments.
311
312
313 Version 8.10 25-Jun-2010
314 ------------------------
315
316 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
317 THEN.
318
319 2. (*ACCEPT) was not working when inside an atomic group.
320
321 3. Inside a character class, \B is treated as a literal by default, but
322 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
323 causes the error). The code is unchanged, but I tidied the documentation.
324
325 4. Inside a character class, PCRE always treated \R and \X as literals,
326 whereas Perl faults them if its -w option is set. I have changed PCRE so
327 that it faults them when PCRE_EXTRA is set.
328
329 5. Added support for \N, which always matches any character other than
330 newline. (It is the same as "." when PCRE_DOTALL is not set.)
331
332 6. When compiling pcregrep with newer versions of gcc which may have
333 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
334 declared with attribute warn_unused_result" were given. Just casting the
335 result to (void) does not stop the warnings; a more elaborate fudge is
336 needed. I've used a macro to implement this.
337
338 7. Minor change to pcretest.c to avoid a compiler warning.
339
340 8. Added four artifical Unicode properties to help with an option to make
341 \s etc use properties (see next item). The new properties are: Xan
342 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
343
344 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
345 use Unicode properties. (*UCP) at the start of a pattern can be used to set
346 this option. Modified pcretest to add /W to test this facility. Added
347 REG_UCP to make it available via the POSIX interface.
348
349 10. Added --line-buffered to pcregrep.
350
351 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
352 studied, and the match started with a letter with a code point greater than
353 127 whose first byte was different to the first byte of the other case of
354 the letter, the other case of this starting letter was not recognized
355 (#976).
356
357 12. If a pattern that was studied started with a repeated Unicode property
358 test, for example, \p{Nd}+, there was the theoretical possibility of
359 setting up an incorrect bitmap of starting bytes, but fortunately it could
360 not have actually happened in practice until change 8 above was made (it
361 added property types that matched character-matching opcodes).
362
363 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
364 possible starting bytes for non-anchored patterns.
365
366 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
367 \R, and also a number of cases that involve Unicode properties, both
368 explicit and implicit when PCRE_UCP is set.
369
370 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
371 input, it could crash or give wrong results if characters with values
372 greater than 0xc0 were present in the subject string. (Detail: it assumed
373 UTF-8 input when processing these items.)
374
375 16. Added a lot of (int) casts to avoid compiler warnings in systems where
376 size_t is 64-bit (#991).
377
378 17. Added a check for running out of memory when PCRE is compiled with
379 --disable-stack-for-recursion (#990).
380
381 18. If the last data line in a file for pcretest does not have a newline on
382 the end, a newline was missing in the output.
383
384 19. The default pcre_chartables.c file recognizes only ASCII characters (values
385 less than 128) in its various bitmaps. However, there is a facility for
386 generating tables according to the current locale when PCRE is compiled. It
387 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
388 characters, are recognized by isspace() and therefore were getting set in
389 these tables, and indeed these tables seem to approximate to ISO 8859. This
390 caused a problem in UTF-8 mode when pcre_study() was used to create a list
391 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
392 which of course cannot start UTF-8 characters. I have changed the code so
393 that only real ASCII characters (less than 128) and the correct starting
394 bytes for UTF-8 encodings are set for characters greater than 127 when in
395 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
396 altogether.)
397
398 20. Added the /T option to pcretest so as to be able to run tests with non-
399 standard character tables, thus making it possible to include the tests
400 used for 19 above in the standard set of tests.
401
402 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
403 reference to a subpattern the other side of a comment that contains an
404 opening parenthesis caused either an internal compiling error, or a
405 reference to the wrong subpattern.
406
407
408 Version 8.02 19-Mar-2010
409 ------------------------
410
411 1. The Unicode data tables have been updated to Unicode 5.2.0.
412
413 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
414 configured.
415
416 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
417 original author of that file, following a query about its status.
418
419 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
420 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
421
422 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
423 quantifier applied to a forward-referencing subroutine call, could compile
424 incorrect code or give the error "internal error: previously-checked
425 referenced subpattern not found".
426
427 6. Both MS Visual Studio and Symbian OS have problems with initializing
428 variables to point to external functions. For these systems, therefore,
429 pcre_malloc etc. are now initialized to local functions that call the
430 relevant global functions.
431
432 7. There were two entries missing in the vectors called coptable and poptable
433 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
434 I've fixed the data, and added a kludgy way of testing at compile time that
435 the lengths are correct (equal to the number of opcodes).
436
437 8. Following on from 7, I added a similar kludge to check the length of the
438 eint vector in pcreposix.c.
439
440 9. Error texts for pcre_compile() are held as one long string to avoid too
441 much relocation at load time. To find a text, the string is searched,
442 counting zeros. There was no check for running off the end of the string,
443 which could happen if a new error number was added without updating the
444 string.
445
446 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
447
448 11. \K was not working if it appeared in an atomic group or in a group that
449 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
450 \K is "not well defined" if used in an assertion. PCRE now accepts it if
451 the assertion is positive, but not if it is negative.
452
453 12. Change 11 fortuitously reduced the size of the stack frame used in the
454 "match()" function of pcre_exec.c by one pointer. Forthcoming
455 implementation of support for (*MARK) will need an extra pointer on the
456 stack; I have reserved it now, so that the stack frame size does not
457 decrease.
458
459 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
460 item in branch that calls a recursion is a subroutine call - as in the
461 second branch in the above example - was incorrectly given the compile-
462 time error "recursive call could loop indefinitely" because pcre_compile()
463 was not correctly checking the subroutine for matching a non-empty string.
464
465 14. The checks for overrunning compiling workspace could trigger after an
466 overrun had occurred. This is a "should never occur" error, but it can be
467 triggered by pathological patterns such as hundreds of nested parentheses.
468 The checks now trigger 100 bytes before the end of the workspace.
469
470 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
471
472
473 Version 8.01 19-Jan-2010
474 ------------------------
475
476 1. If a pattern contained a conditional subpattern with only one branch (in
477 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
478 computed the wrong minimum data length (which is of course zero for such
479 subpatterns). This could cause incorrect "no match" results.
480
481 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
482 the pattern is reset in the first branch, pcre_compile() failed with
483 "internal error: code overflow at offset...". This happened only when
484 the reset was to the original external option setting. (An optimization
485 abstracts leading options settings into an external setting, which was the
486 cause of this.)
487
488 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
489 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
490 assertion pattern did not match (meaning that the assertion was true), it
491 was incorrectly treated as false if the SKIP had been reached during the
492 matching. This also applied to assertions used as conditions.
493
494 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
495 assertion subpattern, including such a pattern used as a condition,
496 unpredictable results occurred, instead of the error return
497 PCRE_ERROR_DFA_UITEM.
498
499 5. The C++ GlobalReplace function was not working like Perl for the special
500 situation when an empty string is matched. It now does the fancy magic
501 stuff that is necessary.
502
503 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
504 removed. (These were left over from very, very early versions of PCRE.)
505
506 7. Some cosmetic changes to the code to make life easier when compiling it
507 as part of something else:
508
509 (a) Change DEBUG to PCRE_DEBUG.
510
511 (b) In pcre_compile(), rename the member of the "branch_chain" structure
512 called "current" as "current_branch", to prevent a collision with the
513 Linux macro when compiled as a kernel module.
514
515 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
516 prevent a collision with the Linux macro when compiled as a kernel
517 module.
518
519 8. In pcre_compile() there are some checks for integer overflows that used to
520 cast potentially large values to (double). This has been changed to that
521 when building, a check for int64_t is made, and if it is found, it is used
522 instead, thus avoiding the use of floating point arithmetic. (There is no
523 other use of FP in PCRE.) If int64_t is not found, the fallback is to
524 double.
525
526 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
527 2005 (difference between two addresses compared to an unsigned value).
528
529 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
530 custom one, because of the following reported problem in Windows:
531
532 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
533 under Win32.
534 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
535 therefore missing the function definition.
536 - The compiler thus generates a "C" signature for the test function.
537 - The linker fails to find the "C" function.
538 - PCRE fails to configure if asked to do so against libbz2.
539
540 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
541 messages were output:
542
543 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
544 rerunning libtoolize, to keep the correct libtool macros in-tree.
545 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
546
547 I have done both of these things.
548
549 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
550 most of the time, it *can* run out if it is given a pattern that contains a
551 runaway infinite recursion. I updated the discussion in the pcrestack man
552 page.
553
554 13. Now that we have gone to the x.xx style of version numbers, the minor
555 version may start with zero. Using 08 or 09 is a bad idea because users
556 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
557 interpreted as invalid octal numbers. I've updated the previous comment in
558 configure.ac, and also added a check that gives an error if 08 or 09 are
559 used.
560
561 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
562 causing partial matching to fail when the end of the subject matched \W
563 in a UTF-8 pattern where \W was quantified with a minimum of 3.
564
565 15. There were some discrepancies between the declarations in pcre_internal.h
566 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
567 their definitions. The declarations used "const uschar *" and the
568 definitions used USPTR. Even though USPTR is normally defined as "const
569 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
570 reported that: "This difference in casting confuses some C++ compilers, for
571 example, SunCC recognizes above declarations as different functions and
572 generates broken code for hbpcre." I have changed the declarations to use
573 USPTR.
574
575 16. GNU libtool is named differently on some systems. The autogen.sh script now
576 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
577 (FreeBSD).
578
579 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
580 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
581 comment: "Figure out how to create a longlong from a string: strtoll and
582 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
583 instance, but it only takes 2 args instead of 3!"
584
585 18. A subtle bug concerned with back references has been fixed by a change of
586 specification, with a corresponding code fix. A pattern such as
587 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
588 refers, was giving matches when it shouldn't. For example, xa=xaaa would
589 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
590 same bug. Such groups have to be quantified to be useful, or contained
591 inside another quantified group. (If there's no repetition, the reference
592 can never match.) The problem arises because, having left the group and
593 moved on to the rest of the pattern, a later failure that backtracks into
594 the group uses the captured value from the final iteration of the group
595 rather than the correct earlier one. I have fixed this in PCRE by forcing
596 any group that contains a reference to itself to be an atomic group; that
597 is, there cannot be any backtracking into it once it has completed. This is
598 similar to recursive and subroutine calls.
599
600
601 Version 8.00 19-Oct-09
602 ----------------------
603
604 1. The table for translating pcre_compile() error codes into POSIX error codes
605 was out-of-date, and there was no check on the pcre_compile() error code
606 being within the table. This could lead to an OK return being given in
607 error.
608
609 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
610 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
611 in a Windows environment.
612
613 3. The pcregrep --count option prints the count for each file even when it is
614 zero, as does GNU grep. However, pcregrep was also printing all files when
615 --files-with-matches was added. Now, when both options are given, it prints
616 counts only for those files that have at least one match. (GNU grep just
617 prints the file name in this circumstance, but including the count seems
618 more useful - otherwise, why use --count?) Also ensured that the
619 combination -clh just lists non-zero counts, with no names.
620
621 4. The long form of the pcregrep -F option was incorrectly implemented as
622 --fixed_strings instead of --fixed-strings. This is an incompatible change,
623 but it seems right to fix it, and I didn't think it was worth preserving
624 the old behaviour.
625
626 5. The command line items --regex=pattern and --regexp=pattern were not
627 recognized by pcregrep, which required --regex pattern or --regexp pattern
628 (with a space rather than an '='). The man page documented the '=' forms,
629 which are compatible with GNU grep; these now work.
630
631 6. No libpcreposix.pc file was created for pkg-config; there was just
632 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
633
634 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
635 when UCP support is not needed, by modifying the Python script that
636 generates it from Unicode data files. This should not matter if the module
637 is correctly used as a library, but I received one complaint about 50K of
638 unwanted data. My guess is that the person linked everything into his
639 program rather than using a library. Anyway, it does no harm.
640
641 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
642 was a minimum greater than 1 for a wide character in a possessive
643 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
644 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
645 character. Chaos in the form of incorrect output or a compiling loop could
646 result.
647
648 9. The restrictions on what a pattern can contain when partial matching is
649 requested for pcre_exec() have been removed. All patterns can now be
650 partially matched by this function. In addition, if there are at least two
651 slots in the offset vector, the offset of the earliest inspected character
652 for the match and the offset of the end of the subject are set in them when
653 PCRE_ERROR_PARTIAL is returned.
654
655 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
656 synonymous with PCRE_PARTIAL, for backwards compatibility, and
657 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
658 and may be more useful for multi-segment matching.
659
660 11. Partial matching with pcre_exec() is now more intuitive. A partial match
661 used to be given if ever the end of the subject was reached; now it is
662 given only if matching could not proceed because another character was
663 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
664 string "Z", which now yields "no match" instead of "partial match". In the
665 case of pcre_dfa_exec(), "no match" is given if every matching path for the
666 final character ended with (*FAIL).
667
668 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
669 if the pattern had a "must contain" character that was already found in the
670 earlier partial match, unless partial matching was again requested. For
671 example, with the pattern /dog.(body)?/, the "must contain" character is
672 "g". If the first part-match was for the string "dog", restarting with
673 "sbody" failed. This bug has been fixed.
674
675 13. The string returned by pcre_dfa_exec() after a partial match has been
676 changed so that it starts at the first inspected character rather than the
677 first character of the match. This makes a difference only if the pattern
678 starts with a lookbehind assertion or \b or \B (\K is not supported by
679 pcre_dfa_exec()). It's an incompatible change, but it makes the two
680 matching functions compatible, and I think it's the right thing to do.
681
682 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
683 so that the demonstration program is easily available in environments where
684 PCRE has not been installed from source.
685
686 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
687 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
688 library.
689
690 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
691 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
692 is not the first non-POSIX option to be added. Clearly some people find
693 these options useful.
694
695 17. If a caller to the POSIX matching function regexec() passes a non-zero
696 value for nmatch with a NULL value for pmatch, the value of
697 nmatch is forced to zero.
698
699 18. RunGrepTest did not have a test for the availability of the -u option of
700 the diff command, as RunTest does. It now checks in the same way as
701 RunTest, and also checks for the -b option.
702
703 19. If an odd number of negated classes containing just a single character
704 interposed, within parentheses, between a forward reference to a named
705 subpattern and the definition of the subpattern, compilation crashed with
706 an internal error, complaining that it could not find the referenced
707 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
708 [The bug was that it was starting one character too far in when skipping
709 over the character class, thus treating the ] as data rather than
710 terminating the class. This meant it could skip too much.]
711
712 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
713 /g option in pcretest when the pattern contains \K, which makes it possible
714 to have an empty string match not at the start, even when the pattern is
715 anchored. Updated pcretest and pcredemo to use this option.
716
717 21. If the maximum number of capturing subpatterns in a recursion was greater
718 than the maximum at the outer level, the higher number was returned, but
719 with unset values at the outer level. The correct (outer level) value is
720 now given.
721
722 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
723 PCRE did not set those parentheses (unlike Perl). I have now found a way to
724 make it do so. The string so far is captured, making this feature
725 compatible with Perl.
726
727 23. The tests have been re-organized, adding tests 11 and 12, to make it
728 possible to check the Perl 5.10 features against Perl 5.10.
729
730 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
731 pattern matches a fixed length string. PCRE did not allow this; now it
732 does. Neither allows recursion.
733
734 25. I finally figured out how to implement a request to provide the minimum
735 length of subject string that was needed in order to match a given pattern.
736 (It was back references and recursion that I had previously got hung up
737 on.) This code has now been added to pcre_study(); it finds a lower bound
738 to the length of subject needed. It is not necessarily the greatest lower
739 bound, but using it to avoid searching strings that are too short does give
740 some useful speed-ups. The value is available to calling programs via
741 pcre_fullinfo().
742
743 26. While implementing 25, I discovered to my embarrassment that pcretest had
744 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
745 study optimizations had never been tested with that matching function.
746 Oops. What is worse, even when it was passed study data, there was a bug in
747 pcre_dfa_exec() that meant it never actually used it. Double oops. There
748 were also very few tests of studied patterns with pcre_dfa_exec().
749
750 27. If (?| is used to create subpatterns with duplicate numbers, they are now
751 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
752 on the other side of the coin, they are no longer allowed to have different
753 names, because these cannot be distinguished in PCRE, and this has caused
754 confusion. (This is a difference from Perl.)
755
756 28. When duplicate subpattern names are present (necessarily with different
757 numbers, as required by 27 above), and a test is made by name in a
758 conditional pattern, either for a subpattern having been matched, or for
759 recursion in such a pattern, all the associated numbered subpatterns are
760 tested, and the overall condition is true if the condition is true for any
761 one of them. This is the way Perl works, and is also more like the way
762 testing by number works.
763
764
765 Version 7.9 11-Apr-09
766 ---------------------
767
768 1. When building with support for bzlib/zlib (pcregrep) and/or readline
769 (pcretest), all targets were linked against these libraries. This included
770 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
771 libraries. This caused unwanted dependencies to be created. This problem
772 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
773 pcretest is linked with readline.
774
775 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
776 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
777 moved outside it again, because FALSE and TRUE are already defined in AIX,
778 but BOOL is not.
779
780 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
781 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
782
783 4. The pcregrep documentation said spaces were inserted as well as colons (or
784 hyphens) following file names and line numbers when outputting matching
785 lines. This is not true; no spaces are inserted. I have also clarified the
786 wording for the --colour (or --color) option.
787
788 5. In pcregrep, when --colour was used with -o, the list of matching strings
789 was not coloured; this is different to GNU grep, so I have changed it to be
790 the same.
791
792 6. When --colo(u)r was used in pcregrep, only the first matching substring in
793 each matching line was coloured. Now it goes on to look for further matches
794 of any of the test patterns, which is the same behaviour as GNU grep.
795
796 7. A pattern that could match an empty string could cause pcregrep to loop; it
797 doesn't make sense to accept an empty string match in pcregrep, so I have
798 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
799 seems to be how GNU grep behaves.
800
801 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
802 start or after a newline", because the conditional assertion was not being
803 correctly handled. The rule now is that both the assertion and what follows
804 in the first alternative must satisfy the test.
805
806 9. If auto-callout was enabled in a pattern with a conditional group whose
807 condition was an assertion, PCRE could crash during matching, both with
808 pcre_exec() and pcre_dfa_exec().
809
810 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
811 used for matching.
812
813 11. Unicode property support in character classes was not working for
814 characters (bytes) greater than 127 when not in UTF-8 mode.
815
816 12. Added the -M command line option to pcretest.
817
818 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
819
820 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
821
822 16. Added comments and documentation about mis-use of no_arg in the C++
823 wrapper.
824
825 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
826 from Martin Jerabek that uses macro names for all relevant character and
827 string constants.
828
829 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
830 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
831 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
832 these, but not everybody uses configure.
833
834 19. A conditional group that had only one branch was not being correctly
835 recognized as an item that could match an empty string. This meant that an
836 enclosing group might also not be so recognized, causing infinite looping
837 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
838 with the subject "ab", where knowledge that the repeated group can match
839 nothing is needed in order to break the loop.
840
841 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
842 exec(), but without supplying a callout function, matching went wrong.
843
844 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
845 leak if the size of the offset vector was greater than 30. When the vector
846 is smaller, the saved offsets during recursion go onto a local stack
847 vector, but for larger vectors malloc() is used. It was failing to free
848 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
849 error, in fact).
850
851 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
852 heapframe that is used only when UTF-8 support is enabled. This caused no
853 problem, but was untidy.
854
855 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
856 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
857 included within another project.
858
859 24. Steven Van Ingelgem's patches to add more options to the CMake support,
860 slightly modified by me:
861
862 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
863 not building pcregrep.
864
865 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
866 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
867
868 25. Forward references, both numeric and by name, in patterns that made use of
869 duplicate group numbers, could behave incorrectly or give incorrect errors,
870 because when scanning forward to find the reference group, PCRE was not
871 taking into account the duplicate group numbers. A pattern such as
872 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
873
874 26. Changed a few more instances of "const unsigned char *" to USPTR, making
875 the feature of a custom pointer more persuasive (as requested by a user).
876
877 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
878 pcretest.c, inside #ifndefs, because it seems they are sometimes already
879 pre-defined.
880
881 28. Added support for (*UTF8) at the start of a pattern.
882
883 29. Arrange for flags added by the "release type" setting in CMake to be shown
884 in the configuration summary.
885
886
887 Version 7.8 05-Sep-08
888 ---------------------
889
890 1. Replaced UCP searching code with optimized version as implemented for Ad
891 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
892 stage table and inline lookup instead of a function, giving speed ups of 2
893 to 5 times on some simple patterns that I tested. Permission was given to
894 distribute the MultiStage2.py script that generates the tables (it's not in
895 the tarball, but is in the Subversion repository).
896
897 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
898 scripts.
899
900 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
901 a group with a zero qualifier. The result of the study could be incorrect,
902 or the function might crash, depending on the pattern.
903
904 4. Caseless matching was not working for non-ASCII characters in back
905 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
906 It now works when Unicode Property Support is available.
907
908 5. In pcretest, an escape such as \x{de} in the data was always generating
909 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
910 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
911 truncation.
912
913 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
914
915 7. Added two (int) casts to pcregrep when printing the difference of two
916 pointers, in case they are 64-bit values.
917
918 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
919 test 2 if it fails.
920
921 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
922 and a #define of that name to empty if it is not externally set. This is to
923 allow users of MSVC to set it if necessary.
924
925 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
926 the convenience functions in the pcre_get.c source file.
927
928 11. An option change at the start of a pattern that had top-level alternatives
929 could cause overwriting and/or a crash. This command provoked a crash in
930 some environments:
931
932 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
933
934 This potential security problem was recorded as CVE-2008-2371.
935
936 12. For a pattern where the match had to start at the beginning or immediately
937 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
938 pcre_dfa_exec() could read past the end of the passed subject if there was
939 no match. To help with detecting such bugs (e.g. with valgrind), I modified
940 pcretest so that it places the subject at the end of its malloc-ed buffer.
941
942 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
943 exec() might read past the end of the data buffer in UTF-8 mode.
944
945 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
946 the data contained the byte 0x85 as part of a UTF-8 character within its
947 first line. This applied both to normal and DFA matching.
948
949 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
950 /^[^d]*?$/8 failed to match "abc".
951
952 16. Added a missing copyright notice to pcrecpp_internal.h.
953
954 17. Make it more clear in the documentation that values returned from
955 pcre_exec() in ovector are byte offsets, not character counts.
956
957 18. Tidied a few places to stop certain compilers from issuing warnings.
958
959 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
960 supplied by Stefan Weber. I made a further small update for 7.8 because
961 there is a change of source arrangements: the pcre_searchfuncs.c module is
962 replaced by pcre_ucd.c.
963
964
965 Version 7.7 07-May-08
966 ---------------------
967
968 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
969 a string to a long long, pretend we don't even have a long long." This is
970 done by checking for the strtoq, strtoll, and _strtoi64 functions.
971
972 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
973 pre-7.6 versions, which defined a global no_arg variable instead of putting
974 it in the RE class. (See also #8 below.)
975
976 3. Remove a line of dead code, identified by coverity and reported by Nuno
977 Lopes.
978
979 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
980
981 (1) The include/exclude patterns were being applied to the whole pathnames
982 of files, instead of just to the final components.
983
984 (2) If there was more than one level of directory, the subdirectories were
985 skipped unless they satisfied the include/exclude conditions. This is
986 inconsistent with GNU grep (and could even be seen as contrary to the
987 pcregrep specification - which I improved to make it absolutely clear).
988 The action now is always to scan all levels of directory, and just
989 apply the include/exclude patterns to regular files.
990
991 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
992 --exclude_dir in the tests to avoid scanning .svn directories.
993
994 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
995 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
996 doesn't support NULs in patterns.
997
998 7. Added some missing "const"s to declarations of static tables in
999 pcre_compile.c and pcre_dfa_exec.c.
1000
1001 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1002 caused by fix #2 above. (Subsequently also a second patch to fix the
1003 first patch. And a third patch - this was a messy problem.)
1004
1005 9. Applied Craig's patch to remove the use of push_back().
1006
1007 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1008 matching function regexec().
1009
1010 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1011 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1012 references. PCRE supports relative numbers with this syntax (I don't think
1013 Oniguruma does).
1014
1015 12. Previously, a group with a zero repeat such as (...){0} was completely
1016 omitted from the compiled regex. However, this means that if the group
1017 was called as a subroutine from elsewhere in the pattern, things went wrong
1018 (an internal error was given). Such groups are now left in the compiled
1019 pattern, with a new opcode that causes them to be skipped at execution
1020 time.
1021
1022 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1023 to the way PCRE behaves:
1024
1025 (a) A lone ] character is dis-allowed (Perl treats it as data).
1026
1027 (b) A back reference to an unmatched subpattern matches an empty string
1028 (Perl fails the current match path).
1029
1030 (c) A data ] in a character class must be notated as \] because if the
1031 first data character in a class is ], it defines an empty class. (In
1032 Perl it is not possible to have an empty class.) The empty class []
1033 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1034 The negative empty class [^] matches any one character, independently
1035 of the DOTALL setting.
1036
1037 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1038 non-existent subpattern following a character class starting with ']' and
1039 containing () gave an internal compiling error instead of "reference to
1040 non-existent subpattern". Fortunately, when the pattern did exist, the
1041 compiled code was correct. (When scanning forwards to check for the
1042 existencd of the subpattern, it was treating the data ']' as terminating
1043 the class, so got the count wrong. When actually compiling, the reference
1044 was subsequently set up correctly.)
1045
1046 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1047 it was being rejected as not supported by pcre_dfa_exec(), even though
1048 other assertions are supported. I have made pcre_dfa_exec() support
1049 (*FAIL).
1050
1051 16. The implementation of 13c above involved the invention of a new opcode,
1052 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1053 cannot be changed at match time, I realized I could make a small
1054 improvement to matching performance by compiling OP_ALLANY instead of
1055 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1056 on the OP_ANY path.
1057
1058 17. Compiling pcretest on Windows with readline support failed without the
1059 following two fixes: (1) Make the unistd.h include conditional on
1060 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1061
1062 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1063 ncurses library to be included for pcretest when ReadLine support is
1064 requested, but also to allow for it to be overridden. This patch came from
1065 Daniel Bergström.
1066
1067 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1068 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1069 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1070 spotting this.
1071
1072
1073 Version 7.6 28-Jan-08
1074 ---------------------
1075
1076 1. A character class containing a very large number of characters with
1077 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1078 overflow.
1079
1080 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1081 HAVE_LONG_LONG is not defined.
1082
1083 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1084 bring them up to date and include new features. This patch includes:
1085
1086 - Fixed PH's badly added libz and libbz2 support.
1087 - Fixed a problem with static linking.
1088 - Added pcredemo. [But later removed - see 7 below.]
1089 - Fixed dftables problem and added an option.
1090 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1091 HAVE_LONG_LONG.
1092 - Added readline support for pcretest.
1093 - Added an listing of the option settings after cmake has run.
1094
1095 4. A user submitted a patch to Makefile that makes it easy to create
1096 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1097 Makefile.am that cause it to include this special target, without
1098 affecting anything else. Note that the same mingw target plus all
1099 the other distribution libraries and programs are now supported
1100 when configuring with CMake (see 6 below) instead of with
1101 Configure/Make.
1102
1103 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1104 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1105 exported in the Windows port". It has not yet been confirmed that the patch
1106 solves the problem, but it does no harm.
1107
1108 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1109 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1110 with CMake, and also correct the comment about stack recursion.
1111
1112 7. Remove the automatic building of pcredemo from the ./configure system and
1113 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1114 of a program that users should build themselves after PCRE is installed, so
1115 building it automatically is not really right. What is more, it gave
1116 trouble in some build environments.
1117
1118 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1119
1120
1121 Version 7.5 10-Jan-08
1122 ---------------------
1123
1124 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1125 values in parens when parsing an RE using the C++ wrapper."
1126
1127 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1128 Characters greater than 255 were excluded from the class instead of being
1129 included.
1130
1131 3. The same bug as (2) above applied to negated POSIX classes such as
1132 [:^space:].
1133
1134 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1135 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1136 I have changed it.
1137
1138 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1139 first named subpattern) and a construct such as (?&a) would reference the
1140 first named subpattern whose name started with "a" (in other words, the
1141 length check was missing). Both these problems are fixed. "Subpattern name
1142 expected" is now given for (?&) (a zero-length name), and this patch also
1143 makes it give the same error for \k'' (previously it complained that that
1144 was a reference to a non-existent subpattern).
1145
1146 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1147 this is right because (?- can be followed by option settings as well as by
1148 digits. I have, however, made the messages clearer.
1149
1150 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1151 than the number used in the conditional) now cause a compile-time error.
1152 This is actually not compatible with Perl, which accepts such patterns, but
1153 treats the conditional as always being FALSE (as PCRE used to), but it
1154 seems to me that giving a diagnostic is better.
1155
1156 8. Change "alphameric" to the more common word "alphanumeric" in comments
1157 and messages.
1158
1159 9. Fix two occurrences of "backslash" in comments that should have been
1160 "backspace".
1161
1162 10. Remove two redundant lines of code that can never be obeyed (their function
1163 was moved elsewhere).
1164
1165 11. The program that makes PCRE's Unicode character property table had a bug
1166 which caused it to generate incorrect table entries for sequences of
1167 characters that have the same character type, but are in different scripts.
1168 It amalgamated them into a single range, with the script of the first of
1169 them. In other words, some characters were in the wrong script. There were
1170 thirteen such cases, affecting characters in the following ranges:
1171
1172 U+002b0 - U+002c1
1173 U+0060c - U+0060d
1174 U+0061e - U+00612
1175 U+0064b - U+0065e
1176 U+0074d - U+0076d
1177 U+01800 - U+01805
1178 U+01d00 - U+01d77
1179 U+01d9b - U+01dbf
1180 U+0200b - U+0200f
1181 U+030fc - U+030fe
1182 U+03260 - U+0327f
1183 U+0fb46 - U+0fbb1
1184 U+10450 - U+1049d
1185
1186 12. The -o option (show only the matching part of a line) for pcregrep was not
1187 compatible with GNU grep in that, if there was more than one match in a
1188 line, it showed only the first of them. It now behaves in the same way as
1189 GNU grep.
1190
1191 13. If the -o and -v options were combined for pcregrep, it printed a blank
1192 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1193 does the same. The return code can be used to tell if there were any
1194 non-matching lines.
1195
1196 14. Added --file-offsets and --line-offsets to pcregrep.
1197
1198 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1199 infinitely looping recursion. The bug was that positive lookaheads were not
1200 being skipped when checking for a possible empty match (negative lookaheads
1201 and both kinds of lookbehind were skipped).
1202
1203 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1204 inclusion of <windows.h> to before rather than after the definition of
1205 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1206
1207 17. Specifying a possessive quantifier with a specific limit for a Unicode
1208 character property caused pcre_compile() to compile bad code, which led at
1209 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1210 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1211 caused the error; without that there was no problem.
1212
1213 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1214
1215 19. Added --enable-pcretest-libreadline.
1216
1217 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1218 RE::GlobalReplace(). As a result, the number of replacements returned was
1219 double what it should be. I removed one of the increments, but Craig sent a
1220 later patch that removed the other one (the right fix) and added unit tests
1221 that check the return values (which was not done before).
1222
1223 21. Several CMake things:
1224
1225 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1226 the names libpcre and libpcreposix, not just pcre and pcreposix.
1227
1228 (2) The above change means that pcretest and pcregrep are now correctly
1229 linked with the newly-built libraries, not previously installed ones.
1230
1231 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1232
1233 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1234 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1235 UTF-8 newline character). The key issue is that the pattern starts .*;
1236 this means that the match must be either at the beginning, or after a
1237 newline. The bug was in the code for advancing after a failed match and
1238 checking that the new position followed a newline. It was not taking
1239 account of UTF-8 characters correctly.
1240
1241 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1242 character classes. PCRE was not treating the sequence [:...:] as a
1243 character class unless the ... were all letters. Perl, however, seems to
1244 allow any characters between [: and :], though of course it rejects as
1245 unknown any "names" that contain non-letters, because all the known class
1246 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1247 for example, whereas PCRE did not - it did not recognize a POSIX character
1248 class. This seemed a bit dangerous, so the code has been changed to be
1249 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1250 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1251 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1252 Perl does, and where it didn't before.
1253
1254 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1255 Windows environments %n is disabled by default.
1256
1257
1258 Version 7.4 21-Sep-07
1259 ---------------------
1260
1261 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1262 means that a class such as [\s] counted as "explicit reference to CR or
1263 LF". That isn't really right - the whole point of the change was to try to
1264 help when there was an actual mention of one of the two characters. So now
1265 the change happens only if \r or \n (or a literal CR or LF) character is
1266 encountered.
1267
1268 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1269 of both had grown to the point where there were only 3 bits left.
1270 Fortunately, there was spare space in the data structure, and so I have
1271 moved the internal flags into a new 16-bit field to free up more option
1272 bits.
1273
1274 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1275 but did not set the internal JCHANGED flag - either of these is enough to
1276 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1277 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1278 start sets both bits.
1279
1280 4. Added options (at build time, compile time, exec time) to change \R from
1281 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1282
1283 5. doc/pcresyntax.html was missing from the distribution.
1284
1285 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1286 compatibility, even though it is no longer used.
1287
1288 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1289 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1290 windows.h file is present (where different names are used). [This was
1291 reversed later after testing - see 16 below.]
1292
1293 8. Changed all #include <config.h> to #include "config.h". There were also
1294 some further <pcre.h> cases that I changed to "pcre.h".
1295
1296 9. When pcregrep was used with the --colour option, it missed the line ending
1297 sequence off the lines that it output.
1298
1299 10. It was pointed out to me that arrays of string pointers cause lots of
1300 relocations when a shared library is dynamically loaded. A technique of
1301 using a single long string with a table of offsets can drastically reduce
1302 these. I have refactored PCRE in four places to do this. The result is
1303 dramatic:
1304
1305 Originally: 290
1306 After changing UCP table: 187
1307 After changing error message table: 43
1308 After changing table of "verbs" 36
1309 After changing table of Posix names 22
1310
1311 Thanks to the folks working on Gregex for glib for this insight.
1312
1313 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1314 unicode-properties was also set.
1315
1316 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1317
1318 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1319 checked only for CRLF.
1320
1321 14. Added casts to pcretest.c to avoid compiler warnings.
1322
1323 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1324
1325 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1326 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1327 entirely. This removes changes made in 7 above.
1328
1329 17. The CMake files have been updated, and there is now more information about
1330 building with CMake in the NON-UNIX-USE document.
1331
1332
1333 Version 7.3 28-Aug-07
1334 ---------------------
1335
1336 1. In the rejigging of the build system that eventually resulted in 7.1, the
1337 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1338 brackets there is not right, since it causes compilers to look for an
1339 installed pcre.h, not the version that is in the source that is being
1340 compiled (which of course may be different). I have changed it back to:
1341
1342 #include "pcre.h"
1343
1344 I have a vague recollection that the change was concerned with compiling in
1345 different directories, but in the new build system, that is taken care of
1346 by the VPATH setting the Makefile.
1347
1348 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1349 when the subject happened to end in the byte 0x85 (e.g. if the last
1350 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1351 characters but of course it shouldn't be taken as a newline when it is part
1352 of another character. The bug was that, for an unlimited repeat of . in
1353 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1354 characters when looking for a newline.
1355
1356 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1357
1358 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1359 in debug output.
1360
1361 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1362 long printing in the pcrecpp unittest when running under MinGW.
1363
1364 6. ESC_K was left out of the EBCDIC table.
1365
1366 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1367 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1368 limit also applies to "virtual nesting" when a pattern is recursive, and in
1369 this case 1000 isn't so big. I have been able to remove this limit at the
1370 expense of backing off one optimization in certain circumstances. Normally,
1371 when pcre_exec() would call its internal match() function recursively and
1372 immediately return the result unconditionally, it uses a "tail recursion"
1373 feature to save stack. However, when a subpattern that can match an empty
1374 string has an unlimited repetition quantifier, it no longer makes this
1375 optimization. That gives it a stack frame in which to save the data for
1376 checking that an empty string has been matched. Previously this was taken
1377 from the 1000-entry workspace that had been reserved. So now there is no
1378 explicit limit, but more stack is used.
1379
1380 8. Applied Daniel's patches to solve problems with the import/export magic
1381 syntax that is required for Windows, and which was going wrong for the
1382 pcreposix and pcrecpp parts of the library. These were overlooked when this
1383 problem was solved for the main library.
1384
1385 9. There were some crude static tests to avoid integer overflow when computing
1386 the size of patterns that contain repeated groups with explicit upper
1387 limits. As the maximum quantifier is 65535, the maximum group length was
1388 set at 30,000 so that the product of these two numbers did not overflow a
1389 32-bit integer. However, it turns out that people want to use groups that
1390 are longer than 30,000 bytes (though not repeat them that many times).
1391 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1392 made it possible to implement the integer overflow checks in a much more
1393 dynamic way, which I have now done. The artificial limitation on group
1394 length has been removed - we now have only the limit on the total length of
1395 the compiled pattern, which depends on the LINK_SIZE setting.
1396
1397 10. Fixed a bug in the documentation for get/copy named substring when
1398 duplicate names are permitted. If none of the named substrings are set, the
1399 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1400 empty string.
1401
1402 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1403 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1404 because the ] is interpreted as the first data character and the
1405 terminating ] is not found. PCRE has been made compatible with Perl in this
1406 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1407 cause memory overwriting.
1408
1409 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1410 string has been matched (to stop an infinite loop). It was not recognizing
1411 a conditional subpattern that could match an empty string if that
1412 subpattern was within another subpattern. For example, it looped when
1413 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1414 condition was not nested. This bug has been fixed.
1415
1416 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1417 past the start of the subject in the presence of bytes with the top bit
1418 set, for example "\x8aBCD".
1419
1420 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1421 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1422
1423 14. Optimized (?!) to (*FAIL).
1424
1425 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1426 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1427 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1428 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1429 does: it's just the validity check that is more restrictive.
1430
1431 16. Inserted checks for integer overflows during escape sequence (backslash)
1432 processing, and also fixed erroneous offset values for syntax errors during
1433 backslash processing.
1434
1435 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1436 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1437
1438 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1439 caused an overrun.
1440
1441 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1442 something other than just ASCII characters) inside a group that had an
1443 unlimited repeat caused a loop at compile time (while checking to see
1444 whether the group could match an empty string).
1445
1446 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1447 [\P{Any}] did so. (Error in the code for printing property names.)
1448
1449 21. An orphan \E inside a character class could cause a crash.
1450
1451 22. A repeated capturing bracket such as (A)? could cause a wild memory
1452 reference during compilation.
1453
1454 23. There are several functions in pcre_compile() that scan along a compiled
1455 expression for various reasons (e.g. to see if it's fixed length for look
1456 behind). There were bugs in these functions when a repeated \p or \P was
1457 present in the pattern. These operators have additional parameters compared
1458 with \d, etc, and these were not being taken into account when moving along
1459 the compiled data. Specifically:
1460
1461 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1462 length.
1463
1464 (b) An item such as \pL+ within a repeated group could cause crashes or
1465 loops.
1466
1467 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1468 "reference to non-existent subpattern" error.
1469
1470 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1471
1472 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1473 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1474
1475 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1476
1477 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1478 character were causing crashes (broken optimization).
1479
1480 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1481 \p or \P) caused a compile-time loop.
1482
1483 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1484 break. For example, the unstudied pattern [\r\n]A does not match the string
1485 "\r\nA" because change 7.0/46 below moves the current point on by two
1486 characters after failing to match at the start. However, the pattern \nA
1487 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1488 the same is true. There doesn't seem any very clean way out of this, but
1489 what I have chosen to do makes the common cases work: PCRE now takes note
1490 of whether there can be an explicit match for \r or \n anywhere in the
1491 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1492 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1493 pattern has explicit CR or LF references.
1494
1495 29. Added (*CR) etc for changing newline setting at start of pattern.
1496
1497
1498 Version 7.2 19-Jun-07
1499 ---------------------
1500
1501 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1502 which is apparently normally available under Windows.
1503
1504 2. Re-jig the pcregrep tests with different newline settings in an attempt
1505 to make them independent of the local environment's newline setting.
1506
1507 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1508
1509 4. Some of the "internals" tests were previously cut out when the link size
1510 was not 2, because the output contained actual offsets. The recent new
1511 "Z" feature of pcretest means that these can be cut out, making the tests
1512 usable with all link sizes.
1513
1514 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1515 stack recursion. This gives a massive performance boost under BSD, but just
1516 a small improvement under Linux. However, it saves one field in the frame
1517 in all cases.
1518
1519 6. Added more features from the forthcoming Perl 5.10:
1520
1521 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1522 recursion call. It refers to the nth most recently opened parentheses.
1523
1524 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1525 to be opened parentheses.
1526
1527 (c) Conditions that refer to capturing parentheses can be specified
1528 relatively, for example, (?(-2)... or (?(+3)...
1529
1530 (d) \K resets the start of the current match so that everything before
1531 is not part of it.
1532
1533 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1534
1535 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1536 reference syntax.
1537
1538 (g) (?| introduces a group in which the numbering of parentheses in each
1539 alternative starts with the same number.
1540
1541 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1542
1543 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1544 PCRE_INFO_JCHANGED.
1545
1546 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1547 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1548 for detecting groups that can match an empty string.
1549
1550 9. A pattern with a very large number of alternatives (more than several
1551 hundred) was running out of internal workspace during the pre-compile
1552 phase, where pcre_compile() figures out how much memory will be needed. A
1553 bit of new cunning has reduced the workspace needed for groups with
1554 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1555 workspace instead of running out of the 4096 that are available.
1556
1557 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1558
1559 11. Applied patch from Google to remove an optimization that didn't quite work.
1560 The report of the bug said:
1561
1562 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1563 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1564 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1565
1566 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1567 it matched the wrong number of bytes.
1568
1569
1570 Version 7.1 24-Apr-07
1571 ---------------------
1572
1573 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1574 that is more "standard", making use of automake and other Autotools. There
1575 is some re-arrangement of the files and adjustment of comments consequent
1576 on this.
1577
1578 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1579 for recursive directory scanning broke on some systems because the files
1580 are not scanned in any specific order and on different systems the order
1581 was different. A call to "sort" has been inserted into RunGrepTest for the
1582 approprate test as a short-term fix. In the longer term there may be an
1583 alternative.
1584
1585 3. I had an email from Eric Raymond about problems translating some of PCRE's
1586 man pages to HTML (despite the fact that I distribute HTML pages, some
1587 people do their own conversions for various reasons). The problems
1588 concerned the use of low-level troff macros .br and .in. I have therefore
1589 removed all such uses from the man pages (some were redundant, some could
1590 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1591 HTML has been updated to handle .nf/.fi and to complain if it encounters
1592 .br or .in.
1593
1594 4. Updated comments in configure.ac that get placed in config.h.in and also
1595 arranged for config.h to be included in the distribution, with the name
1596 config.h.generic, for the benefit of those who have to compile without
1597 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1598
1599 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1600 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1601 makevp.bat for the current PCRE, using the additional files
1602 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1603
1604 6. A Windows user reported a minor discrepancy with test 2, which turned out
1605 to be caused by a trailing space on an input line that had got lost in his
1606 copy. The trailing space was an accident, so I've just removed it.
1607
1608 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1609 that is needed.
1610
1611 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1612 as "const" (a) because they are and (b) because it helps the PHP
1613 maintainers who have recently made a script to detect big data structures
1614 in the php code that should be moved to the .rodata section. I remembered
1615 to update Builducptable as well, so it won't revert if ucptable.h is ever
1616 re-created.
1617
1618 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1619 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1620 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1621 support is not required. This saves 1.5-2K of code, which is important in
1622 some applications.
1623
1624 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1625 so as not to refer to the tables, even though these functions will never be
1626 called when UTF-8 support is disabled. Otherwise there are problems with a
1627 shared library.
1628
1629 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1630
1631 (a) It was defining its arguments as char * instead of void *.
1632
1633 (b) It was assuming that all moves were upwards in memory; this was true
1634 a long time ago when I wrote it, but is no longer the case.
1635
1636 The emulated memove() is provided for those environments that have neither
1637 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1638 is clearly not the case, as these two bugs were recently reported.
1639
1640 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1641 and Detrail to create the HTML documentation, the .txt form of the man
1642 pages, and it removes trailing spaces from listed files. It also creates
1643 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1644 case, it wraps all the #defines with #ifndefs. This script should be run
1645 before "make dist".
1646
1647 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1648 with Unicode property support.
1649
1650 (a) For a maximizing quantifier, if the two different cases of the
1651 character were of different lengths in their UTF-8 codings (there are
1652 some cases like this - I found 11), and the matching function had to
1653 back up over a mixture of the two cases, it incorrectly assumed they
1654 were both the same length.
1655
1656 (b) When PCRE was configured to use the heap rather than the stack for
1657 recursion during matching, it was not correctly preserving the data for
1658 the other case of a UTF-8 character when checking ahead for a match
1659 while processing a minimizing repeat. If the check also involved
1660 matching a wide character, but failed, corruption could cause an
1661 erroneous result when trying to check for a repeat of the original
1662 character.
1663
1664 13. Some tidying changes to the testing mechanism:
1665
1666 (a) The RunTest script now detects the internal link size and whether there
1667 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1668 values substituted by "configure". (The RunGrepTest script already did
1669 this for UTF-8.) The configure.ac script no longer substitutes the
1670 relevant variables.
1671
1672 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1673 with length and offset values. This means that the output is different
1674 for different internal link sizes. Test 2 is skipped for link sizes
1675 other than 2 because of this, bypassing the problem. Unfortunately,
1676 there was also a test in test 3 (the locale tests) that used /B and
1677 failed for link sizes other than 2. Rather than cut the whole test out,
1678 I have added a new /Z option to pcretest that replaces the length and
1679 offset values with spaces. This is now used to make test 3 independent
1680 of link size. (Test 2 will be tidied up later.)
1681
1682 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1683 segmentation fault instead of returning the appropriate error message.
1684
1685 15. In multiline mode when the newline sequence was set to "any", the pattern
1686 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1687 This doesn't seem right; it now treats the CRLF combination as the line
1688 ending, and so does not match in that case. It's only a pattern such as ^$
1689 that would hit this one: something like ^ABC$ would have failed after \r
1690 and then tried again after \r\n.
1691
1692 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1693 in an attempt to make files that differ only in their line terminators
1694 compare equal. This works on Linux.
1695
1696 17. Under certain error circumstances pcregrep might try to free random memory
1697 as it exited. This is now fixed, thanks to valgrind.
1698
1699 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1700 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1701 was because its rules for how to advance for /g after matching an empty
1702 string at the end of a line did not allow for this case. They now check for
1703 it specially.
1704
1705 20. pcretest is supposed to handle patterns and data of any length, by
1706 extending its buffers when necessary. It was getting this wrong when the
1707 buffer for a data line had to be extended.
1708
1709 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1710 CRLF as a newline sequence.
1711
1712 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1713 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1714 I have nevertheless tidied it up.
1715
1716 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1717
1718 24. Added a man page for pcre-config.
1719
1720
1721 Version 7.0 19-Dec-06
1722 ---------------------
1723
1724 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1725 moving to gcc 4.1.1.
1726
1727 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1728 sys/time.h, which is documented as needed for this function. It doesn't
1729 seem to matter on Linux, but it showed up on some releases of OS X.
1730
1731 3. It seems that there are systems where bytes whose values are greater than
1732 127 match isprint() in the "C" locale. The "C" locale should be the
1733 default when a C program starts up. In most systems, only ASCII printing
1734 characters match isprint(). This difference caused the output from pcretest
1735 to vary, making some of the tests fail. I have changed pcretest so that:
1736
1737 (a) When it is outputting text in the compiled version of a pattern, bytes
1738 other than 32-126 are always shown as hex escapes.
1739
1740 (b) When it is outputting text that is a matched part of a subject string,
1741 it does the same, unless a different locale has been set for the match
1742 (using the /L modifier). In this case, it uses isprint() to decide.
1743
1744 4. Fixed a major bug that caused incorrect computation of the amount of memory
1745 required for a compiled pattern when options that changed within the
1746 pattern affected the logic of the preliminary scan that determines the
1747 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1748 that the computed length was too small. The symptoms of this bug were
1749 either the PCRE error "internal error: code overflow" from pcre_compile(),
1750 or a glibc crash with a message such as "pcretest: free(): invalid next
1751 size (fast)". Examples of patterns that provoked this bug (shown in
1752 pcretest format) are:
1753
1754 /(?-x: )/x
1755 /(?x)(?-x: \s*#\s*)/
1756 /((?i)[\x{c0}])/8
1757 /(?i:[\x{c0}])/8
1758
1759 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1760 is now done differently.
1761
1762 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1763 wrapper classes; (b) implement a new function in the C++ scanner that is
1764 more efficient than the old way of doing things because it avoids levels of
1765 recursion in the regex matching; (c) add a paragraph to the documentation
1766 for the FullMatch() function.
1767
1768 6. The escape sequence \n was being treated as whatever was defined as
1769 "newline". Not only was this contrary to the documentation, which states
1770 that \n is character 10 (hex 0A), but it also went horribly wrong when
1771 "newline" was defined as CRLF. This has been fixed.
1772
1773 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1774 was being set to -1 for the "end of line" case (supposedly a value that no
1775 character can have). Though this value is never used (the check for end of
1776 line is "zero bytes in current character"), it caused compiler complaints.
1777 I've changed it to 0xffffffff.
1778
1779 8. In pcre_version.c, the version string was being built by a sequence of
1780 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1781 string (as it is for production releases) called a macro with an empty
1782 argument. The C standard says the result of this is undefined. The gcc
1783 compiler treats it as an empty string (which was what was wanted) but it is
1784 reported that Visual C gives an error. The source has been hacked around to
1785 avoid this problem.
1786
1787 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1788 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1789 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1790 of them did).
1791
1792 10. Originally, pcretest opened its input and output without "b"; then I was
1793 told that "b" was needed in some environments, so it was added for release
1794 5.0 to both the input and output. (It makes no difference on Unix-like
1795 systems.) Later I was told that it is wrong for the input on Windows. I've
1796 now abstracted the modes into two macros, to make it easier to fiddle with
1797 them, and removed "b" from the input mode under Windows.
1798
1799 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1800
1801 12. Added -help and --help to pcretest as an official way of being reminded
1802 of the options.
1803
1804 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1805 and pcrecpp.cc because they annoy compilers at high warning levels.
1806
1807 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1808
1809 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1810 scripts are not C programs :-) and which was not noticed because it works
1811 on Linux.
1812
1813 16. pcretest is supposed to handle any length of pattern and data line (as one
1814 line or as a continued sequence of lines) by extending its input buffer if
1815 necessary. This feature was broken for very long pattern lines, leading to
1816 a string of junk being passed to pcre_compile() if the pattern was longer
1817 than about 50K.
1818
1819 17. I have done a major re-factoring of the way pcre_compile() computes the
1820 amount of memory needed for a compiled pattern. Previously, there was code
1821 that made a preliminary scan of the pattern in order to do this. That was
1822 OK when PCRE was new, but as the facilities have expanded, it has become
1823 harder and harder to keep it in step with the real compile phase, and there
1824 have been a number of bugs (see for example, 4 above). I have now found a
1825 cunning way of running the real compile function in a "fake" mode that
1826 enables it to compute how much memory it would need, while actually only
1827 ever using a few hundred bytes of working memory and without too many
1828 tests of the mode. This should make future maintenance and development
1829 easier. A side effect of this work is that the limit of 200 on the nesting
1830 depth of parentheses has been removed (though this was never a serious
1831 limitation, I suspect). However, there is a downside: pcre_compile() now
1832 runs more slowly than before (30% or more, depending on the pattern). I
1833 hope this isn't a big issue. There is no effect on runtime performance.
1834
1835 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1836 newline (only possible for the last line of a file) and it was a
1837 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1838
1839 19. Added additional timing features to pcretest. (1) The -tm option now times
1840 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1841 separate command line item, by a number that specifies the number of
1842 repeats to use when timing. The default is 50000; this gives better
1843 precision, but takes uncomfortably long for very large patterns.
1844
1845 20. Extended pcre_study() to be more clever in cases where a branch of a
1846 subpattern has no definite first character. For example, (a*|b*)[cd] would
1847 previously give no result from pcre_study(). Now it recognizes that the
1848 first character must be a, b, c, or d.
1849
1850 21. There was an incorrect error "recursive call could loop indefinitely" if
1851 a subpattern (or the entire pattern) that was being tested for matching an
1852 empty string contained only one non-empty item after a nested subpattern.
1853 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1854 incorrectly, because the \d was being skipped in the check.
1855
1856 22. The pcretest program now has a new pattern option /B and a command line
1857 option -b, which is equivalent to adding /B to every pattern. This causes
1858 it to show the compiled bytecode, without the additional information that
1859 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1860 is the same as /B/I).
1861
1862 23. A new optimization is now able automatically to treat some sequences such
1863 as a*b as a*+b. More specifically, if something simple (such as a character
1864 or a simple class like \d) has an unlimited quantifier, and is followed by
1865 something that cannot possibly match the quantified thing, the quantifier
1866 is automatically "possessified".
1867
1868 24. A recursive reference to a subpattern whose number was greater than 39
1869 went wrong under certain circumstances in UTF-8 mode. This bug could also
1870 have affected the operation of pcre_study().
1871
1872 25. Realized that a little bit of performance could be had by replacing
1873 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1874
1875 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1876
1877 27. Possessive quantifiers such as a++ were previously implemented by turning
1878 them into atomic groups such as ($>a+). Now they have their own opcodes,
1879 which improves performance. This includes the automatically created ones
1880 from 23 above.
1881
1882 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1883 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1884 the first matched character to be a colon. This applied both to named and
1885 numbered groups.
1886
1887 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1888
1889 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1890 building PCRE on Windows easier, so I have included it in the distribution.
1891
1892 31. There is now a check in pcretest against a ridiculously large number being
1893 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1894 loop, the loop is abandoned.
1895
1896 32. Forward references to subpatterns in conditions such as (?(2)...) where
1897 subpattern 2 is defined later cause pcre_compile() to search forwards in
1898 the pattern for the relevant set of parentheses. This search went wrong
1899 when there were unescaped parentheses in a character class, parentheses
1900 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1901
1902 33. "Subroutine" calls and backreferences were previously restricted to
1903 referencing subpatterns earlier in the regex. This restriction has now
1904 been removed.
1905
1906 34. Added a number of extra features that are going to be in Perl 5.10. On the
1907 whole, these are just syntactic alternatives for features that PCRE had
1908 previously implemented using the Python syntax or my own invention. The
1909 other formats are all retained for compatibility.
1910
1911 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1912 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1913 also .NET compatible.
1914
1915 (b) A recursion or subroutine call to a named group can now be defined as
1916 (?&name) as well as (?P>name).
1917
1918 (c) A backreference to a named group can now be defined as \k<name> or
1919 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1920 5.10, are also .NET compatible.
1921
1922 (d) A conditional reference to a named group can now use the syntax
1923 (?(<name>) or (?('name') as well as (?(name).
1924
1925 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1926 groups (named and numbered) that are never evaluated inline, but can be
1927 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1928 is always false. There may be only one alternative in such a group.
1929
1930 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1931 as the simple (?(R). The condition is true only if the most recent
1932 recursion is that of the given number or name. It does not search out
1933 through the entire recursion stack.
1934
1935 (g) The escape \gN or \g{N} has been added, where N is a positive or
1936 negative number, specifying an absolute or relative reference.
1937
1938 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1939 some "unreachable code" warnings.
1940
1941 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1942 things, this adds five new scripts.
1943
1944 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1945 There were also incompatibilities regarding the handling of \Q..\E inside
1946 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1947 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1948
1949 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1950 matches an empty string, and forcibly breaks the loop. There were bugs in
1951 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1952 against aaaa the result was just "a" rather than "aaaa", for example. Two
1953 separate and independent bugs (that affected different cases) have been
1954 fixed.
1955
1956 39. Refactored the code to abolish the use of different opcodes for small
1957 capturing bracket numbers. This is a tidy that I avoided doing when I
1958 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1959 The new approach is not only tidier, it makes it possible to reduce the
1960 memory needed to fix the previous bug (38).
1961
1962 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1963 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1964 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1965 mode.
1966
1967 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1968 report.
1969
1970 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1971 copy construction and assignment in the C++ wrapper.
1972
1973 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1974 couple of bugs that could have given wrong results in the "--newline=crlf"
1975 case.
1976
1977 44. Added a number of casts and did some reorganization of signed/unsigned int
1978 variables following suggestions from Dair Grant. Also renamed the variable
1979 "this" as "item" because it is a C++ keyword.
1980
1981 45. Arranged for dftables to add
1982
1983 #include "pcre_internal.h"
1984
1985 to pcre_chartables.c because without it, gcc 4.x may remove the array
1986 definition from the final binary if PCRE is built into a static library and
1987 dead code stripping is activated.
1988
1989 46. For an unanchored pattern, if a match attempt fails at the start of a
1990 newline sequence, and the newline setting is CRLF or ANY, and the next two
1991 characters are CRLF, advance by two characters instead of one.
1992
1993
1994 Version 6.7 04-Jul-06
1995 ---------------------
1996
1997 1. In order to handle tests when input lines are enormously long, pcretest has
1998 been re-factored so that it automatically extends its buffers when
1999 necessary. The code is crude, but this _is_ just a test program. The
2000 default size has been increased from 32K to 50K.
2001
2002 2. The code in pcre_study() was using the value of the re argument before
2003 testing it for NULL. (Of course, in any sensible call of the function, it
2004 won't be NULL.)
2005
2006 3. The memmove() emulation function in pcre_internal.h, which is used on
2007 systems that lack both memmove() and bcopy() - that is, hardly ever -
2008 was missing a "static" storage class specifier.
2009
2010 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2011 containing an extended class (one that cannot be represented by a bitmap
2012 because it contains high-valued characters or Unicode property items, e.g.
2013 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2014 pattern, but PCRE should not loop if you do not (it no longer does).
2015 [Detail: two cases were found: (a) a repeated subpattern containing an
2016 extended class; (b) a recursive reference to a subpattern that followed a
2017 previous extended class. It wasn't skipping over the extended class
2018 correctly when UTF-8 mode was not set.]
2019
2020 5. A negated single-character class was not being recognized as fixed-length
2021 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2022 compile error "lookbehind assertion is not fixed length".
2023
2024 6. The RunPerlTest auxiliary script was showing an unexpected difference
2025 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2026 write a Perl script that can interpret lines of an input file either as
2027 byte characters or as UTF-8, which is what "perltest" was being required to
2028 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2029 can't do is switch easily at run time between having the "use utf8;" pragma
2030 or not. In the end, I fudged it by using the RunPerlTest script to insert
2031 "use utf8;" explicitly for the UTF-8 tests.
2032
2033 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2034 the end of the subject string, contrary to the documentation and to what
2035 Perl does. This was true of both matching functions. Now it matches only at
2036 the start of the subject and immediately after *internal* newlines.
2037
2038 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2039 a pointer to an int instead of a pointer to an unsigned long int. This
2040 caused problems on 64-bit systems.
2041
2042 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2043 instance of the 'standard' template library not being so standard".
2044
2045 10. There was no check on the number of named subpatterns nor the maximum
2046 length of a subpattern name. The product of these values is used to compute
2047 the size of the memory block for a compiled pattern. By supplying a very
2048 long subpattern name and a large number of named subpatterns, the size
2049 computation could be caused to overflow. This is now prevented by limiting
2050 the length of names to 32 characters, and the number of named subpatterns
2051 to 10,000.
2052
2053 11. Subpatterns that are repeated with specific counts have to be replicated in
2054 the compiled pattern. The size of memory for this was computed from the
2055 length of the subpattern and the repeat count. The latter is limited to
2056 65535, but there was no limit on the former, meaning that integer overflow
2057 could in principle occur. The compiled length of a repeated subpattern is
2058 now limited to 30,000 bytes in order to prevent this.
2059
2060 12. Added the optional facility to have named substrings with the same name.
2061
2062 13. Added the ability to use a named substring as a condition, using the
2063 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2064 are numbers (not recommended). Forward references are permitted.
2065
2066 14. Added forward references in named backreferences (if you see what I mean).
2067
2068 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2069 pattern could run off the end of the subject. For example, the pattern
2070 "(?s)(.{1,5})"8 did this with the subject "ab".
2071
2072 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2073 PCRE_CASELESS was set when matching characters that were quantified with ?
2074 or *.
2075
2076 17. A character class other than a single negated character that had a minimum
2077 but no maximum quantifier - for example [ab]{6,} - was not handled
2078 correctly by pce_dfa_exec(). It would match only one character.
2079
2080 18. A valid (though odd) pattern that looked like a POSIX character
2081 class but used an invalid character after [ (for example [[,abc,]]) caused
2082 pcre_compile() to give the error "Failed: internal error: code overflow" or
2083 in some cases to crash with a glibc free() error. This could even happen if
2084 the pattern terminated after [[ but there just happened to be a sequence of
2085 letters, a binary zero, and a closing ] in the memory that followed.
2086
2087 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2088 over the years. Originally (before any Unicode support), just the bottom 8
2089 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2090 output from "man perlunicode" includes this:
2091
2092 The regular expression compiler produces polymorphic opcodes. That
2093 is, the pattern adapts to the data and automatically switches to
2094 the Unicode character scheme when presented with Unicode data--or
2095 instead uses a traditional byte scheme when presented with byte
2096 data.
2097
2098 Sadly, a wide octal escape does not cause a switch, and in a string with
2099 no other multibyte characters, these octal escapes are treated as before.
2100 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2101 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2102 Unicode string.
2103
2104 I have not perpetrated such confusion in PCRE. Up till now, it took just
2105 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2106 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2107 translate to the appropriate multibyte character.
2108
2109 29. Applied some refactoring to reduce the number of warnings from Microsoft
2110 and Borland compilers. This has included removing the fudge introduced
2111 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2112 a warning about an unused variable.
2113
2114 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2115 characters since release 4.0, because Perl (from release 5.004) does not.
2116 [Or at least, is documented not to: some releases seem to be in conflict
2117 with the documentation.] However, when a pattern was studied with
2118 pcre_study() and all its branches started with \s, PCRE still included VT
2119 as a possible starting character. Of course, this did no harm; it just
2120 caused an unnecessary match attempt.
2121
2122 22. Removed a now-redundant internal flag bit that recorded the fact that case
2123 dependency changed within the pattern. This was once needed for "required
2124 byte" processing, but is no longer used. This recovers a now-scarce options
2125 bit. Also moved the least significant internal flag bit to the most-
2126 significant bit of the word, which was not previously used (hangover from
2127 the days when it was an int rather than a uint) to free up another bit for
2128 the future.
2129
2130 23. Added support for CRLF line endings as well as CR and LF. As well as the
2131 default being selectable at build time, it can now be changed at runtime
2132 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2133 specify that it is scanning data with non-default line endings.
2134
2135 24. Changed the definition of CXXLINK to make it agree with the definition of
2136 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2137
2138 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2139 recursions. This makes a big different to stack usage for some patterns.
2140
2141 26. If a subpattern containing a named recursion or subroutine reference such
2142 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2143 the space required for the compiled pattern went wrong and gave too small a
2144 value. Depending on the environment, this could lead to "Failed: internal
2145 error: code overflow at offset 49" or "glibc detected double free or
2146 corruption" errors.
2147
2148 27. Applied patches from Google (a) to support the new newline modes and (b) to
2149 advance over multibyte UTF-8 characters in GlobalReplace.
2150
2151 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2152 difference for some implementation of PCRE in some Windows version.
2153
2154 29. Added some extra testing facilities to pcretest:
2155
2156 \q<number> in a data line sets the "match limit" value
2157 \Q<number> in a data line sets the "match recursion limt" value
2158 -S <number> sets the stack size, where <number> is in megabytes
2159
2160 The -S option isn't available for Windows.
2161
2162
2163 Version 6.6 06-Feb-06
2164 ---------------------
2165
2166 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2167 in pcreposix.h. I have copied the definition from pcre.h.
2168
2169 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2170 because pcre.h is no longer a built file.
2171
2172 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2173 not normally included in the compiled code.
2174
2175
2176 Version 6.5 01-Feb-06
2177 ---------------------
2178
2179 1. When using the partial match feature with pcre_dfa_exec(), it was not
2180 anchoring the second and subsequent partial matches at the new starting
2181 point. This could lead to incorrect results. For example, with the pattern
2182 /1234/, partially matching against "123" and then "a4" gave a match.
2183
2184 2. Changes to pcregrep:
2185
2186 (a) All non-match returns from pcre_exec() were being treated as failures
2187 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2188 error message is output. Some extra information is given for the
2189 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2190 probably the only errors that are likely to be caused by users (by
2191 specifying a regex that has nested indefinite repeats, for instance).
2192 If there are more than 20 of these errors, pcregrep is abandoned.
2193
2194 (b) A binary zero was treated as data while matching, but terminated the
2195 output line if it was written out. This has been fixed: binary zeroes
2196 are now no different to any other data bytes.
2197
2198 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2199 used to set a locale for matching. The --locale=xxxx long option has
2200 been added (no short equivalent) to specify a locale explicitly on the
2201 pcregrep command, overriding the environment variables.
2202
2203 (d) When -B was used with -n, some line numbers in the output were one less
2204 than they should have been.
2205
2206 (e) Added the -o (--only-matching) option.
2207
2208 (f) If -A or -C was used with -c (count only), some lines of context were
2209 accidentally printed for the final match.
2210
2211 (g) Added the -H (--with-filename) option.
2212
2213 (h) The combination of options -rh failed to suppress file names for files
2214 that were found from directory arguments.
2215
2216 (i) Added the -D (--devices) and -d (--directories) options.
2217
2218 (j) Added the -F (--fixed-strings) option.
2219
2220 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2221
2222 (l) Added the --colo(u)r option.
2223
2224 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2225 is not present by default.
2226
2227 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2228 items such as (?R) or (?1), when the recursion could match a number of
2229 alternatives. If it matched one of the alternatives, but subsequently,
2230 outside the recursion, there was a failure, the code tried to back up into
2231 the recursion. However, because of the way PCRE is implemented, this is not
2232 possible, and the result was an incorrect result from the match.
2233
2234 In order to prevent this happening, the specification of recursion has
2235 been changed so that all such subpatterns are automatically treated as
2236 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2237
2238 4. I had overlooked the fact that, in some locales, there are characters for
2239 which isalpha() is true but neither isupper() nor islower() are true. In
2240 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2241 and ordfeminine) are like this. This affected the treatment of \w and \W
2242 when they appeared in character classes, but not when they appeared outside
2243 a character class. The bit map for "word" characters is now created
2244 separately from the results of isalnum() instead of just taking it from the
2245 upper, lower, and digit maps. (Plus the underscore character, of course.)
2246
2247 5. The above bug also affected the handling of POSIX character classes such as
2248 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2249 permanent tables. Instead, the bit maps for such a class were previously
2250 created as the appropriate unions of the upper, lower, and digit bitmaps.
2251 Now they are created by subtraction from the [[:word:]] class, which has
2252 its own bitmap.
2253
2254 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2255 It is created by subtracting the vertical space characters (\x09, \x0a,
2256 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2257 subtraction was done in the overall bitmap for a character class, meaning
2258 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2259 be recognized. This bug has been fixed.
2260
2261 7. Patches from the folks at Google:
2262
2263 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2264 real life, but is still worth protecting against".
2265
2266 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2267 regular expressions".
2268
2269 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2270 have it.
2271
2272 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2273 "configure" and the latter not, in order to fix a problem somebody had
2274 with compiling the Arg class on HP-UX.
2275
2276 (e) Improve the error-handling of the C++ wrapper a little bit.
2277
2278 (f) New tests for checking recursion limiting.
2279
2280 8. The pcre_memmove() function, which is used only if the environment does not
2281 have a standard memmove() function (and is therefore rarely compiled),
2282 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2283 returning a result (though PCRE never actually uses the result).
2284
2285 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2286 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2287 returned instead of calling malloc() with an overflowing number that would
2288 most likely cause subsequent chaos.
2289
2290 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2291
2292 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2293 with this option is matched, the nmatch and pmatch options of regexec() are
2294 ignored.
2295
2296 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2297 provided in case anyone wants to the the POSIX interface with UTF-8
2298 strings.
2299
2300 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2301 C++ linking (needed for some HP-UX environments).
2302
2303 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2304 (unused parameter) and in the pcre_printint() function (omitted "default"
2305 switch label when the default is to do nothing).
2306
2307 15. Added some code to make it possible, when PCRE is compiled as a C++
2308 library, to replace subject pointers for pcre_exec() with a smart pointer
2309 class, thus making it possible to process discontinuous strings.
2310
2311 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2312 much the same function. They were added by different people who were trying
2313 to make PCRE easy to compile on non-Unix systems. It has been suggested
2314 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2315 for compiling on Windows systems. I have therefore replaced it with
2316 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2317 defaults to "extern" for C or "extern C" for C++, which works fine on
2318 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2319 SCOPE with something explicit in config.h. In addition:
2320
2321 (a) pcreposix.h still had just "extern" instead of either of these macros;
2322 I have replaced it with PCRE_DATA_SCOPE.
2323
2324 (b) Functions such as _pcre_xclass(), which are internal to the library,
2325 but external in the C sense, all had PCRE_EXPORT in their definitions.
2326 This is apparently wrong for the Windows case, so I have removed it.
2327 (It makes no difference on Unix-like systems.)
2328
2329 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2330 of recursive calls to match(). This is different to MATCH_LIMIT because
2331 that limits the total number of calls to match(), not all of which increase
2332 the depth of recursion. Limiting the recursion depth limits the amount of
2333 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2334 when PCRE is compiled, and changed at run time. A patch from Google adds
2335 this functionality to the C++ interface.
2336
2337 18. Changes to the handling of Unicode character properties:
2338
2339 (a) Updated the table to Unicode 4.1.0.
2340
2341 (b) Recognize characters that are not in the table as "Cn" (undefined).
2342
2343 (c) I revised the way the table is implemented to a much improved format
2344 which includes recognition of ranges. It now supports the ranges that
2345 are defined in UnicodeData.txt, and it also amalgamates other
2346 characters into ranges. This has reduced the number of entries in the
2347 table from around 16,000 to around 3,000, thus reducing its size
2348 considerably. I realized I did not need to use a tree structure after
2349 all - a binary chop search is just as efficient. Having reduced the
2350 number of entries, I extended their size from 6 bytes to 8 bytes to
2351 allow for more data.
2352
2353 (d) Added support for Unicode script names via properties such as \p{Han}.
2354
2355 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2356 matching that character.
2357
2358 20. When matching a repeated Unicode property with a minimum greater than zero,
2359 (for example \pL{2,}), PCRE could look past the end of the subject if it
2360 reached it while seeking the minimum number of characters. This could
2361 happen only if some of the characters were more than one byte long, because
2362 there is a check for at least the minimum number of bytes.
2363
2364 21. Refactored the implementation of \p and \P so as to be more general, to
2365 allow for more different types of property in future. This has changed the
2366 compiled form incompatibly. Anybody with saved compiled patterns that use
2367 \p or \P will have to recompile them.
2368
2369 22. Added "Any" and "L&" to the supported property types.
2370
2371 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2372 but give a compile time error if the value is greater than 0xff.
2373
2374 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2375 accidentally not being installed or uninstalled.
2376
2377 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2378 made were to insert the current release number. This seemed silly, because
2379 it made things harder for people building PCRE on systems that don't run
2380 "configure". I have turned pcre.h into a distributed file, no longer built
2381 by "configure", with the version identification directly included. There is
2382 no longer a pcre.h.in file.
2383
2384 However, this change necessitated a change to the pcre-config script as
2385 well. It is built from pcre-config.in, and one of the substitutions was the
2386 release number. I have updated configure.ac so that ./configure now finds
2387 the release number by grepping pcre.h.
2388
2389 26. Added the ability to run the tests under valgrind.
2390
2391
2392 Version 6.4 05-Sep-05
2393 ---------------------
2394
2395 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2396 "--" to be printed when multiple files were scanned, even when none of the
2397 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2398 consider it to be a bug, and have restored the previous behaviour.
2399
2400 2. A couple of code tidies to get rid of compiler warnings.
2401
2402 3. The pcretest program used to cheat by referring to symbols in the library
2403 whose names begin with _pcre_. These are internal symbols that are not
2404 really supposed to be visible externally, and in some environments it is
2405 possible to suppress them. The cheating is now confined to including
2406 certain files from the library's source, which is a bit cleaner.
2407
2408 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2409 file's purpose clearer.
2410
2411 5. Reorganized pcre_ucp_findchar().
2412
2413
2414 Version 6.3 15-Aug-05
2415 ---------------------
2416
2417 1. The file libpcre.pc.in did not have general read permission in the tarball.
2418
2419 2. There were some problems when building without C++ support:
2420
2421 (a) If C++ support was not built, "make install" and "make test" still
2422 tried to test it.
2423
2424 (b) There were problems when the value of CXX was explicitly set. Some
2425 changes have been made to try to fix these, and ...
2426
2427 (c) --disable-cpp can now be used to explicitly disable C++ support.
2428
2429 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2430 backslash in a target when C++ was disabled. This confuses some
2431 versions of "make", apparently. Using an intermediate variable solves
2432 this. (Same for CPP_LOBJ.)
2433
2434 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2435 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2436 necessary on certain architectures.
2437
2438 4. Added a setting of -export-symbols-regex to the link command to remove
2439 those symbols that are exported in the C sense, but actually are local
2440 within the library, and not documented. Their names all begin with
2441 "_pcre_". This is not a perfect job, because (a) we have to except some
2442 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2443 available (and never for static libraries). I have made a note to try to
2444 find a way round (a) in the future.
2445
2446
2447 Version 6.2 01-Aug-05
2448 ---------------------
2449
2450 1. There was no test for integer overflow of quantifier values. A construction
2451 such as {1111111111111111} would give undefined results. What is worse, if
2452 a minimum quantifier for a parenthesized subpattern overflowed and became
2453 negative, the calculation of the memory size went wrong. This could have
2454 led to memory overwriting.
2455
2456 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2457
2458 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2459 operating environments where this matters.
2460
2461 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2462 PCRE options from within the C++ wrapper.
2463
2464 5. Named capturing subpatterns were not being correctly counted when a pattern
2465 was compiled. This caused two problems: (a) If there were more than 100
2466 such subpatterns, the calculation of the memory needed for the whole
2467 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2468 back references of the form \12, where the number was greater than 9, were
2469 not recognized as back references, even though there were sufficient
2470 previous subpatterns.
2471
2472 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2473 versions of gcc, e.g. 2.95.4.
2474
2475
2476 Version 6.1 21-Jun-05
2477 ---------------------
2478
2479 1. There was one reference to the variable "posix" in pcretest.c that was not
2480 surrounded by "#if !defined NOPOSIX".
2481
2482 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2483 the cross-check on the old pcre_info() function, for the benefit of the
2484 cut-down version of PCRE that is currently imported into Exim.
2485
2486 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2487 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2488 patterns that start (?i) but I don't think that matters. The use of (?i) is
2489 just an example; this all applies to the other options as well.
2490
2491 4. Since libtool seems to echo the compile commands it is issuing, the output
2492 from "make" can be reduced a bit by putting "@" in front of each libtool
2493 compile command.
2494
2495 5. Patch from the folks at Google for configure.in to be a bit more thorough
2496 in checking for a suitable C++ installation before trying to compile the
2497 C++ stuff. This should fix a reported problem when a compiler was present,
2498 but no suitable headers.
2499
2500 6. The man pages all had just "PCRE" as their title. I have changed them to
2501 be the relevant file name. I have also arranged that these names are
2502 retained in the file doc/pcre.txt, which is a concatenation in text format
2503 of all the man pages except the little individual ones for each function.
2504
2505 7. The NON-UNIX-USE file had not been updated for the different set of source
2506 files that come with release 6. I also added a few comments about the C++
2507 wrapper.
2508
2509
2510 Version 6.0 07-Jun-05
2511 ---------------------
2512
2513 1. Some minor internal re-organization to help with my DFA experiments.
2514
2515 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2516 didn't matter for the library itself when fully configured, but did matter
2517 when compiling without UCP support, or within Exim, where the ucp files are
2518 not imported.
2519
2520 3. Refactoring of the library code to split up the various functions into
2521 different source modules. The addition of the new DFA matching code (see
2522 below) to a single monolithic source would have made it really too
2523 unwieldy, quite apart from causing all the code to be include in a
2524 statically linked application, when only some functions are used. This is
2525 relevant even without the DFA addition now that patterns can be compiled in
2526 one application and matched in another.
2527
2528 The downside of splitting up is that there have to be some external
2529 functions and data tables that are used internally in different modules of
2530 the library but which are not part of the API. These have all had their
2531 names changed to start with "_pcre_" so that they are unlikely to clash
2532 with other external names.
2533
2534 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2535 a different (DFA) algorithm. Although it is slower than the original
2536 function, it does have some advantages for certain types of matching
2537 problem.
2538
2539 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2540 including restarting after a partial match.
2541
2542 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2543 defined when compiling for Windows was sent to me. I have put it into the
2544 code, though I have no means of testing or verifying it.
2545
2546 7. Added the pcre_refcount() auxiliary function.
2547
2548 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2549 match before or at the first newline in the subject string. In pcretest,
2550 the /f option on a pattern can be used to set this.
2551
2552 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2553 would behave wrongly. This has been present in PCRE since release 4.0.
2554
2555 10. A number of changes to the pcregrep command:
2556
2557 (a) Refactored how -x works; insert ^(...)$ instead of setting
2558 PCRE_ANCHORED and checking the length, in preparation for adding
2559 something similar for -w.
2560
2561 (b) Added the -w (match as a word) option.
2562
2563 (c) Refactored the way lines are read and buffered so as to have more
2564 than one at a time available.
2565
2566 (d) Implemented a pcregrep test script.
2567
2568 (e) Added the -M (multiline match) option. This allows patterns to match
2569 over several lines of the subject. The buffering ensures that at least
2570 8K, or the rest of the document (whichever is the shorter) is available
2571 for matching (and similarly the previous 8K for lookbehind assertions).
2572
2573 (f) Changed the --help output so that it now says
2574
2575 -w, --word-regex(p)
2576
2577 instead of two lines, one with "regex" and the other with "regexp"
2578 because that confused at least one person since the short forms are the
2579 same. (This required a bit of code, as the output is generated
2580 automatically from a table. It wasn't just a text change.)
2581
2582 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2583 option but starts with a hyphen. Could be a pattern or a path name
2584 starting with a hyphen, for instance.
2585
2586 (h) "-" can be given as a file name to represent stdin.
2587
2588 (i) When file names are being printed, "(standard input)" is used for
2589 the standard input, for compatibility with GNU grep. Previously
2590 "<stdin>" was used.
2591
2592 (j) The option --label=xxx can be used to supply a name to be used for
2593 stdin when file names are being printed. There is no short form.
2594
2595 (k) Re-factored the options decoding logic because we are going to add
2596 two more options that take data. Such options can now be given in four
2597 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2598
2599 (l) Added the -A, -B, and -C options for requesting that lines of context
2600 around matches be printed.
2601
2602 (m) Added the -L option to print the names of files that do not contain
2603 any matching lines, that is, the complement of -l.
2604
2605 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2606 continue to scan other files.
2607
2608 (o) The -s option was incorrectly implemented. For compatibility with other
2609 greps, it now suppresses the error message for a non-existent or non-
2610 accessible file (but not the return code). There is a new option called
2611 -q that suppresses the output of matching lines, which was what -s was
2612 previously doing.
2613
2614 (p) Added --include and --exclude options to specify files for inclusion
2615 and exclusion when recursing.
2616
2617 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2618 Hopefully, it now does.
2619
2620 12. Missing cast in pcre_study().
2621
2622 13. Added an "uninstall" target to the makefile.
2623
2624 14. Replaced "extern" in the function prototypes in Makefile.in with
2625 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2626 world, but is set differently for Windows.
2627
2628 15. Added a second compiling function called pcre_compile2(). The only
2629 difference is that it has an extra argument, which is a pointer to an
2630 integer error code. When there is a compile-time failure, this is set
2631 non-zero, in addition to the error test pointer being set to point to an
2632 error message. The new argument may be NULL if no error number is required
2633 (but then you may as well call pcre_compile(), which is now just a
2634 wrapper). This facility is provided because some applications need a
2635 numeric error indication, but it has also enabled me to tidy up the way
2636 compile-time errors are handled in the POSIX wrapper.
2637
2638 16. Added VPATH=.libs to the makefile; this should help when building with one
2639 prefix path and installing with another. (Or so I'm told by someone who
2640 knows more about this stuff than I do.)
2641
2642 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2643 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2644 match everything, including newlines. This is not POSIX-compatible, but
2645 somebody wanted the feature. From pcretest it can be activated by using
2646 both the P and the s flags.
2647
2648 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2649
2650 19. libpcre.pc was being incorrectly installed as executable.
2651
2652 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2653 it now also looks for '\r' so that it will work unmodified on Windows.
2654
2655 21. Added Google's contributed C++ wrapper to the distribution.
2656
2657 22. Added some untidy missing memory free() calls in pcretest, to keep
2658 Electric Fence happy when testing.
2659
2660
2661
2662 Version 5.0 13-Sep-04
2663 ---------------------
2664
2665 1. Internal change: literal characters are no longer packed up into items
2666 containing multiple characters in a single byte-string. Each character
2667 is now matched using a separate opcode. However, there may be more than one
2668 byte in the character in UTF-8 mode.
2669
2670 2. The pcre_callout_block structure has two new fields: pattern_position and
2671 next_item_length. These contain the offset in the pattern to the next match
2672 item, and its length, respectively.
2673
2674 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2675 insertion of callouts before each pattern item. Added the /C option to
2676 pcretest to make use of this.
2677
2678 4. On the advice of a Windows user, the lines
2679
2680 #if defined(_WIN32) || defined(WIN32)
2681 _setmode( _fileno( stdout ), 0x8000 );
2682 #endif /* defined(_WIN32) || defined(WIN32) */
2683
2684 have been added to the source of pcretest. This apparently does useful
2685 magic in relation to line terminators.
2686
2687 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2688 for the benefit of those environments where the "b" makes a difference.
2689
2690 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2691 to know about it. I have put a hack into configure.in that adds in code
2692 to set GCC=yes if CC=icc. This seems to end up at a point in the
2693 generated configure script that is early enough to affect the setting of
2694 compiler options, which is what is needed, but I have no means of testing
2695 whether it really works. (The user who reported this had patched the
2696 generated configure script, which of course I cannot do.)
2697
2698 LATER: After change 22 below (new libtool files), the configure script
2699 seems to know about icc (and also ecc). Therefore, I have commented out
2700 this hack in configure.in.
2701
2702 7. Added support for pkg-config (2 patches were sent in).
2703
2704 8. Negated POSIX character classes that used a combination of internal tables
2705 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2706 [[:^ascii]]. Typically, they would match almost any characters. The other
2707 POSIX classes were not broken in this way.
2708
2709 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2710 to find the match, as PCRE was deluded into thinking that the match had to
2711 start at the start point or following a newline. The same bug applied to
2712 patterns with negative forward assertions or any backward assertions
2713 preceding ".*" at the start, unless the pattern required a fixed first
2714 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2715
2716 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2717 starting at the last subject character, bytes beyond the end of the subject
2718 string were read.
2719
2720 11. Renamed the variable "class" as "classbits" to make life easier for C++
2721 users. (Previously there was a macro definition, but it apparently wasn't
2722 enough.)
2723
2724 12. Added the new field "tables" to the extra data so that tables can be passed
2725 in at exec time, or the internal tables can be re-selected. This allows
2726 a compiled regex to be saved and re-used at a later time by a different
2727 program that might have everything at different addresses.
2728
2729 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2730 -R library as well as a -L library.
2731
2732 14. The debugging options of pcretest (-d on the command line or D on a
2733 pattern) showed incorrect output for anything following an extended class
2734 that contained multibyte characters and which was followed by a quantifier.
2735
2736 15. Added optional support for general category Unicode character properties
2737 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2738 support. It adds about 90K to the size of the library. The meanings of the
2739 inbuilt class escapes such as \d and \s have NOT been changed.
2740
2741 16. Updated pcredemo.c to include calls to free() to release the memory for the
2742 compiled pattern.
2743
2744 17. The generated file chartables.c was being created in the source directory
2745 instead of in the building directory. This caused the build to fail if the
2746 source directory was different from the building directory, and was
2747 read-only.
2748
2749 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2750 file. No doubt somebody will tell me if they don't make sense... Also added
2751 Dan Mooney's comments about building on OpenVMS.
2752
2753 19. Added support for partial matching via the PCRE_PARTIAL option for
2754 pcre_exec() and the \P data escape in pcretest.
2755
2756 20. Extended pcretest with 3 new pattern features:
2757
2758 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2759 write the compiled pattern to the file whose name is "rest-of-line".
2760 This is a straight binary dump of the data, with the saved pointer to
2761 the character tables forced to be NULL. The study data, if any, is
2762 written too. After writing, pcretest reads a new pattern.
2763
2764 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2765 compiled pattern from the given file. There must not be any
2766 occurrences of "<" in the file name (pretty unlikely); if there are,
2767 pcretest will instead treat the initial "<" as a pattern delimiter.
2768 After reading in the pattern, pcretest goes on to read data lines as
2769 usual.
2770
2771 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2772 and 16-bit fields in a compiled pattern, to simulate a pattern that
2773 was compiled on a host of opposite endianness.
2774
2775 21. The pcre-exec() function can now cope with patterns that were compiled on
2776 hosts of opposite endianness, with this restriction:
2777
2778 As for any compiled expression that is saved and used later, the tables
2779 pointer field cannot be preserved; the extra_data field in the arguments
2780 to pcre_exec() should be used to pass in a tables address if a value
2781 other than the default internal tables were used at compile time.
2782
2783 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2784 now diagnosed as an error. Previously, most of the time, a negative number
2785 would have been treated as zero, but if in addition "ovector" was passed as
2786 NULL, a crash could occur.
2787
2788 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2789 new versions from the libtool 1.5 distribution (the last one is a copy of
2790 a file called libtool.m4). This seems to have fixed the need to patch
2791 "configure" to support Darwin 1.3 (which I used to do). However, I still
2792 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2793 workstation).
2794
2795 24. Changed the PCRE licence to be the more standard "BSD" licence.
2796
2797
2798 Version 4.5 01-Dec-03
2799 ---------------------
2800
2801 1. There has been some re-arrangement of the code for the match() function so
2802 that it can be compiled in a version that does not call itself recursively.
2803 Instead, it keeps those local variables that need separate instances for
2804 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2805 needs to "recurse". Keeping track of where control must go is done by means
2806 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2807 hide most of the details from the main code, and operates only if
2808 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2809 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2810 operating.
2811
2812 To make it easier for callers to provide specially tailored get/free
2813 functions for this usage, two new functions, pcre_stack_malloc, and
2814 pcre_stack_free, are used. They are always called in strict stacking order,
2815 and the size of block requested is always the same.
2816
2817 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2818 PCRE has been compiled to use the stack or the heap for recursion. The
2819 -C option of pcretest uses this to show which version is compiled.
2820
2821 A new data escape \S, is added to pcretest; it causes the amounts of store
2822 obtained and freed by both kinds of malloc/free at match time to be added
2823 to the output.
2824
2825 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2826 what's available on my current Linux desktop machine.
2827
2828 3. When matching a UTF-8 string, the test for a valid string at the start has
2829 been extended. If start_offset is not zero, PCRE now checks that it points
2830 to a byte that is the start of a UTF-8 character. If not, it returns
2831 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2832 this is necessary because there may be backward assertions in the pattern.
2833 When matching the same subject several times, it may save resources to use
2834 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2835
2836 4. The code for checking the validity of UTF-8 strings has been tightened so
2837 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2838 containing "overlong sequences".
2839
2840 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2841 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2842 should have been "&", but it just so happened that all the cases this let
2843 through by mistake were picked up later in the function.
2844
2845 6. I had used a variable called "isblank" - this is a C99 function, causing
2846 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2847
2848 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2849 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2850 so the version is shown; (c) stop "make test" echoing "./RunTest".
2851
2852 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2853
2854 9. The prototype for memmove() for systems that don't have it was using
2855 size_t, but the inclusion of the header that defines size_t was later. I've
2856 moved the #includes for the C headers earlier to avoid this.
2857
2858 10. Added some adjustments to the code to make it easier to compiler on certain
2859 special systems:
2860
2861 (a) Some "const" qualifiers were missing.
2862 (b) Added the macro EXPORT before all exported functions; by default this
2863 is defined to be empty.
2864 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2865 that it reads its output file name as an argument instead of writing
2866 to the standard output and assuming this can be redirected.
2867
2868 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2869 class containing characters with values greater than 255, PCRE compilation
2870 went into a loop.
2871
2872 12. A recursive reference to a subpattern that was within another subpattern
2873 that had a minimum quantifier of zero caused PCRE to crash. For example,
2874 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2875 recursion. If the recursively-called subpattern itself had a zero repeat,
2876 that was OK.
2877
2878 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2879 buffer into which it was copied (for escape processing) was still set at
2880 1024, so long lines caused crashes.
2881
2882 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2883 "internal error: code overflow...". This applied to any character class
2884 that was followed by a possessive quantifier.
2885
2886 15. Modified the Makefile to add libpcre.la as a prerequisite for
2887 libpcreposix.la because I was told this is needed for a parallel build to
2888 work.
2889
2890 16. If a pattern that contained .* following optional items at the start was
2891 studied, the wrong optimizing data was generated, leading to matching
2892 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2893 matching string must start with a or b or c. The correct conclusion for
2894 this pattern is that a match can start with any character.
2895
2896
2897 Version 4.4 13-Aug-03
2898 ---------------------
2899
2900 1. In UTF-8 mode, a character class containing characters with values between
2901 127 and 255 was not handled correctly if the compiled pattern was studied.
2902 In fixing this, I have also improved the studying algorithm for such
2903 classes (slightly).
2904
2905 2. Three internal functions had redundant arguments passed to them. Removal
2906 might give a very teeny performance improvement.
2907
2908 3. Documentation bug: the value of the capture_top field in a callout is *one
2909 more than* the number of the hightest numbered captured substring.
2910
2911 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2912 in incorrectly linking with a previously installed version. They now link
2913 explicitly with libpcre.la.
2914
2915 5. configure.in no longer needs to recognize Cygwin specially.
2916
2917 6. A problem in pcre.in for Windows platforms is fixed.
2918
2919 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2920 pcretest, it used to include the size of the study block as part of its
2921 output. Unfortunately, the structure contains a field that has a different
2922 size on different hardware architectures. This meant that the tests that
2923 showed this size failed. As the block is currently always of a fixed size,
2924 this information isn't actually particularly useful in pcretest output, so
2925 I have just removed it.
2926
2927 8. Three pre-processor statements accidentally did not start in column 1.
2928 Sadly, there are *still* compilers around that complain, even though
2929 standard C has not required this for well over a decade. Sigh.
2930
2931 9. In pcretest, the code for checking callouts passed small integers in the
2932 callout_data field, which is a void * field. However, some picky compilers
2933 complained about the casts involved for this on 64-bit systems. Now
2934 pcretest passes the address of the small integer instead, which should get
2935 rid of the warnings.
2936
2937 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2938 both compile and run time, and gives an error if an invalid UTF-8 sequence
2939 is found. There is a option for disabling this check in cases where the
2940 string is known to be correct and/or the maximum performance is wanted.
2941
2942 11. In response to a bug report, I changed one line in Makefile.in from
2943
2944 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2945 to
2946 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2947
2948 to look similar to other lines, but I have no way of telling whether this
2949 is the right thing to do, as I do not use Windows. No doubt I'll get told
2950 if it's wrong...
2951
2952
2953 Version 4.3 21-May-03
2954 ---------------------
2955
2956 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2957 Makefile.
2958
2959 2. Some refactoring to improve the quality of the code:
2960
2961 (i) The utf8_table... variables are now declared "const".
2962
2963 (ii) The code for \cx, which used the "case flipping" table to upper case
2964 lower case letters, now just substracts 32. This is ASCII-specific,
2965 but the whole concept of \cx is ASCII-specific, so it seems
2966 reasonable.
2967
2968 (iii) PCRE was using its character types table to recognize decimal and
2969 hexadecimal digits in the pattern. This is silly, because it handles
2970 only 0-9, a-f, and A-F, but the character types table is locale-
2971 specific, which means strange things might happen. A private
2972 table is now used for this - though it costs 256 bytes, a table is
2973 much faster than multiple explicit tests. Of course, the standard
2974 character types table is still used for matching digits in subject
2975 strings against \d.
2976
2977 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2978 ending in _t are). So I've renamed it as ESC_tee.
2979
2980 3. The first argument for regexec() in the POSIX wrapper should have been
2981 defined as "const".
2982
2983 4. Changed pcretest to use malloc() for its buffers so that they can be
2984 Electric Fenced for debugging.
2985
2986 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2987 to read one or more bytes before the start of the subject string. Often this
2988 had no effect on PCRE's behaviour, but in some circumstances it could
2989 provoke a segmentation fault.
2990
2991 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2992 to try to read one or more bytes before the start of the subject string.
2993
2994 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2995 UTF-8 support could misbehave in various ways if the subject string
2996 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2997 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2998 back over UTF-8 characters.)
2999
3000
3001 Version 4.2 14-Apr-03
3002 ---------------------
3003
3004 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3005
3006 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3007 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3008 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3009 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3010 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3011 and BUILD_EXEEXT
3012 Note: automatic setting of the BUILD variables is not yet working
3013 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3014 compile-time but not at link-time
3015 [LINK]: use for linking executables only
3016 make different versions for Windows and non-Windows
3017 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3018 libraries
3019 [LINK_FOR_BUILD]: new variable
3020 [OBJEXT]: use throughout
3021 [EXEEXT]: use throughout
3022 <winshared>: new target
3023 <wininstall>: new target
3024 <dftables.o>: use native compiler
3025 <dftables>: use native linker
3026 <install>: handle Windows platform correctly
3027 <clean>: ditto
3028 <check>: ditto
3029 copy DLL to top builddir before testing
3030
3031 As part of these changes, -no-undefined was removed again. This was reported
3032 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3033 in any case.
3034
3035 3. Some tidies to get rid of compiler warnings:
3036
3037 . In the match_data structure, match_limit was an unsigned long int, whereas
3038 match_call_count was an int. I've made them both unsigned long ints.
3039
3040 . In pcretest the fact that a const uschar * doesn't automatically cast to
3041 a void * provoked a warning.
3042
3043 . Turning on some more compiler warnings threw up some "shadow" variables
3044 and a few more missing casts.
3045
3046 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3047 option, a class that contained a single character with a value between 128
3048 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3049
3050 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3051 option, a class that contained several characters, but with at least one
3052 whose value was between 128 and 255 caused PCRE to crash.
3053
3054
3055 Version 4.1 12-Mar-03
3056 ---------------------
3057
3058 1. Compiling with gcc -pedantic found a couple of places where casts were
3059 needed, and a string in dftables.c that was longer than standard compilers are
3060 required to support.
3061
3062 2. Compiling with Sun's compiler found a few more places where the code could
3063 be tidied up in order to avoid warnings.
3064
3065 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3066 first of these names is deprecated in the latest Autoconf in favour of the name
3067 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3068 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3069 analogy I have changed it to CFLAGS_FOR_BUILD.
3070
3071 4. Added -no-undefined to the linking command in the Makefile, because this is
3072 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3073 linking step for the pcreposix library.
3074
3075 5. PCRE was failing to diagnose the case of two named groups with the same
3076 name.
3077
3078 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3079 literal character that is needed in the subject for a match, and scans along to
3080 ensure that it is present before embarking on the full matching process. This
3081 saves time in cases of nested unlimited repeats that are never going to match.
3082 Problem: the scan can take a lot of time if the subject is very long (e.g.
3083 megabytes), thus penalizing straightforward matches. It is now done only if the
3084 amount of subject to be scanned is less than 1000 bytes.
3085
3086 7. A lesser problem with the same optimization is that it was recording the
3087 first character of an anchored pattern as "needed", thus provoking a search
3088 right along the subject, even when the first match of the pattern was going to
3089 fail. The "needed" character is now not set for anchored patterns, unless it
3090 follows something in the pattern that is of non-fixed length. Thus, it still
3091 fulfils its original purpose of finding quick non-matches in cases of nested
3092 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3093
3094
3095 Version 4.0 17-Feb-03
3096 ---------------------
3097
3098 1. If a comment in an extended regex that started immediately after a meta-item
3099 extended to the end of string, PCRE compiled incorrect data. This could lead to
3100 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3101
3102 2. Moved to autoconf 2.53 and libtool 1.4.2.
3103
3104 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3105 the special perltest8 script is no longer needed - all the tests can be run
3106 from a single perltest script.
3107
3108 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3109 by \s. It has now been removed in PCRE. This means it isn't recognized as
3110 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3111 class [:space:] *does* include VT, thereby creating a mess.
3112
3113 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3114 space and tab.
3115
3116 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3117 its new features into the main test script, reducing the number of scripts.
3118
3119 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3120 were backward compatible, and made the (?i) apply to the whole pattern, as if
3121 /i were given. Now it behaves more logically, and applies the option setting
3122 only to what follows. PCRE has been changed to follow suit. However, if it
3123 finds options settings right at the start of the pattern, it extracts them into
3124 the global options, as before. Thus, they show up in the info data.
3125
3126 8. Added support for the \Q...\E escape sequence. Characters in between are
3127 treated as literals. This is slightly different from Perl in that $ and @ are
3128 also handled as literals inside the quotes. In Perl, they will cause variable
3129 interpolation. Note the following examples:
3130
3131 Pattern PCRE matches Perl matches
3132
3133 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3134 \Qabc\$xyz\E abc\$xyz abc\$xyz
3135 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3136
3137 For compatibility with Perl, \Q...\E sequences are recognized inside character
3138 classes as well as outside them.
3139
3140 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3141 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3142 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3143 signed/unsigned warnings.
3144
3145 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3146 option for pcretest, so I've replaced it by a simple function that does just
3147 that job.
3148
3149 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3150 "pcregrep -".
3151
3152 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3153 Java package. This provides some syntactic sugar for simple cases of what my
3154 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3155 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3156 item, you can use this simplified notation. Note that only makes sense with
3157 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3158 greediness, whatever the setting of the PCRE_UNGREEDY option.
3159
3160 13. A change of greediness default within a pattern was not taking effect at
3161 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3162 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3163 was abstracted outside.
3164
3165 14. PCRE now supports the \G assertion. It is true when the current matching
3166 position is at the start point of the match. This differs from \A when the
3167 starting offset is non-zero. Used with the /g option of pcretest (or similar
3168 code), it works in the same way as it does for Perl's /g option. If all
3169 alternatives of a regex begin with \G, the expression is anchored to the start
3170 match position, and the "anchored" flag is set in the compiled expression.
3171
3172 15. Some bugs concerning the handling of certain option changes within patterns
3173 have been fixed. These applied to options other than (?ims). For example,
3174 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3175 been the other way round. Some of this was related to change 7 above.
3176
3177 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3178 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3179 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3180 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3181
3182 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3183 mode. Unlike ".", it always matches newline, whatever the setting of
3184 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3185 assertions. Perl allows it, but it doesn't (in general) work because it can't
3186 calculate the length of the lookbehind. At least, that's the case for Perl
3187 5.8.0 - I've been told they are going to document that it doesn't work in
3188 future.
3189
3190 18. Added an error diagnosis for escapes that PCRE does not support: these are
3191 \L, \l, \N, \P, \p, \U, \u, and \X.
3192
3193 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3194 reading past the end of the pattern in cases such as /[abcd/.
3195
3196 20. PCRE was getting more memory than necessary for patterns with classes that
3197 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3198
3199 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3200 compiling PCRE for use with Virtual Pascal.
3201
3202 22. Small fix to the Makefile to make it work properly if the build is done
3203 outside the source tree.
3204
3205 23. Added a new extension: a condition to go with recursion. If a conditional
3206 subpattern starts with (?(R) the "true" branch is used if recursion has
3207 happened, whereas the "false" branch is used only at the top level.
3208
3209 24. When there was a very long string of literal characters (over 255 bytes
3210 without UTF support, over 250 bytes with UTF support), the computation of how
3211 much memory was required could be incorrect, leading to segfaults or other
3212 strange effects.
3213
3214 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3215 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3216 there was a subsequent back reference to those brackets. This meant that, for
3217 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3218 possible to check for precisely this case. All we can do is abandon the
3219 optimization if .* occurs inside capturing brackets when there are any back
3220 references whatsoever. (See below for a better fix that came later.)
3221
3222 26. The handling of the optimization for finding the first character of a
3223 non-anchored pattern, and for finding a character that is required later in the
3224 match were failing in some cases. This didn't break the matching; it just
3225 failed to optimize when it could. The way this is done has been re-implemented.
3226
3227 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3228
3229 28. Added a new feature that provides some of the functionality that Perl
3230 provides with (?{...}). The facility is termed a "callout". The way it is done
3231 in PCRE is for the caller to provide an optional function, by setting
3232 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3233 global variable. By default it is unset, which disables all calling out. To get
3234 the function called, the regex must include (?C) at appropriate points. This
3235 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3236 This provides a means of identifying different callout points. When PCRE
3237 reaches such a point in the regex, if pcre_callout has been set, the external
3238 function is called. It is provided with data in a structure called
3239 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3240 matching continues; if it returns a non-zero value, the match at the current
3241 point fails. However, backtracking will occur if possible. [This was changed
3242 later and other features added - see item 49 below.]
3243
3244 29. pcretest is upgraded to test the callout functionality. It provides a
3245 callout function that displays information. By default, it shows the start of
3246 the match and the current position in the text. There are some new data escapes
3247 to vary what happens:
3248
3249 \C+ in addition, show current contents of captured substrings
3250 \C- do not supply a callout function
3251 \C!n return 1 when callout number n is reached
3252 \C!n!m return 1 when callout number n is reached for the mth time
3253
3254 30. If pcregrep was called with the -l option and just a single file name, it
3255 output "<stdin>" if a match was found, instead of the file name.
3256
3257 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3258 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3259 pcre_exec(). This saves a malloc/free per call. The default value of
3260 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3261 when configuring.
3262
3263 32. The default maximum size of a compiled pattern is 64K. There have been a
3264 few cases of people hitting this limit. The code now uses macros to handle the
3265 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3266 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3267 configuring. Tests 2 and 5 work only with 2-byte links because they output
3268 debugging information about compiled patterns.
3269
3270 33. Internal code re-arrangements:
3271
3272 (a) Moved the debugging function for printing out a compiled regex into
3273 its own source file (printint.c) and used #include to pull it into
3274 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3275 separate copies.
3276
3277 (b) Defined the list of op-code names for debugging as a macro in
3278 internal.h so that it is next to the definition of the opcodes.
3279
3280 (c) Defined a table of op-code lengths for simpler skipping along compiled
3281 code. This is again a macro in internal.h so that it is next to the
3282 definition of the opcodes.
3283
3284 34. Added support for recursive calls to individual subpatterns, along the
3285 lines of Robin Houston's patch (but implemented somewhat differently).
3286
3287 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3288 allow it to read and process whole directories in Win32. This code was
3289 contributed by Lionel Fourquaux; it has not been tested by me.
3290
3291 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3292 used to name a group. Names consist of alphanumerics and underscores, and must
3293 be unique. Back references use the syntax (?P=name) and recursive calls use
3294 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3295 numbers. The function pcre_fullinfo() can be used after compilation to extract
3296 a name/number map. There are three relevant calls:
3297
3298 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3299 PCRE_INFO_NAMECOUNT yields the number of entries
3300 PCRE_INFO_NAMETABLE yields a pointer to the map.
3301
3302 The map is a vector of fixed-size entries. The size of each entry depends on
3303 the length of the longest name used. The first two bytes of each entry are the
3304 group number, most significant byte first. There follows the corresponding
3305 name, zero terminated. The names are in alphabetical order.
3306
3307 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3308 case instead of 255. Making it the same both with and without UTF-8 support
3309 means that the same test output works with both.
3310
3311 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3312 calling malloc() with a zero argument.
3313
3314 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3315 optimization. I've improved things by keeping a bitmap of backreferences with
3316 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3317 fact referenced, the optimization can be applied. It is unlikely that a
3318 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3319 the match to follow \n) will appear inside brackets with a number greater than
3320 31, but if it does, any back reference > 31 suppresses the optimization.
3321
3322 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3323 of disabling numbered capturing parentheses. Any opening parenthesis that is
3324 not followed by ? behaves as if it were followed by ?: but named parentheses
3325 can still be used for capturing (and they will acquire numbers in the usual
3326 way).
3327
3328 41. Redesigned the return codes from the match() function into yes/no/error so
3329 that errors can be passed back from deep inside the nested calls. A malloc
3330 failure while inside a recursive subpattern call now causes the
3331 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3332
3333 42. It is now possible to set a limit on the number of times the match()
3334 function is called in a call to pcre_exec(). This facility makes it possible to
3335 limit the amount of recursion and backtracking, though not in a directly
3336 obvious way, because the match() function is used in a number of different
3337 circumstances. The count starts from zero for each position in the subject
3338 string (for non-anchored patterns). The default limit is, for compatibility, a
3339 large number, namely 10 000 000. You can change this in two ways:
3340
3341 (a) When configuring PCRE before making, you can use --with-match-limit=n
3342 to set a default value for the compiled library.
3343
3344 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3345 a different value is set. See 45 below.
3346
3347 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3348
3349 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3350 of things that can be changed at compile time. The first argument specifies
3351 what is wanted and the second points to where the information is to be placed.
3352 The current list of available information is:
3353
3354 PCRE_CONFIG_UTF8
3355
3356 The output is an integer that is set to one if UTF-8 support is available;
3357 otherwise it is set to zero.
3358
3359 PCRE_CONFIG_NEWLINE
3360
3361 The output is an integer that it set to the value of the code that is used for
3362 newline. It is either LF (10) or CR (13).
3363
3364 PCRE_CONFIG_LINK_SIZE
3365
3366 The output is an integer that contains the number of bytes used for internal
3367 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3368
3369 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3370
3371 The output is an integer that contains the threshold above which the POSIX
3372 interface uses malloc() for output vectors. See item 31 above.
3373
3374 PCRE_CONFIG_MATCH_LIMIT
3375
3376 The output is an unsigned integer that contains the default limit of the number
3377 of match() calls in a pcre_exec() execution. See 42 above.
3378
3379 44. pcretest has been upgraded by the addition of the -C option. This causes it
3380 to extract all the available output from the new pcre_config() function, and to
3381 output it. The program then exits immediately.
3382
3383 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3384 order to support additional features. One way would have been to define
3385 pcre_exec2() (for example) with extra arguments, but this would not have been
3386 extensible, and would also have required all calls to the original function to
3387 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3388 is used for passing in "extra" data from pcre_study().
3389
3390 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3391 contains the following fields:
3392
3393 flags a bitmap indicating which of the following fields are set
3394 study_data opaque data from pcre_study()
3395 match_limit a way of specifying a limit on match() calls for a specific
3396 call to pcre_exec()
3397 callout_data data for callouts (see 49 below)
3398
3399 The flag bits are also defined in pcre.h, and are
3400
3401 PCRE_EXTRA_STUDY_DATA
3402 PCRE_EXTRA_MATCH_LIMIT
3403 PCRE_EXTRA_CALLOUT_DATA
3404
3405 The pcre_study() function now returns one of these new pcre_extra blocks, with
3406 the actual study data pointed to by the study_data field, and the
3407 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3408 before. That is, this change is entirely upwards-compatible and requires no
3409 change to existing code.
3410
3411 If you want to pass in additional data to pcre_exec(), you can either place it
3412 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3413 block.
3414
3415 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3416 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3417 times with different match limits, until it finds the minimum value needed for
3418 pcre_exec() to complete. The value is then output. This can be instructive; for
3419 most simple matches the number is quite small, but for pathological cases it
3420 gets very large very quickly.
3421
3422 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3423 returns the size of the data block pointed to by the study_data field in a
3424 pcre_extra block, that is, the value that was passed as the argument to
3425 pcre_malloc() when PCRE was getting memory in which to place the information
3426 created by pcre_study(). The fourth argument should point to a size_t variable.
3427 pcretest has been extended so that this information is shown after a successful
3428 pcre_study() call when information about the compiled regex is being displayed.
3429
3430 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3431 because what follows is always an absolute path. (Later: it turns out that this
3432 is more than cosmetic for MinGW, because it doesn't like empty path
3433 components.)
3434
3435 49. Some changes have been made to the callout feature (see 28 above):
3436
3437 (i) A callout function now has three choices for what it returns:
3438
3439 0 => success, carry on matching
3440 > 0 => failure at this point, but backtrack if possible
3441 < 0 => serious error, return this value from pcre_exec()
3442
3443 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3444 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3445 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3446 use by callout functions. It will never be used by PCRE itself.
3447
3448 (ii) The pcre_extra structure (see 45 above) has a void * field called
3449 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3450 pcre_callout_block structure has a field of the same name. The contents of
3451 the field passed in the pcre_extra structure are passed to the callout
3452 function in the corresponding field in the callout block. This makes it
3453 easier to use the same callout-containing regex from multiple threads. For
3454 testing, the pcretest program has a new data escape
3455
3456 \C*n pass the number n (may be negative) as callout_data
3457
3458 If the callout function in pcretest receives a non-zero value as
3459 callout_data, it returns that value.
3460
3461 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3462 there were some redundant $(CFLAGS) in commands that are now specified as
3463 $(LINK), which already includes $(CFLAGS).
3464
3465 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3466 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3467 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3468 one-byte characters throughout. Note that case-insensitive matching applies
3469 only to characters whose values are less than 256. PCRE doesn't support the
3470 notion of cases for higher-valued characters.
3471
3472 (i) A character class whose characters are all within 0-255 is handled as
3473 a bit map, and the map is inverted for negative classes. Previously, a
3474 character > 255 always failed to match such a class; however it should
3475 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3476
3477 (ii) A negated character class with a single character < 255 is coded as
3478 "not this character" (OP_NOT). This wasn't working properly when the test
3479 character was multibyte, either singly or repeated.
3480
3481 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3482 mode, for example: \x{100}{2,3}.
3483
3484 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3485 singly or repeated) now correctly test multibyte characters. However,
3486 PCRE doesn't recognize any characters with values greater than 255 as
3487 digits, spaces, or word characters. Such characters always match \D, \S,
3488 and \W, and never match \d, \s, or \w.
3489
3490 (v) Classes may now contain characters and character ranges with values
3491 greater than 255. For example: [ab\x{100}-\x{400}].
3492
3493 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3494 PCRE in UTF-8 mode.
3495
3496 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3497 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3498 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3499 value.)
3500
3501 53. The single man page has become too large. I have therefore split it up into
3502 a number of separate man pages. These also give rise to individual HTML pages;
3503 these are now put in a separate directory, and there is an index.html page that
3504 lists them all. Some hyperlinking between the pages has been installed.
3505
3506 54. Added convenience functions for handling named capturing parentheses.
3507
3508 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3509 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3510 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3511 are faulted.
3512
3513 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3514 calling configure. These values are used when compiling the dftables.c program
3515 which is run to generate the source of the default character tables. They
3516 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3517 you will need to set these values.
3518
3519 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3520
3521
3522 Version 3.9 02-Jan-02
3523 ---------------------
3524
3525 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3526
3527 2. If --disable-static was given, the building process failed when trying to
3528 build pcretest and pcregrep. (For some reason it was using libtool to compile
3529 them, which is not right, as they aren't part of the library.)
3530
3531
3532 Version 3.8 18-Dec-01
3533 ---------------------
3534
3535 1. The experimental UTF-8 code was completely screwed up. It was packing the
3536 bytes in the wrong order. How dumb can you get?
3537
3538
3539 Version 3.7 29-Oct-01
3540 ---------------------
3541
3542 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3543 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3544 this didn't happen under Solaris 8, where I normally test things.
3545
3546 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3547 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3548 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3549 this fix an hour or so after the initial 3.7 release.)
3550
3551
3552 Version 3.6 23-Oct-01
3553 ---------------------
3554
3555 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3556 offsets passed as NULL with zero offset count.
3557
3558 2. The config.guess and config.sub files had not been updated when I moved to
3559 the latest autoconf.
3560
3561
3562 Version 3.5 15-Aug-01
3563 ---------------------
3564
3565 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3566 had been forgotten.
3567
3568 2. By using declared but undefined structures, we can avoid using "void"
3569 definitions in pcre.h while keeping the internal definitions of the structures
3570 private.
3571
3572 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3573 user point of view, this means that both static and shared libraries are built
3574 by default, but this can be individually controlled. More of the work of
3575 handling this static/shared cases is now inside libtool instead of PCRE's make
3576 file.
3577
3578 4. The pcretest utility is now installed along with pcregrep because it is
3579 useful for users (to test regexs) and by doing this, it automatically gets
3580 relinked by libtool. The documentation has been turned into a man page, so
3581 there are now .1, .txt, and .html versions in /doc.
3582
3583 5. Upgrades to pcregrep:
3584 (i) Added long-form option names like gnu grep.
3585 (ii) Added --help to list all options with an explanatory phrase.
3586 (iii) Added -r, --recursive to recurse into sub-directories.
3587 (iv) Added -f, --file to read patterns from a file.
3588
3589 6. pcre_exec() was referring to its "code" argument before testing that
3590 argument for NULL (and giving an error if it was NULL).
3591
3592 7. Upgraded Makefile.in to allow for compiling in a different directory from
3593 the source directory.
3594
3595 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3596 options bits, the pointer it was passed was to an int instead of to an unsigned
3597 long int. This mattered only on 64-bit systems.
3598
3599 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3600 generated) instead of pcre.in, which it its source. Also made the same change
3601 in several of the .c files.
3602
3603 10. A new release of gcc defines printf() as a macro, which broke pcretest
3604 because it had an ifdef in the middle of a string argument for printf(). Fixed
3605 by using separate calls to printf().
3606
3607 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3608 script, to force use of CR or LF instead of \n in the source. On non-Unix
3609 systems, the value can be set in config.h.
3610
3611 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3612 absolute limit. Changed the text of the error message to make this clear, and
3613 likewise updated the man page.
3614
3615 13. The limit of 99 on the number of capturing subpatterns has been removed.
3616 The new limit is 65535, which I hope will not be a "real" limit.
3617
3618
3619 Version 3.4 22-Aug-00
3620 ---------------------
3621
3622 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3623
3624 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3625
3626
3627 Version 3.3 01-Aug-00
3628 ---------------------
3629
3630 1. If an octal character was given, but the value was greater than \377, it
3631 was not getting masked to the least significant bits, as documented. This could
3632 lead to crashes in some systems.
3633
3634 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3635 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3636
3637 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3638 These just pass their arguments on to (pcre_free)(), but they are provided
3639 because some uses of PCRE bind it to non-C systems that can call its functions,
3640 but cannot call free() or pcre_free() directly.
3641
3642 4. Add "make test" as a synonym for "make check". Corrected some comments in
3643 the Makefile.
3644
3645 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3646 Makefile.
3647
3648 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3649 command called pgrep for grepping around the active processes.
3650
3651 7. Added the beginnings of support for UTF-8 character strings.
3652
3653 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3654 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3655 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3656 out for the ar command.)
3657
3658
3659 Version 3.2 12-May-00
3660 ---------------------
3661
3662 This is purely a bug fixing release.
3663
3664 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3665 of ZA. This was just one example of several cases that could provoke this bug,
3666 which was introduced by change 9 of version 2.00. The code for breaking
3667 infinite loops after an iteration that matches an empty string was't working
3668 correctly.
3669
3670 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3671 when matched against abbab (for example). After matching an empty string, it
3672 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3673 caused it to match further down the string than it should.
3674
3675 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3676 was there because it doesn't seem to be needed, and it causes trouble on some
3677 systems, as it is not a Standard C header. It has been removed.
3678
3679 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3680 were reported on the Macintosh. The changes were from
3681
3682 while ((c = *(++ptr)) != 0 && c != '\n');
3683 to
3684 while ((c = *(++ptr)) != 0 && c != '\n') ;
3685
3686 Totally extraordinary, but if that's what it takes...
3687
3688 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3689 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3690 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3691 assumes the way PCRE uses memmove() (always moving upwards).
3692
3693 6. PCRE is being used in one environment where strchr() is not available. There
3694 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3695 faster code anyway.
3696
3697
3698 Version 3.1 09-Feb-00
3699 ---------------------
3700
3701 The only change in this release is the fixing of some bugs in Makefile.in for
3702 the "install" target:
3703
3704 (1) It was failing to install pcreposix.h.
3705
3706 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3707
3708
3709 Version 3.0 01-Feb-00
3710 ---------------------
3711
3712 1. Add support for the /+ modifier to perltest (to output $` like it does in
3713 pcretest).
3714
3715 2. Add support for the /g modifier to perltest.
3716
3717 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3718 matches null strings.
3719
3720 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3721 pattern. Perl treats empty patterns specially - it reuses the most recent
3722 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3723 effect.
3724
3725 5. The POSIX interface was broken in that it was just handing over the POSIX
3726 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3727 required a bigger vector, with some working space on the end. This means that
3728 the POSIX wrapper now has to get and free some memory, and copy the results.
3729
3730 6. Added some simple autoconf support, placing the test data and the
3731 documentation in separate directories, re-organizing some of the
3732 information files, and making it build pcre-config (a GNU standard). Also added
3733 libtool support for building PCRE as a shared library, which is now the
3734 default.
3735
3736 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3737 09 are not valid octal constants. Single digits will be used for minor values
3738 less than 10.
3739
3740 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3741 existing programs that set these in the POSIX interface can use PCRE without
3742 modification.
3743
3744 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3745 return all that pcre_info() returns, plus additional data. The pcre_info()
3746 function is retained for compatibility, but is considered to be obsolete.
3747
3748 10. Added experimental recursion feature (?R) to handle one common case that
3749 Perl 5.6 will be able to do with (?p{...}).
3750
3751 11. Added support for POSIX character classes like [:alpha:], which Perl is
3752 adopting.
3753
3754
3755 Version 2.08 31-Aug-99
3756 ----------------------
3757
3758 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3759 trying to match at the startoffset position, but instead was moving forward to
3760 the next newline as if a previous match had failed.
3761
3762 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3763 and could get into a loop if a null string was matched other than at the start
3764 of the subject.
3765
3766 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3767 be distinguished at compile time, and for completeness also added PCRE_DATE.
3768
3769 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3770 in GnuWin32 environments.
3771
3772
3773 Version 2.07 29-Jul-99
3774 ----------------------
3775
3776 1. The documentation is now supplied in plain text form and HTML as well as in
3777 the form of man page sources.
3778
3779 2. C++ compilers don't like assigning (void *) values to other pointer types.
3780 In particular this affects malloc(). Although there is no problem in Standard
3781 C, I've put in casts to keep C++ compilers happy.
3782
3783 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3784 should be (const char *).
3785
3786 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3787 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3788 However, I haven't made this a standard facility. The documentation doesn't
3789 mention it, and the Makefile doesn't support it.
3790
3791 5. The Makefile now contains an "install" target, with editable destinations at
3792 the top of the file. The pcretest program is not installed.
3793
3794 6. pgrep -V now gives the PCRE version number and date.
3795
3796 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3797 causing the entire string to be ignored, instead of just the last character.
3798
3799 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3800 non-matching string, it can take a very, very long time, even for strings of
3801 quite modest length, because of the nested recursion. PCRE now does better in
3802 some of these cases. It does this by remembering the last required literal
3803 character in the pattern, and pre-searching the subject to ensure it is present
3804 before running the real match. In other words, it applies a heuristic to detect
3805 some types of certain failure quickly, and in the above example, if presented
3806 with a string that has no trailing " it gives "no match" very quickly.
3807
3808 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3809 other alternatives are tried instead.
3810
3811
3812 Version 2.06 09-Jun-99
3813 ----------------------
3814
3815 1. Change pcretest's output for amount of store used to show just the code
3816 space, because the remainder (the data block) varies in size between 32-bit and
3817 64-bit systems.
3818
3819 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3820 start matching at. This allows lookbehinds to work when searching for multiple
3821 occurrences in a string.
3822
3823 3. Added additional options to pcretest for testing multiple occurrences:
3824
3825 /+ outputs the rest of the string that follows a match
3826 /g loops for multiple occurrences, using the new startoffset argument
3827 /G loops for multiple occurrences by passing an incremented pointer
3828
3829 4. PCRE wasn't doing the "first character" optimization for patterns starting
3830 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3831 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3832 the letter 'x'. On long subject strings, this gives a significant speed-up.
3833
3834
3835 Version 2.05 21-Apr-99
3836 ----------------------
3837
3838 1. Changed the type of magic_number from int to long int so that it works
3839 properly on 16-bit systems.
3840
3841 2. Fixed a bug which caused patterns starting with .* not to work correctly
3842 when the subject string contained newline characters. PCRE was assuming
3843 anchoring for such patterns in all cases, which is not correct because .* will
3844 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3845 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3846 must be retried after every newline in the subject.
3847
3848
3849 Version 2.04 18-Feb-99
3850 ----------------------
3851
3852 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3853 computation of the store needed to hold the pattern was incorrect (too large).
3854 If such patterns were nested a few deep, this could multiply and become a real
3855 problem.
3856
3857 2. Added /M option to pcretest to show the memory requirement of a specific
3858 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3859
3860 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3861 compiled in such a way that the backtracking after subsequent failure was
3862 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3863 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3864
3865
3866 Version 2.03 02-Feb-99
3867 ----------------------
3868
3869 1. Fixed typo and small mistake in man page.
3870
3871 2. Added 4th condition (GPL supersedes if conflict) and created separate
3872 LICENCE file containing the conditions.
3873
3874 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3875 Perl, that is the internal \ allows the delimiter to be included in the
3876 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3877 the final delimiter, add \ to the end of the pattern (to test the error).
3878
3879 4. Added the convenience functions for extracting substrings after a successful
3880 match. Updated pcretest to make it able to test these functions.
3881
3882
3883 Version 2.02 14-Jan-99
3884 ----------------------
3885
3886 1. Initialized the working variables associated with each extraction so that
3887 their saving and restoring doesn't refer to uninitialized store.
3888
3889 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3890 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3891 fix the problem.
3892
3893 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3894 calls, and wasn't printing the correct value for compiling calls. Increased the
3895 default value of LOOPREPEAT, and the number of significant figures in the
3896 times.
3897
3898 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3899
3900 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3901 a building problem on Windows NT with a FAT file system.
3902
3903
3904 Version 2.01 21-Oct-98
3905 ----------------------
3906
3907 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3908 to character tables built by pcre_maketables() in the current locale. If NULL
3909 is passed, the default tables are used.
3910
3911
3912 Version 2.00 24-Sep-98
3913 ----------------------
3914
3915 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3916 it any more.
3917
3918 2. Allow quantification of (?>) groups, and make it work correctly.
3919
3920 3. The first character computation wasn't working for (?>) groups.
3921
3922 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3923 end of the subject) and add 5.005's \z, which really does match only at the
3924 very end of the subject.
3925
3926 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3927
3928 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3929 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3930 localized options. All options to pcre_study() were also removed.
3931
3932 7. Add other new features from 5.005:
3933
3934 $(?<= positive lookbehind
3935 $(?<! negative lookbehind
3936 (?imsx-imsx) added the unsetting capability
3937 such a setting is global if at outer level; local otherwise
3938 (?imsx-imsx:) non-capturing groups with option setting
3939 (?(cond)re|re) conditional pattern matching
3940
3941 A backreference to itself in a repeated group matches the previous
3942 captured string.
3943
3944 8. General tidying up of studying (both automatic and via "study")
3945 consequential on the addition of new assertions.
3946
3947 9. As in 5.005, unlimited repeated groups that could match an empty substring
3948 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3949 runtime if any iteration does actually match an empty substring.
3950
3951 10. Include the RunTest script in the distribution.
3952
3953 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3954 discrepancies, some of which were old and were also with respect to 5.004. They
3955 have now been fixed.
3956
3957
3958 Version 1.09 28-Apr-98
3959 ----------------------
3960
3961 1. A negated single character class followed by a quantifier with a minimum
3962 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3963 program crashes, or just wrong answers. This did not apply to negated classes
3964 containing more than one character, or to minima other than one.
3965
3966
3967 Version 1.08 27-Mar-98
3968 ----------------------
3969
3970 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3971
3972 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3973 latter must appear before anything that relies on it in the pattern.
3974
3975
3976 Version 1.07 16-Feb-98
3977 ----------------------
3978
3979 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3980 repeat of a potentially empty string).
3981
3982
3983 Version 1.06 23-Jan-98
3984 ----------------------
3985
3986 1. Added Markus Oberhumer's little patches for C++.
3987
3988 2. Literal strings longer than 255 characters were broken.
3989
3990
3991 Version 1.05 23-Dec-97
3992 ----------------------
3993
3994 1. Negated character classes containing more than one character were failing if
3995 PCRE_CASELESS was set at run time.
3996
3997
3998 Version 1.04 19-Dec-97
3999 ----------------------
4000
4001 1. Corrected the man page, where some "const" qualifiers had been omitted.
4002
4003 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4004 input syntax.
4005
4006 3. Fixed memory leak which occurred when a regex with back references was
4007 matched with an offsets vector that wasn't big enough. The temporary memory
4008 that is used in this case wasn't being freed if the match failed.
4009
4010 4. Tidied pcretest to ensure it frees memory that it gets.
4011
4012 5. Temporary memory was being obtained in the case where the passed offsets
4013 vector was exactly big enough.
4014
4015 6. Corrected definition of offsetof() from change 5 below.
4016
4017 7. I had screwed up change 6 below and broken the rules for the use of
4018 setjmp(). Now fixed.
4019
4020
4021 Version 1.03 18-Dec-97
4022 ----------------------
4023
4024 1. A erroneous regex with a missing opening parenthesis was correctly
4025 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4026 on some systems.
4027
4028 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4029 it was reported that one broken compiler failed on the former because "code" is
4030 also an independent variable.
4031
4032 3. The erroneous regex a[]b caused an array overrun reference.
4033
4034 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4035 fail on data ending with that character. (It was going on too far, and checking
4036 the next character, typically a binary zero.) This was specific to the
4037 optimized code for single-character negative classes.
4038
4039 5. Added a contributed patch from the TIN world which does the following:
4040
4041 + Add an undef for memmove, in case the the system defines a macro for it.
4042
4043 + Add a definition of offsetof(), in case there isn't one. (I don't know
4044 the reason behind this - offsetof() is part of the ANSI standard - but
4045 it does no harm).
4046
4047 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4048 most of the places where whitespace preceded '#'. I have given up and
4049 allowed the remaining 2 cases to be at the margin.
4050
4051 + Rename some variables in pcre to eliminate shadowing. This seems very
4052 pedantic, but does no harm, of course.
4053
4054 6. Moved the call to setjmp() into its own function, to get rid of warnings
4055 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4056
4057 7. Constructs such as \d{8,} were compiling into the equivalent of
4058 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4059 outcome, but in this particular case used more store than had been allocated,
4060 which caused the bug to be discovered because it threw up an internal error.
4061
4062 8. The debugging code in both pcre and pcretest for outputting the compiled
4063 form of a regex was going wrong in the case of back references followed by
4064 curly-bracketed repeats.
4065
4066
4067 Version 1.02 12-Dec-97
4068 ----------------------
4069
4070 1. Typos in pcre.3 and comments in the source fixed.
4071
4072 2. Applied a contributed patch to get rid of places where it used to remove
4073 'const' from variables, and fixed some signed/unsigned and uninitialized
4074 variable warnings.
4075
4076 3. Added the "runtest" target to Makefile.
4077
4078 4. Set default compiler flag to -O2 rather than just -O.
4079
4080
4081 Version 1.01 19-Nov-97
4082 ----------------------
4083
4084 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4085 like /([ab]*)*/, that is, for classes with more than one character in them.
4086
4087 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4088 as /((?>a*))*/ (a PCRE_EXTRA facility).
4089
4090
4091 Version 1.00 18-Nov-97
4092 ----------------------
4093
4094 1. Added compile-time macros to support systems such as SunOS4 which don't have
4095 memmove() or strerror() but have other things that can be used instead.
4096
4097 2. Arranged that "make clean" removes the executables.
4098
4099
4100 Version 0.99 27-Oct-97
4101 ----------------------
4102
4103 1. Fixed bug in code for optimizing classes with only one character. It was
4104 initializing a 32-byte map regardless, which could cause it to run off the end
4105 of the memory it had got.
4106
4107 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4108
4109
4110 Version 0.98 22-Oct-97
4111 ----------------------
4112
4113 1. Fixed bug in code for handling temporary memory usage when there are more
4114 back references than supplied space in the ovector. This could cause segfaults.
4115
4116
4117 Version 0.97 21-Oct-97
4118 ----------------------
4119
4120 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4121
4122 2. Optimized negated single characters not to use a bit map.
4123
4124 3. Brought error texts together as macro definitions; clarified some of them;
4125 fixed one that was wrong - it said "range out of order" when it meant "invalid
4126 escape sequence".
4127
4128 4. Changed some char * arguments to const char *.
4129
4130 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4131
4132 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4133 pcretest.
4134
4135
4136 Version 0.96 16-Oct-97
4137 ----------------------
4138
4139 1. Added a simple "pgrep" utility to the distribution.
4140
4141 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4142 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4143 where "ddd" means "one or more decimal digits".
4144
4145 3. Fixed serious bug. If a pattern had a back reference, but the call to
4146 pcre_exec() didn't supply a large enough ovector to record the related
4147 identifying subpattern, the match always failed. PCRE now remembers the number
4148 of the largest back reference, and gets some temporary memory in which to save
4149 the offsets during matching if necessary, in order to ensure that
4150 backreferences always work.
4151
4152 4. Increased the compatibility with Perl in a number of ways:
4153
4154 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4155 to request this handling. The option can be set at compile or exec time.
4156
4157 (b) $ matches before a terminating newline by default; an option
4158 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4159 mode). The option can be set at compile or exec time.
4160
4161 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4162 the same as Perl's. If the decimal number it represents is less than 10
4163 or there aren't that many previous left capturing parentheses, an octal
4164 escape is read. Inside a character class, it's always an octal escape,
4165 even if it is a single digit.
4166
4167 (d) An escaped but undefined alphabetic character is taken as a literal,
4168 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4169 escapes.
4170
4171 (e) {0} is now permitted. (The previous item is removed from the compiled
4172 pattern).
4173
4174 5. Changed all the names of code files so that the basic parts are no longer
4175 than 10 characters, and abolished the teeny "globals.c" file.
4176
4177 6. Changed the handling of character classes; they are now done with a 32-byte
4178 bit map always.
4179
4180 7. Added the -d and /D options to pcretest to make it possible to look at the
4181 internals of compilation without having to recompile pcre.
4182
4183
4184 Version 0.95 23-Sep-97
4185 ----------------------
4186
4187 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4188 \x20 at the start of a run of normal characters. These were being treated as
4189 real characters, instead of the source characters being re-checked.
4190
4191
4192 Version 0.94 18-Sep-97
4193 ----------------------
4194
4195 1. The functions are now thread-safe, with the caveat that the global variables
4196 containing pointers to malloc() and free() or alternative functions are the
4197 same for all threads.
4198
4199 2. Get pcre_study() to generate a bitmap of initial characters for non-
4200 anchored patterns when this is possible, and use it if passed to pcre_exec().
4201
4202
4203 Version 0.93 15-Sep-97
4204 ----------------------
4205
4206 1. /(b)|(:+)/ was computing an incorrect first character.
4207
4208 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4209 but not actually doing anything yet.
4210
4211 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4212 as Perl does (e.g. [-az] or [az-]).
4213
4214 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4215 all possible positions.
4216
4217 5. Split up into different modules to avoid including unneeded functions in a
4218 compiled binary. However, compile and exec are still in one module. The "study"
4219 function is split off.
4220
4221 6. The character tables are now in a separate module whose source is generated
4222 by an auxiliary program - but can then be edited by hand if required. There are
4223 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4224 toupper() in the code.
4225
4226 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4227 make them global. Abolish the function for setting them, as the caller can now
4228 set them directly.
4229
4230
4231 Version 0.92 11-Sep-97
4232 ----------------------
4233
4234 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4235 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4236
4237 2. Caseless matching was not working in character classes if the characters in
4238 the pattern were in upper case.
4239
4240 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4241
4242 4. Make PCRE_ANCHORED public and accept as a compile option.
4243
4244 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4245 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4246 pass them.
4247
4248 6. Give an error if bad option bits passed at compile or run time.
4249
4250 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4251 pcretest to cause it to pass that flag.
4252
4253 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4254 options, and the first character, if set.
4255
4256 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4257
4258
4259 Version 0.91 10-Sep-97
4260 ----------------------
4261
4262 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4263 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4264
4265 2. PCRE was looping on encountering an indefinitely repeated back reference to
4266 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4267 Perl does - treats the match as successful.
4268
4269 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12