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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 691 - (show annotations)
Sun Sep 11 14:31:21 2011 UTC (8 years, 1 month ago) by ph10
File size: 213658 byte(s)
Error occurred while calculating annotation data.
Final source and document tidies for 8.20-RC1.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.20 12-Sep-2011
5 ------------------------
6
7 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
8 a POSIX class. After further experiments with Perl, which convinced me that
9 Perl has bugs and confusions, a closing square bracket is no longer allowed
10 in a POSIX name. This bug also affected patterns with classes that started
11 with full stops.
12
13 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no captured
14 substring, but while checking the failing first alternative, substring 1 is
15 temporarily captured. If the output vector supplied to pcre_exec() was not
16 big enough for this capture, the yield of the function was still zero
17 ("insufficient space for captured substrings"). This cannot be totally fixed
18 without adding another stack variable, which seems a lot of expense for a
19 edge case. However, I have improved the situation in cases such as
20 /(a)(b)x|abc/ matched against "abc", where the return code indicates that
21 fewer than the maximum number of slots in the ovector have been set.
22
23 3. Related to (2) above: when there are more back references in a pattern than
24 slots in the output vector, pcre_exec() uses temporary memory during
25 matching, and copies in the captures as far as possible afterwards. It was
26 using the entire output vector, but this conflicts with the specification
27 that only 2/3 is used for passing back captured substrings. Now it uses only
28 the first 2/3, for compatibility. This is, of course, another edge case.
29
30 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
31 main code base, and can be used by building with --enable-jit. When this is
32 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
33 runtime --no-jit option is given.
34
35 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
36 ovector, the return from the function was zero, implying that there were
37 other matches that did not fit. The correct "exactly full" value is now
38 returned.
39
40 6. If a subpattern that was called recursively or as a subroutine contained
41 (*PRUNE) or any other control that caused it to give a non-standard return,
42 invalid errors such as "Error -26 (nested recursion at the same subject
43 position)" or even infinite loops could occur.
44
45
46 Version 8.13 16-Aug-2011
47 ------------------------
48
49 1. The Unicode data tables have been updated to Unicode 6.0.0.
50
51 2. Two minor typos in pcre_internal.h have been fixed.
52
53 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
54 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
55 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
56
57 4. There were a number of related bugs in the code for matching backrefences
58 caselessly in UTF-8 mode when codes for the characters concerned were
59 different numbers of bytes. For example, U+023A and U+2C65 are an upper
60 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
61 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
62 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
63 2-byte code at the end of the subject (it thought there wasn't enough data
64 left).
65
66 5. Comprehensive information about what went wrong is now returned by
67 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
68 as the output vector has at least 2 elements. The offset of the start of
69 the failing character and a reason code are placed in the vector.
70
71 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
72 now returned is for the first byte of the failing character, instead of the
73 last byte inspected. This is an incompatible change, but I hope it is small
74 enough not to be a problem. It makes the returned offset consistent with
75 pcre_exec() and pcre_dfa_exec().
76
77 7. pcretest now gives a text phrase as well as the error number when
78 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
79 failure, the offset and reason code are output.
80
81 8. When \R was used with a maximizing quantifier it failed to skip backwards
82 over a \r\n pair if the subsequent match failed. Instead, it just skipped
83 back over a single character (\n). This seems wrong (because it treated the
84 two characters as a single entity when going forwards), conflicts with the
85 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
86 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
87 has been changed.
88
89 9. Some internal refactoring has changed the processing so that the handling
90 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
91 time (the PCRE_DOTALL option was changed this way some time ago: version
92 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
93 which was always a bit of a fudge. It also means that there is one less
94 argument for the match() function, which reduces its stack requirements
95 slightly. This change also fixes an incompatibility with Perl: the pattern
96 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
97
98 10. More internal refactoring has drastically reduced the number of recursive
99 calls to match() for possessively repeated groups such as (abc)++ when
100 using pcre_exec().
101
102 11. While implementing 10, a number of bugs in the handling of groups were
103 discovered and fixed:
104
105 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
106 (a|)*(?1) gave a compile-time internal error.
107 ((a|)+)+ did not notice that the outer group could match an empty string.
108 (^a|^)+ was not marked as anchored.
109 (.*a|.*)+ was not marked as matching at start or after a newline.
110
111 12. Yet more internal refactoring has removed another argument from the match()
112 function. Special calls to this function are now indicated by setting a
113 value in a variable in the "match data" data block.
114
115 13. Be more explicit in pcre_study() instead of relying on "default" for
116 opcodes that mean there is no starting character; this means that when new
117 ones are added and accidentally left out of pcre_study(), testing should
118 pick them up.
119
120 14. The -s option of pcretest has been documented for ages as being an old
121 synonym of -m (show memory usage). I have changed it to mean "force study
122 for every regex", that is, assume /S for every regex. This is similar to -i
123 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
124 using it. It makes it easier to run collections of tests with and without
125 study enabled, and thereby test pcre_study() more easily. All the standard
126 tests are now run with and without -s (but some patterns can be marked as
127 "never study" - see 20 below).
128
129 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
130 restoration of the capturing data to the outer values was not happening
131 correctly.
132
133 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
134 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
135 pattern had matched an empty string, and so incorrectly returned a no
136 match.
137
138 17. There was optimizing code for the last branch of non-capturing parentheses,
139 and also for the obeyed branch of a conditional subexpression, which used
140 tail recursion to cut down on stack usage. Unfortunately, now that there is
141 the possibility of (*THEN) occurring in these branches, tail recursion is
142 no longer possible because the return has to be checked for (*THEN). These
143 two optimizations have therefore been removed.
144
145 18. If a pattern containing \R was studied, it was assumed that \R always
146 matched two bytes, thus causing the minimum subject length to be
147 incorrectly computed because \R can also match just one byte.
148
149 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
150 was incorrectly computed.
151
152 20. If /S is present twice on a test pattern in pcretest input, it now
153 *disables* studying, thereby overriding the use of -s on the command line
154 (see 14 above). This is necessary for one or two tests to keep the output
155 identical in both cases.
156
157 21. When (*ACCEPT) was used in an assertion that matched an empty string and
158 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
159
160 22. When an atomic group that contained a capturing parenthesis was
161 successfully matched, but the branch in which it appeared failed, the
162 capturing was not being forgotten if a higher numbered group was later
163 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
164 group 1 to "a", when in fact it should be unset. This applied to multi-
165 branched capturing and non-capturing groups, repeated or not, and also to
166 positive assertions (capturing in negative assertions does not happen
167 in PCRE) and also to nested atomic groups.
168
169 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
170 subject after a captured substring, to make it easier to tell which of a
171 number of identical substrings has been captured.
172
173 24. The way atomic groups are processed by pcre_exec() has been changed so that
174 if they are repeated, backtracking one repetition now resets captured
175 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
176 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
177 "aaa". Previously, it would have been "a". As part of this code
178 refactoring, the way recursive calls are handled has also been changed.
179
180 25. If an assertion condition captured any substrings, they were not passed
181 back unless some other capturing happened later. For example, if
182 (?(?=(a))a) was matched against "a", no capturing was returned.
183
184 26. When studying a pattern that contained subroutine calls or assertions,
185 the code for finding the minimum length of a possible match was handling
186 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
187 group 1 called group 2 while simultaneously a separate group 2 called group
188 1). A stack overflow occurred in this case. I have fixed this by limiting
189 the recursion depth to 10.
190
191 27. Updated RunTest.bat in the distribution to the version supplied by Tom
192 Fortmann. This supports explicit test numbers on the command line, and has
193 argument validation and error reporting.
194
195 28. An instance of \X with an unlimited repeat could fail if at any point the
196 first character it looked at was a mark character.
197
198 29. Some minor code refactoring concerning Unicode properties and scripts
199 should reduce the stack requirement of match() slightly.
200
201 30. Added the '=' option to pcretest to check the setting of unused capturing
202 slots at the end of the pattern, which are documented as being -1, but are
203 not included in the return count.
204
205 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
206 compiled something random. Now it gives a compile-time error (as does
207 Perl).
208
209 32. A *MARK encountered during the processing of a positive assertion is now
210 recorded and passed back (compatible with Perl).
211
212 33. If --only-matching or --colour was set on a pcregrep call whose pattern
213 had alternative anchored branches, the search for a second match in a line
214 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
215 matched the line "0102" twice. The same bug affected patterns that started
216 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
217 twice.
218
219 34. Previously, PCRE did not allow quantification of assertions. However, Perl
220 does, and because of capturing effects, quantifying parenthesized
221 assertions may at times be useful. Quantifiers are now allowed for
222 parenthesized assertions.
223
224 35. A minor code tidy in pcre_compile() when checking options for \R usage.
225
226 36. \g was being checked for fancy things in a character class, when it should
227 just be a literal "g".
228
229 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
230 appearance of a nested POSIX class supersedes an apparent external class.
231 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
232 unescaped square brackets may also appear as part of class names. For
233 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
234 more like Perl. (But see 8.20/1 above.)
235
236 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
237 was because it thought it was \N{name}, which is not supported).
238
239 39. Add minix to OS list not supporting the -S option in pcretest.
240
241 40. PCRE tries to detect cases of infinite recursion at compile time, but it
242 cannot analyze patterns in sufficient detail to catch mutual recursions
243 such as ((?1))((?2)). There is now a runtime test that gives an error if a
244 subgroup is called recursively as a subpattern for a second time at the
245 same position in the subject string. In previous releases this might have
246 been caught by the recursion limit, or it might have run out of stack.
247
248 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
249 happen only once. PCRE was, however incorrectly giving a compile time error
250 "recursive call could loop indefinitely" because it cannot analyze the
251 pattern in sufficient detail. The compile time test no longer happens when
252 PCRE is compiling a conditional subpattern, but actual runaway loops are
253 now caught at runtime (see 40 above).
254
255 42. It seems that Perl allows any characters other than a closing parenthesis
256 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
257 has been changed to be the same.
258
259 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
260 as not to get warnings when autogen.sh is called. Also changed
261 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
262
263 44. To help people who use pcregrep to scan files containing exceedingly long
264 lines, the following changes have been made:
265
266 (a) The default value of the buffer size parameter has been increased from
267 8K to 20K. (The actual buffer used is three times this size.)
268
269 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
270 PCRE is built.
271
272 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
273 to be set at run time.
274
275 (d) Numerical values in pcregrep options can be followed by K or M, for
276 example --buffer-size=50K.
277
278 (e) If a line being scanned overflows pcregrep's buffer, an error is now
279 given and the return code is set to 2.
280
281 45. Add a pointer to the latest mark to the callout data block.
282
283 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
284 partial match of an empty string instead of no match. This was specific to
285 the use of ".".
286
287 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
288 complete match instead of a partial match. This bug was dependent on both
289 the PCRE_UTF8 and PCRE_DOTALL options being set.
290
291 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
292 starting byte set, because \b was not being ignored.
293
294
295 Version 8.12 15-Jan-2011
296 ------------------------
297
298 1. Fixed some typos in the markup of the man pages, and wrote a script that
299 checks for such things as part of the documentation building process.
300
301 2. On a big-endian 64-bit system, pcregrep did not correctly process the
302 --match-limit and --recursion-limit options (added for 8.11). In
303 particular, this made one of the standard tests fail. (The integer value
304 went into the wrong half of a long int.)
305
306 3. If the --colour option was given to pcregrep with -v (invert match), it
307 did strange things, either producing crazy output, or crashing. It should,
308 of course, ignore a request for colour when reporting lines that do not
309 match.
310
311 4. Another pcregrep bug caused similar problems if --colour was specified with
312 -M (multiline) and the pattern match finished with a line ending.
313
314 5. In pcregrep, when a pattern that ended with a literal newline sequence was
315 matched in multiline mode, the following line was shown as part of the
316 match. This seems wrong, so I have changed it.
317
318 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
319 the check for further matches in the same line (so they could be coloured)
320 to overrun the end of the current line. If another match was found, it was
321 incorrectly shown (and then shown again when found in the next line).
322
323 7. If pcregrep was compiled under Windows, there was a reference to the
324 function pcregrep_exit() before it was defined. I am assuming this was
325 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
326 reported by a user. I've moved the definition above the reference.
327
328
329 Version 8.11 10-Dec-2010
330 ------------------------
331
332 1. (*THEN) was not working properly if there were untried alternatives prior
333 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
334 backtracked to try for "b" instead of moving to the next alternative branch
335 at the same level (in this case, to look for "c"). The Perl documentation
336 is clear that when (*THEN) is backtracked onto, it goes to the "next
337 alternative in the innermost enclosing group".
338
339 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
340 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
341 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
342 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
343 (*THEN).
344
345 3. If \s appeared in a character class, it removed the VT character from
346 the class, even if it had been included by some previous item, for example
347 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
348 of \s, but is part of the POSIX "space" class.)
349
350 4. A partial match never returns an empty string (because you can always
351 match an empty string at the end of the subject); however the checking for
352 an empty string was starting at the "start of match" point. This has been
353 changed to the "earliest inspected character" point, because the returned
354 data for a partial match starts at this character. This means that, for
355 example, /(?<=abc)def/ gives a partial match for the subject "abc"
356 (previously it gave "no match").
357
358 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
359 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
360 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
361 has an implication that the given string is incomplete (because a partial
362 match is preferred over a full match). For this reason, these items now
363 give a partial match in this situation. [Aside: previously, the one case
364 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
365 match rather than a full match, which was wrong by the old rules, but is
366 now correct.]
367
368 6. There was a bug in the handling of #-introduced comments, recognized when
369 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
370 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
371 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
372 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
373 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
374 places in pcre_compile().
375
376 7. Related to (6) above, when pcre_compile() was skipping #-introduced
377 comments when looking ahead for named forward references to subpatterns,
378 the only newline sequence it recognized was NL. It now handles newlines
379 according to the set newline convention.
380
381 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
382 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
383 cater for a lack of strerror(). These oversights have been fixed.
384
385 9. Added --match-limit and --recursion-limit to pcregrep.
386
387 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
388
389 11. When the -o option was used, pcregrep was setting a return code of 1, even
390 when matches were found, and --line-buffered was not being honoured.
391
392 12. Added an optional parentheses number to the -o and --only-matching options
393 of pcregrep.
394
395 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
396 can match an empty string. The code to do it in pcretest and pcredemo
397 needed fixing:
398
399 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
400 only one byte after an empty string match just before CRLF (this case
401 just got forgotten; "any" and "anycrlf" were OK).
402
403 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
404 mode when an empty string match preceded an ASCII character followed by
405 a non-ASCII character. (The code for advancing by one character rather
406 than one byte was nonsense.)
407
408 (c) The pcredemo.c sample program did not have any code at all to handle
409 the cases when CRLF is a valid newline sequence.
410
411 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
412 as a starting offset was within the subject string. There is now a new
413 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
414 negative or greater than the length of the string. In order to test this,
415 pcretest is extended to allow the setting of negative starting offsets.
416
417 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
418 starting offset points to the beginning of a UTF-8 character was
419 unnecessarily clumsy. I tidied it up.
420
421 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
422 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
423
424 17. Nobody had reported that the --include_dir option, which was added in
425 release 7.7 should have been called --include-dir (hyphen, not underscore)
426 for compatibility with GNU grep. I have changed it to --include-dir, but
427 left --include_dir as an undocumented synonym, and the same for
428 --exclude-dir, though that is not available in GNU grep, at least as of
429 release 2.5.4.
430
431 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
432 characters from a string of bytes) have been redefined so as not to use
433 loops, in order to improve performance in some environments. At the same
434 time, I abstracted some of the common code into auxiliary macros to save
435 repetition (this should not affect the compiled code).
436
437 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
438 compile-time error is now given if \c is not followed by an ASCII
439 character, that is, a byte less than 128. (In EBCDIC mode, the code is
440 different, and any byte value is allowed.)
441
442 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
443 START_OPTIMIZE option, which is now allowed at compile time - but just
444 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
445 to pcregrep and other applications that have no direct access to PCRE
446 options. The new /Y option in pcretest sets this option when calling
447 pcre_compile().
448
449 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
450 back references. Groups containing recursive back references were forced to
451 be atomic by that change, but in the case of named groups, the amount of
452 memory required was incorrectly computed, leading to "Failed: internal
453 error: code overflow". This has been fixed.
454
455 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
456 pcretest.c, to avoid build problems in some Borland environments.
457
458
459 Version 8.10 25-Jun-2010
460 ------------------------
461
462 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
463 THEN.
464
465 2. (*ACCEPT) was not working when inside an atomic group.
466
467 3. Inside a character class, \B is treated as a literal by default, but
468 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
469 causes the error). The code is unchanged, but I tidied the documentation.
470
471 4. Inside a character class, PCRE always treated \R and \X as literals,
472 whereas Perl faults them if its -w option is set. I have changed PCRE so
473 that it faults them when PCRE_EXTRA is set.
474
475 5. Added support for \N, which always matches any character other than
476 newline. (It is the same as "." when PCRE_DOTALL is not set.)
477
478 6. When compiling pcregrep with newer versions of gcc which may have
479 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
480 declared with attribute warn_unused_result" were given. Just casting the
481 result to (void) does not stop the warnings; a more elaborate fudge is
482 needed. I've used a macro to implement this.
483
484 7. Minor change to pcretest.c to avoid a compiler warning.
485
486 8. Added four artifical Unicode properties to help with an option to make
487 \s etc use properties (see next item). The new properties are: Xan
488 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
489
490 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
491 use Unicode properties. (*UCP) at the start of a pattern can be used to set
492 this option. Modified pcretest to add /W to test this facility. Added
493 REG_UCP to make it available via the POSIX interface.
494
495 10. Added --line-buffered to pcregrep.
496
497 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
498 studied, and the match started with a letter with a code point greater than
499 127 whose first byte was different to the first byte of the other case of
500 the letter, the other case of this starting letter was not recognized
501 (#976).
502
503 12. If a pattern that was studied started with a repeated Unicode property
504 test, for example, \p{Nd}+, there was the theoretical possibility of
505 setting up an incorrect bitmap of starting bytes, but fortunately it could
506 not have actually happened in practice until change 8 above was made (it
507 added property types that matched character-matching opcodes).
508
509 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
510 possible starting bytes for non-anchored patterns.
511
512 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
513 \R, and also a number of cases that involve Unicode properties, both
514 explicit and implicit when PCRE_UCP is set.
515
516 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
517 input, it could crash or give wrong results if characters with values
518 greater than 0xc0 were present in the subject string. (Detail: it assumed
519 UTF-8 input when processing these items.)
520
521 16. Added a lot of (int) casts to avoid compiler warnings in systems where
522 size_t is 64-bit (#991).
523
524 17. Added a check for running out of memory when PCRE is compiled with
525 --disable-stack-for-recursion (#990).
526
527 18. If the last data line in a file for pcretest does not have a newline on
528 the end, a newline was missing in the output.
529
530 19. The default pcre_chartables.c file recognizes only ASCII characters (values
531 less than 128) in its various bitmaps. However, there is a facility for
532 generating tables according to the current locale when PCRE is compiled. It
533 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
534 characters, are recognized by isspace() and therefore were getting set in
535 these tables, and indeed these tables seem to approximate to ISO 8859. This
536 caused a problem in UTF-8 mode when pcre_study() was used to create a list
537 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
538 which of course cannot start UTF-8 characters. I have changed the code so
539 that only real ASCII characters (less than 128) and the correct starting
540 bytes for UTF-8 encodings are set for characters greater than 127 when in
541 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
542 altogether.)
543
544 20. Added the /T option to pcretest so as to be able to run tests with non-
545 standard character tables, thus making it possible to include the tests
546 used for 19 above in the standard set of tests.
547
548 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
549 reference to a subpattern the other side of a comment that contains an
550 opening parenthesis caused either an internal compiling error, or a
551 reference to the wrong subpattern.
552
553
554 Version 8.02 19-Mar-2010
555 ------------------------
556
557 1. The Unicode data tables have been updated to Unicode 5.2.0.
558
559 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
560 configured.
561
562 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
563 original author of that file, following a query about its status.
564
565 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
566 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
567
568 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
569 quantifier applied to a forward-referencing subroutine call, could compile
570 incorrect code or give the error "internal error: previously-checked
571 referenced subpattern not found".
572
573 6. Both MS Visual Studio and Symbian OS have problems with initializing
574 variables to point to external functions. For these systems, therefore,
575 pcre_malloc etc. are now initialized to local functions that call the
576 relevant global functions.
577
578 7. There were two entries missing in the vectors called coptable and poptable
579 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
580 I've fixed the data, and added a kludgy way of testing at compile time that
581 the lengths are correct (equal to the number of opcodes).
582
583 8. Following on from 7, I added a similar kludge to check the length of the
584 eint vector in pcreposix.c.
585
586 9. Error texts for pcre_compile() are held as one long string to avoid too
587 much relocation at load time. To find a text, the string is searched,
588 counting zeros. There was no check for running off the end of the string,
589 which could happen if a new error number was added without updating the
590 string.
591
592 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
593
594 11. \K was not working if it appeared in an atomic group or in a group that
595 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
596 \K is "not well defined" if used in an assertion. PCRE now accepts it if
597 the assertion is positive, but not if it is negative.
598
599 12. Change 11 fortuitously reduced the size of the stack frame used in the
600 "match()" function of pcre_exec.c by one pointer. Forthcoming
601 implementation of support for (*MARK) will need an extra pointer on the
602 stack; I have reserved it now, so that the stack frame size does not
603 decrease.
604
605 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
606 item in branch that calls a recursion is a subroutine call - as in the
607 second branch in the above example - was incorrectly given the compile-
608 time error "recursive call could loop indefinitely" because pcre_compile()
609 was not correctly checking the subroutine for matching a non-empty string.
610
611 14. The checks for overrunning compiling workspace could trigger after an
612 overrun had occurred. This is a "should never occur" error, but it can be
613 triggered by pathological patterns such as hundreds of nested parentheses.
614 The checks now trigger 100 bytes before the end of the workspace.
615
616 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
617
618
619 Version 8.01 19-Jan-2010
620 ------------------------
621
622 1. If a pattern contained a conditional subpattern with only one branch (in
623 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
624 computed the wrong minimum data length (which is of course zero for such
625 subpatterns). This could cause incorrect "no match" results.
626
627 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
628 the pattern is reset in the first branch, pcre_compile() failed with
629 "internal error: code overflow at offset...". This happened only when
630 the reset was to the original external option setting. (An optimization
631 abstracts leading options settings into an external setting, which was the
632 cause of this.)
633
634 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
635 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
636 assertion pattern did not match (meaning that the assertion was true), it
637 was incorrectly treated as false if the SKIP had been reached during the
638 matching. This also applied to assertions used as conditions.
639
640 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
641 assertion subpattern, including such a pattern used as a condition,
642 unpredictable results occurred, instead of the error return
643 PCRE_ERROR_DFA_UITEM.
644
645 5. The C++ GlobalReplace function was not working like Perl for the special
646 situation when an empty string is matched. It now does the fancy magic
647 stuff that is necessary.
648
649 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
650 removed. (These were left over from very, very early versions of PCRE.)
651
652 7. Some cosmetic changes to the code to make life easier when compiling it
653 as part of something else:
654
655 (a) Change DEBUG to PCRE_DEBUG.
656
657 (b) In pcre_compile(), rename the member of the "branch_chain" structure
658 called "current" as "current_branch", to prevent a collision with the
659 Linux macro when compiled as a kernel module.
660
661 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
662 prevent a collision with the Linux macro when compiled as a kernel
663 module.
664
665 8. In pcre_compile() there are some checks for integer overflows that used to
666 cast potentially large values to (double). This has been changed to that
667 when building, a check for int64_t is made, and if it is found, it is used
668 instead, thus avoiding the use of floating point arithmetic. (There is no
669 other use of FP in PCRE.) If int64_t is not found, the fallback is to
670 double.
671
672 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
673 2005 (difference between two addresses compared to an unsigned value).
674
675 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
676 custom one, because of the following reported problem in Windows:
677
678 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
679 under Win32.
680 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
681 therefore missing the function definition.
682 - The compiler thus generates a "C" signature for the test function.
683 - The linker fails to find the "C" function.
684 - PCRE fails to configure if asked to do so against libbz2.
685
686 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
687 messages were output:
688
689 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
690 rerunning libtoolize, to keep the correct libtool macros in-tree.
691 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
692
693 I have done both of these things.
694
695 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
696 most of the time, it *can* run out if it is given a pattern that contains a
697 runaway infinite recursion. I updated the discussion in the pcrestack man
698 page.
699
700 13. Now that we have gone to the x.xx style of version numbers, the minor
701 version may start with zero. Using 08 or 09 is a bad idea because users
702 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
703 interpreted as invalid octal numbers. I've updated the previous comment in
704 configure.ac, and also added a check that gives an error if 08 or 09 are
705 used.
706
707 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
708 causing partial matching to fail when the end of the subject matched \W
709 in a UTF-8 pattern where \W was quantified with a minimum of 3.
710
711 15. There were some discrepancies between the declarations in pcre_internal.h
712 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
713 their definitions. The declarations used "const uschar *" and the
714 definitions used USPTR. Even though USPTR is normally defined as "const
715 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
716 reported that: "This difference in casting confuses some C++ compilers, for
717 example, SunCC recognizes above declarations as different functions and
718 generates broken code for hbpcre." I have changed the declarations to use
719 USPTR.
720
721 16. GNU libtool is named differently on some systems. The autogen.sh script now
722 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
723 (FreeBSD).
724
725 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
726 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
727 comment: "Figure out how to create a longlong from a string: strtoll and
728 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
729 instance, but it only takes 2 args instead of 3!"
730
731 18. A subtle bug concerned with back references has been fixed by a change of
732 specification, with a corresponding code fix. A pattern such as
733 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
734 refers, was giving matches when it shouldn't. For example, xa=xaaa would
735 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
736 same bug. Such groups have to be quantified to be useful, or contained
737 inside another quantified group. (If there's no repetition, the reference
738 can never match.) The problem arises because, having left the group and
739 moved on to the rest of the pattern, a later failure that backtracks into
740 the group uses the captured value from the final iteration of the group
741 rather than the correct earlier one. I have fixed this in PCRE by forcing
742 any group that contains a reference to itself to be an atomic group; that
743 is, there cannot be any backtracking into it once it has completed. This is
744 similar to recursive and subroutine calls.
745
746
747 Version 8.00 19-Oct-09
748 ----------------------
749
750 1. The table for translating pcre_compile() error codes into POSIX error codes
751 was out-of-date, and there was no check on the pcre_compile() error code
752 being within the table. This could lead to an OK return being given in
753 error.
754
755 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
756 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
757 in a Windows environment.
758
759 3. The pcregrep --count option prints the count for each file even when it is
760 zero, as does GNU grep. However, pcregrep was also printing all files when
761 --files-with-matches was added. Now, when both options are given, it prints
762 counts only for those files that have at least one match. (GNU grep just
763 prints the file name in this circumstance, but including the count seems
764 more useful - otherwise, why use --count?) Also ensured that the
765 combination -clh just lists non-zero counts, with no names.
766
767 4. The long form of the pcregrep -F option was incorrectly implemented as
768 --fixed_strings instead of --fixed-strings. This is an incompatible change,
769 but it seems right to fix it, and I didn't think it was worth preserving
770 the old behaviour.
771
772 5. The command line items --regex=pattern and --regexp=pattern were not
773 recognized by pcregrep, which required --regex pattern or --regexp pattern
774 (with a space rather than an '='). The man page documented the '=' forms,
775 which are compatible with GNU grep; these now work.
776
777 6. No libpcreposix.pc file was created for pkg-config; there was just
778 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
779
780 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
781 when UCP support is not needed, by modifying the Python script that
782 generates it from Unicode data files. This should not matter if the module
783 is correctly used as a library, but I received one complaint about 50K of
784 unwanted data. My guess is that the person linked everything into his
785 program rather than using a library. Anyway, it does no harm.
786
787 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
788 was a minimum greater than 1 for a wide character in a possessive
789 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
790 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
791 character. Chaos in the form of incorrect output or a compiling loop could
792 result.
793
794 9. The restrictions on what a pattern can contain when partial matching is
795 requested for pcre_exec() have been removed. All patterns can now be
796 partially matched by this function. In addition, if there are at least two
797 slots in the offset vector, the offset of the earliest inspected character
798 for the match and the offset of the end of the subject are set in them when
799 PCRE_ERROR_PARTIAL is returned.
800
801 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
802 synonymous with PCRE_PARTIAL, for backwards compatibility, and
803 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
804 and may be more useful for multi-segment matching.
805
806 11. Partial matching with pcre_exec() is now more intuitive. A partial match
807 used to be given if ever the end of the subject was reached; now it is
808 given only if matching could not proceed because another character was
809 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
810 string "Z", which now yields "no match" instead of "partial match". In the
811 case of pcre_dfa_exec(), "no match" is given if every matching path for the
812 final character ended with (*FAIL).
813
814 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
815 if the pattern had a "must contain" character that was already found in the
816 earlier partial match, unless partial matching was again requested. For
817 example, with the pattern /dog.(body)?/, the "must contain" character is
818 "g". If the first part-match was for the string "dog", restarting with
819 "sbody" failed. This bug has been fixed.
820
821 13. The string returned by pcre_dfa_exec() after a partial match has been
822 changed so that it starts at the first inspected character rather than the
823 first character of the match. This makes a difference only if the pattern
824 starts with a lookbehind assertion or \b or \B (\K is not supported by
825 pcre_dfa_exec()). It's an incompatible change, but it makes the two
826 matching functions compatible, and I think it's the right thing to do.
827
828 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
829 so that the demonstration program is easily available in environments where
830 PCRE has not been installed from source.
831
832 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
833 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
834 library.
835
836 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
837 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
838 is not the first non-POSIX option to be added. Clearly some people find
839 these options useful.
840
841 17. If a caller to the POSIX matching function regexec() passes a non-zero
842 value for nmatch with a NULL value for pmatch, the value of
843 nmatch is forced to zero.
844
845 18. RunGrepTest did not have a test for the availability of the -u option of
846 the diff command, as RunTest does. It now checks in the same way as
847 RunTest, and also checks for the -b option.
848
849 19. If an odd number of negated classes containing just a single character
850 interposed, within parentheses, between a forward reference to a named
851 subpattern and the definition of the subpattern, compilation crashed with
852 an internal error, complaining that it could not find the referenced
853 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
854 [The bug was that it was starting one character too far in when skipping
855 over the character class, thus treating the ] as data rather than
856 terminating the class. This meant it could skip too much.]
857
858 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
859 /g option in pcretest when the pattern contains \K, which makes it possible
860 to have an empty string match not at the start, even when the pattern is
861 anchored. Updated pcretest and pcredemo to use this option.
862
863 21. If the maximum number of capturing subpatterns in a recursion was greater
864 than the maximum at the outer level, the higher number was returned, but
865 with unset values at the outer level. The correct (outer level) value is
866 now given.
867
868 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
869 PCRE did not set those parentheses (unlike Perl). I have now found a way to
870 make it do so. The string so far is captured, making this feature
871 compatible with Perl.
872
873 23. The tests have been re-organized, adding tests 11 and 12, to make it
874 possible to check the Perl 5.10 features against Perl 5.10.
875
876 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
877 pattern matches a fixed length string. PCRE did not allow this; now it
878 does. Neither allows recursion.
879
880 25. I finally figured out how to implement a request to provide the minimum
881 length of subject string that was needed in order to match a given pattern.
882 (It was back references and recursion that I had previously got hung up
883 on.) This code has now been added to pcre_study(); it finds a lower bound
884 to the length of subject needed. It is not necessarily the greatest lower
885 bound, but using it to avoid searching strings that are too short does give
886 some useful speed-ups. The value is available to calling programs via
887 pcre_fullinfo().
888
889 26. While implementing 25, I discovered to my embarrassment that pcretest had
890 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
891 study optimizations had never been tested with that matching function.
892 Oops. What is worse, even when it was passed study data, there was a bug in
893 pcre_dfa_exec() that meant it never actually used it. Double oops. There
894 were also very few tests of studied patterns with pcre_dfa_exec().
895
896 27. If (?| is used to create subpatterns with duplicate numbers, they are now
897 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
898 on the other side of the coin, they are no longer allowed to have different
899 names, because these cannot be distinguished in PCRE, and this has caused
900 confusion. (This is a difference from Perl.)
901
902 28. When duplicate subpattern names are present (necessarily with different
903 numbers, as required by 27 above), and a test is made by name in a
904 conditional pattern, either for a subpattern having been matched, or for
905 recursion in such a pattern, all the associated numbered subpatterns are
906 tested, and the overall condition is true if the condition is true for any
907 one of them. This is the way Perl works, and is also more like the way
908 testing by number works.
909
910
911 Version 7.9 11-Apr-09
912 ---------------------
913
914 1. When building with support for bzlib/zlib (pcregrep) and/or readline
915 (pcretest), all targets were linked against these libraries. This included
916 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
917 libraries. This caused unwanted dependencies to be created. This problem
918 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
919 pcretest is linked with readline.
920
921 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
922 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
923 moved outside it again, because FALSE and TRUE are already defined in AIX,
924 but BOOL is not.
925
926 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
927 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
928
929 4. The pcregrep documentation said spaces were inserted as well as colons (or
930 hyphens) following file names and line numbers when outputting matching
931 lines. This is not true; no spaces are inserted. I have also clarified the
932 wording for the --colour (or --color) option.
933
934 5. In pcregrep, when --colour was used with -o, the list of matching strings
935 was not coloured; this is different to GNU grep, so I have changed it to be
936 the same.
937
938 6. When --colo(u)r was used in pcregrep, only the first matching substring in
939 each matching line was coloured. Now it goes on to look for further matches
940 of any of the test patterns, which is the same behaviour as GNU grep.
941
942 7. A pattern that could match an empty string could cause pcregrep to loop; it
943 doesn't make sense to accept an empty string match in pcregrep, so I have
944 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
945 seems to be how GNU grep behaves.
946
947 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
948 start or after a newline", because the conditional assertion was not being
949 correctly handled. The rule now is that both the assertion and what follows
950 in the first alternative must satisfy the test.
951
952 9. If auto-callout was enabled in a pattern with a conditional group whose
953 condition was an assertion, PCRE could crash during matching, both with
954 pcre_exec() and pcre_dfa_exec().
955
956 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
957 used for matching.
958
959 11. Unicode property support in character classes was not working for
960 characters (bytes) greater than 127 when not in UTF-8 mode.
961
962 12. Added the -M command line option to pcretest.
963
964 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
965
966 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
967
968 16. Added comments and documentation about mis-use of no_arg in the C++
969 wrapper.
970
971 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
972 from Martin Jerabek that uses macro names for all relevant character and
973 string constants.
974
975 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
976 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
977 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
978 these, but not everybody uses configure.
979
980 19. A conditional group that had only one branch was not being correctly
981 recognized as an item that could match an empty string. This meant that an
982 enclosing group might also not be so recognized, causing infinite looping
983 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
984 with the subject "ab", where knowledge that the repeated group can match
985 nothing is needed in order to break the loop.
986
987 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
988 exec(), but without supplying a callout function, matching went wrong.
989
990 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
991 leak if the size of the offset vector was greater than 30. When the vector
992 is smaller, the saved offsets during recursion go onto a local stack
993 vector, but for larger vectors malloc() is used. It was failing to free
994 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
995 error, in fact).
996
997 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
998 heapframe that is used only when UTF-8 support is enabled. This caused no
999 problem, but was untidy.
1000
1001 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1002 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1003 included within another project.
1004
1005 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1006 slightly modified by me:
1007
1008 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1009 not building pcregrep.
1010
1011 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1012 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1013
1014 25. Forward references, both numeric and by name, in patterns that made use of
1015 duplicate group numbers, could behave incorrectly or give incorrect errors,
1016 because when scanning forward to find the reference group, PCRE was not
1017 taking into account the duplicate group numbers. A pattern such as
1018 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1019
1020 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1021 the feature of a custom pointer more persuasive (as requested by a user).
1022
1023 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1024 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1025 pre-defined.
1026
1027 28. Added support for (*UTF8) at the start of a pattern.
1028
1029 29. Arrange for flags added by the "release type" setting in CMake to be shown
1030 in the configuration summary.
1031
1032
1033 Version 7.8 05-Sep-08
1034 ---------------------
1035
1036 1. Replaced UCP searching code with optimized version as implemented for Ad
1037 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1038 stage table and inline lookup instead of a function, giving speed ups of 2
1039 to 5 times on some simple patterns that I tested. Permission was given to
1040 distribute the MultiStage2.py script that generates the tables (it's not in
1041 the tarball, but is in the Subversion repository).
1042
1043 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1044 scripts.
1045
1046 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1047 a group with a zero qualifier. The result of the study could be incorrect,
1048 or the function might crash, depending on the pattern.
1049
1050 4. Caseless matching was not working for non-ASCII characters in back
1051 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1052 It now works when Unicode Property Support is available.
1053
1054 5. In pcretest, an escape such as \x{de} in the data was always generating
1055 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1056 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1057 truncation.
1058
1059 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1060
1061 7. Added two (int) casts to pcregrep when printing the difference of two
1062 pointers, in case they are 64-bit values.
1063
1064 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1065 test 2 if it fails.
1066
1067 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1068 and a #define of that name to empty if it is not externally set. This is to
1069 allow users of MSVC to set it if necessary.
1070
1071 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1072 the convenience functions in the pcre_get.c source file.
1073
1074 11. An option change at the start of a pattern that had top-level alternatives
1075 could cause overwriting and/or a crash. This command provoked a crash in
1076 some environments:
1077
1078 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1079
1080 This potential security problem was recorded as CVE-2008-2371.
1081
1082 12. For a pattern where the match had to start at the beginning or immediately
1083 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1084 pcre_dfa_exec() could read past the end of the passed subject if there was
1085 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1086 pcretest so that it places the subject at the end of its malloc-ed buffer.
1087
1088 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1089 exec() might read past the end of the data buffer in UTF-8 mode.
1090
1091 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1092 the data contained the byte 0x85 as part of a UTF-8 character within its
1093 first line. This applied both to normal and DFA matching.
1094
1095 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1096 /^[^d]*?$/8 failed to match "abc".
1097
1098 16. Added a missing copyright notice to pcrecpp_internal.h.
1099
1100 17. Make it more clear in the documentation that values returned from
1101 pcre_exec() in ovector are byte offsets, not character counts.
1102
1103 18. Tidied a few places to stop certain compilers from issuing warnings.
1104
1105 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1106 supplied by Stefan Weber. I made a further small update for 7.8 because
1107 there is a change of source arrangements: the pcre_searchfuncs.c module is
1108 replaced by pcre_ucd.c.
1109
1110
1111 Version 7.7 07-May-08
1112 ---------------------
1113
1114 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1115 a string to a long long, pretend we don't even have a long long." This is
1116 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1117
1118 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1119 pre-7.6 versions, which defined a global no_arg variable instead of putting
1120 it in the RE class. (See also #8 below.)
1121
1122 3. Remove a line of dead code, identified by coverity and reported by Nuno
1123 Lopes.
1124
1125 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1126
1127 (1) The include/exclude patterns were being applied to the whole pathnames
1128 of files, instead of just to the final components.
1129
1130 (2) If there was more than one level of directory, the subdirectories were
1131 skipped unless they satisfied the include/exclude conditions. This is
1132 inconsistent with GNU grep (and could even be seen as contrary to the
1133 pcregrep specification - which I improved to make it absolutely clear).
1134 The action now is always to scan all levels of directory, and just
1135 apply the include/exclude patterns to regular files.
1136
1137 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1138 --exclude_dir in the tests to avoid scanning .svn directories.
1139
1140 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1141 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1142 doesn't support NULs in patterns.
1143
1144 7. Added some missing "const"s to declarations of static tables in
1145 pcre_compile.c and pcre_dfa_exec.c.
1146
1147 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1148 caused by fix #2 above. (Subsequently also a second patch to fix the
1149 first patch. And a third patch - this was a messy problem.)
1150
1151 9. Applied Craig's patch to remove the use of push_back().
1152
1153 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1154 matching function regexec().
1155
1156 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1157 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1158 references. PCRE supports relative numbers with this syntax (I don't think
1159 Oniguruma does).
1160
1161 12. Previously, a group with a zero repeat such as (...){0} was completely
1162 omitted from the compiled regex. However, this means that if the group
1163 was called as a subroutine from elsewhere in the pattern, things went wrong
1164 (an internal error was given). Such groups are now left in the compiled
1165 pattern, with a new opcode that causes them to be skipped at execution
1166 time.
1167
1168 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1169 to the way PCRE behaves:
1170
1171 (a) A lone ] character is dis-allowed (Perl treats it as data).
1172
1173 (b) A back reference to an unmatched subpattern matches an empty string
1174 (Perl fails the current match path).
1175
1176 (c) A data ] in a character class must be notated as \] because if the
1177 first data character in a class is ], it defines an empty class. (In
1178 Perl it is not possible to have an empty class.) The empty class []
1179 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1180 The negative empty class [^] matches any one character, independently
1181 of the DOTALL setting.
1182
1183 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1184 non-existent subpattern following a character class starting with ']' and
1185 containing () gave an internal compiling error instead of "reference to
1186 non-existent subpattern". Fortunately, when the pattern did exist, the
1187 compiled code was correct. (When scanning forwards to check for the
1188 existencd of the subpattern, it was treating the data ']' as terminating
1189 the class, so got the count wrong. When actually compiling, the reference
1190 was subsequently set up correctly.)
1191
1192 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1193 it was being rejected as not supported by pcre_dfa_exec(), even though
1194 other assertions are supported. I have made pcre_dfa_exec() support
1195 (*FAIL).
1196
1197 16. The implementation of 13c above involved the invention of a new opcode,
1198 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1199 cannot be changed at match time, I realized I could make a small
1200 improvement to matching performance by compiling OP_ALLANY instead of
1201 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1202 on the OP_ANY path.
1203
1204 17. Compiling pcretest on Windows with readline support failed without the
1205 following two fixes: (1) Make the unistd.h include conditional on
1206 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1207
1208 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1209 ncurses library to be included for pcretest when ReadLine support is
1210 requested, but also to allow for it to be overridden. This patch came from
1211 Daniel Bergström.
1212
1213 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1214 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1215 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1216 spotting this.
1217
1218
1219 Version 7.6 28-Jan-08
1220 ---------------------
1221
1222 1. A character class containing a very large number of characters with
1223 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1224 overflow.
1225
1226 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1227 HAVE_LONG_LONG is not defined.
1228
1229 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1230 bring them up to date and include new features. This patch includes:
1231
1232 - Fixed PH's badly added libz and libbz2 support.
1233 - Fixed a problem with static linking.
1234 - Added pcredemo. [But later removed - see 7 below.]
1235 - Fixed dftables problem and added an option.
1236 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1237 HAVE_LONG_LONG.
1238 - Added readline support for pcretest.
1239 - Added an listing of the option settings after cmake has run.
1240
1241 4. A user submitted a patch to Makefile that makes it easy to create
1242 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1243 Makefile.am that cause it to include this special target, without
1244 affecting anything else. Note that the same mingw target plus all
1245 the other distribution libraries and programs are now supported
1246 when configuring with CMake (see 6 below) instead of with
1247 Configure/Make.
1248
1249 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1250 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1251 exported in the Windows port". It has not yet been confirmed that the patch
1252 solves the problem, but it does no harm.
1253
1254 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1255 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1256 with CMake, and also correct the comment about stack recursion.
1257
1258 7. Remove the automatic building of pcredemo from the ./configure system and
1259 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1260 of a program that users should build themselves after PCRE is installed, so
1261 building it automatically is not really right. What is more, it gave
1262 trouble in some build environments.
1263
1264 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1265
1266
1267 Version 7.5 10-Jan-08
1268 ---------------------
1269
1270 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1271 values in parens when parsing an RE using the C++ wrapper."
1272
1273 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1274 Characters greater than 255 were excluded from the class instead of being
1275 included.
1276
1277 3. The same bug as (2) above applied to negated POSIX classes such as
1278 [:^space:].
1279
1280 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1281 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1282 I have changed it.
1283
1284 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1285 first named subpattern) and a construct such as (?&a) would reference the
1286 first named subpattern whose name started with "a" (in other words, the
1287 length check was missing). Both these problems are fixed. "Subpattern name
1288 expected" is now given for (?&) (a zero-length name), and this patch also
1289 makes it give the same error for \k'' (previously it complained that that
1290 was a reference to a non-existent subpattern).
1291
1292 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1293 this is right because (?- can be followed by option settings as well as by
1294 digits. I have, however, made the messages clearer.
1295
1296 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1297 than the number used in the conditional) now cause a compile-time error.
1298 This is actually not compatible with Perl, which accepts such patterns, but
1299 treats the conditional as always being FALSE (as PCRE used to), but it
1300 seems to me that giving a diagnostic is better.
1301
1302 8. Change "alphameric" to the more common word "alphanumeric" in comments
1303 and messages.
1304
1305 9. Fix two occurrences of "backslash" in comments that should have been
1306 "backspace".
1307
1308 10. Remove two redundant lines of code that can never be obeyed (their function
1309 was moved elsewhere).
1310
1311 11. The program that makes PCRE's Unicode character property table had a bug
1312 which caused it to generate incorrect table entries for sequences of
1313 characters that have the same character type, but are in different scripts.
1314 It amalgamated them into a single range, with the script of the first of
1315 them. In other words, some characters were in the wrong script. There were
1316 thirteen such cases, affecting characters in the following ranges:
1317
1318 U+002b0 - U+002c1
1319 U+0060c - U+0060d
1320 U+0061e - U+00612
1321 U+0064b - U+0065e
1322 U+0074d - U+0076d
1323 U+01800 - U+01805
1324 U+01d00 - U+01d77
1325 U+01d9b - U+01dbf
1326 U+0200b - U+0200f
1327 U+030fc - U+030fe
1328 U+03260 - U+0327f
1329 U+0fb46 - U+0fbb1
1330 U+10450 - U+1049d
1331
1332 12. The -o option (show only the matching part of a line) for pcregrep was not
1333 compatible with GNU grep in that, if there was more than one match in a
1334 line, it showed only the first of them. It now behaves in the same way as
1335 GNU grep.
1336
1337 13. If the -o and -v options were combined for pcregrep, it printed a blank
1338 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1339 does the same. The return code can be used to tell if there were any
1340 non-matching lines.
1341
1342 14. Added --file-offsets and --line-offsets to pcregrep.
1343
1344 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1345 infinitely looping recursion. The bug was that positive lookaheads were not
1346 being skipped when checking for a possible empty match (negative lookaheads
1347 and both kinds of lookbehind were skipped).
1348
1349 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1350 inclusion of <windows.h> to before rather than after the definition of
1351 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1352
1353 17. Specifying a possessive quantifier with a specific limit for a Unicode
1354 character property caused pcre_compile() to compile bad code, which led at
1355 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1356 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1357 caused the error; without that there was no problem.
1358
1359 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1360
1361 19. Added --enable-pcretest-libreadline.
1362
1363 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1364 RE::GlobalReplace(). As a result, the number of replacements returned was
1365 double what it should be. I removed one of the increments, but Craig sent a
1366 later patch that removed the other one (the right fix) and added unit tests
1367 that check the return values (which was not done before).
1368
1369 21. Several CMake things:
1370
1371 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1372 the names libpcre and libpcreposix, not just pcre and pcreposix.
1373
1374 (2) The above change means that pcretest and pcregrep are now correctly
1375 linked with the newly-built libraries, not previously installed ones.
1376
1377 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1378
1379 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1380 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1381 UTF-8 newline character). The key issue is that the pattern starts .*;
1382 this means that the match must be either at the beginning, or after a
1383 newline. The bug was in the code for advancing after a failed match and
1384 checking that the new position followed a newline. It was not taking
1385 account of UTF-8 characters correctly.
1386
1387 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1388 character classes. PCRE was not treating the sequence [:...:] as a
1389 character class unless the ... were all letters. Perl, however, seems to
1390 allow any characters between [: and :], though of course it rejects as
1391 unknown any "names" that contain non-letters, because all the known class
1392 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1393 for example, whereas PCRE did not - it did not recognize a POSIX character
1394 class. This seemed a bit dangerous, so the code has been changed to be
1395 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1396 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1397 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1398 Perl does, and where it didn't before.
1399
1400 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1401 Windows environments %n is disabled by default.
1402
1403
1404 Version 7.4 21-Sep-07
1405 ---------------------
1406
1407 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1408 means that a class such as [\s] counted as "explicit reference to CR or
1409 LF". That isn't really right - the whole point of the change was to try to
1410 help when there was an actual mention of one of the two characters. So now
1411 the change happens only if \r or \n (or a literal CR or LF) character is
1412 encountered.
1413
1414 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1415 of both had grown to the point where there were only 3 bits left.
1416 Fortunately, there was spare space in the data structure, and so I have
1417 moved the internal flags into a new 16-bit field to free up more option
1418 bits.
1419
1420 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1421 but did not set the internal JCHANGED flag - either of these is enough to
1422 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1423 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1424 start sets both bits.
1425
1426 4. Added options (at build time, compile time, exec time) to change \R from
1427 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1428
1429 5. doc/pcresyntax.html was missing from the distribution.
1430
1431 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1432 compatibility, even though it is no longer used.
1433
1434 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1435 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1436 windows.h file is present (where different names are used). [This was
1437 reversed later after testing - see 16 below.]
1438
1439 8. Changed all #include <config.h> to #include "config.h". There were also
1440 some further <pcre.h> cases that I changed to "pcre.h".
1441
1442 9. When pcregrep was used with the --colour option, it missed the line ending
1443 sequence off the lines that it output.
1444
1445 10. It was pointed out to me that arrays of string pointers cause lots of
1446 relocations when a shared library is dynamically loaded. A technique of
1447 using a single long string with a table of offsets can drastically reduce
1448 these. I have refactored PCRE in four places to do this. The result is
1449 dramatic:
1450
1451 Originally: 290
1452 After changing UCP table: 187
1453 After changing error message table: 43
1454 After changing table of "verbs" 36
1455 After changing table of Posix names 22
1456
1457 Thanks to the folks working on Gregex for glib for this insight.
1458
1459 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1460 unicode-properties was also set.
1461
1462 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1463
1464 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1465 checked only for CRLF.
1466
1467 14. Added casts to pcretest.c to avoid compiler warnings.
1468
1469 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1470
1471 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1472 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1473 entirely. This removes changes made in 7 above.
1474
1475 17. The CMake files have been updated, and there is now more information about
1476 building with CMake in the NON-UNIX-USE document.
1477
1478
1479 Version 7.3 28-Aug-07
1480 ---------------------
1481
1482 1. In the rejigging of the build system that eventually resulted in 7.1, the
1483 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1484 brackets there is not right, since it causes compilers to look for an
1485 installed pcre.h, not the version that is in the source that is being
1486 compiled (which of course may be different). I have changed it back to:
1487
1488 #include "pcre.h"
1489
1490 I have a vague recollection that the change was concerned with compiling in
1491 different directories, but in the new build system, that is taken care of
1492 by the VPATH setting the Makefile.
1493
1494 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1495 when the subject happened to end in the byte 0x85 (e.g. if the last
1496 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1497 characters but of course it shouldn't be taken as a newline when it is part
1498 of another character. The bug was that, for an unlimited repeat of . in
1499 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1500 characters when looking for a newline.
1501
1502 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1503
1504 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1505 in debug output.
1506
1507 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1508 long printing in the pcrecpp unittest when running under MinGW.
1509
1510 6. ESC_K was left out of the EBCDIC table.
1511
1512 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1513 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1514 limit also applies to "virtual nesting" when a pattern is recursive, and in
1515 this case 1000 isn't so big. I have been able to remove this limit at the
1516 expense of backing off one optimization in certain circumstances. Normally,
1517 when pcre_exec() would call its internal match() function recursively and
1518 immediately return the result unconditionally, it uses a "tail recursion"
1519 feature to save stack. However, when a subpattern that can match an empty
1520 string has an unlimited repetition quantifier, it no longer makes this
1521 optimization. That gives it a stack frame in which to save the data for
1522 checking that an empty string has been matched. Previously this was taken
1523 from the 1000-entry workspace that had been reserved. So now there is no
1524 explicit limit, but more stack is used.
1525
1526 8. Applied Daniel's patches to solve problems with the import/export magic
1527 syntax that is required for Windows, and which was going wrong for the
1528 pcreposix and pcrecpp parts of the library. These were overlooked when this
1529 problem was solved for the main library.
1530
1531 9. There were some crude static tests to avoid integer overflow when computing
1532 the size of patterns that contain repeated groups with explicit upper
1533 limits. As the maximum quantifier is 65535, the maximum group length was
1534 set at 30,000 so that the product of these two numbers did not overflow a
1535 32-bit integer. However, it turns out that people want to use groups that
1536 are longer than 30,000 bytes (though not repeat them that many times).
1537 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1538 made it possible to implement the integer overflow checks in a much more
1539 dynamic way, which I have now done. The artificial limitation on group
1540 length has been removed - we now have only the limit on the total length of
1541 the compiled pattern, which depends on the LINK_SIZE setting.
1542
1543 10. Fixed a bug in the documentation for get/copy named substring when
1544 duplicate names are permitted. If none of the named substrings are set, the
1545 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1546 empty string.
1547
1548 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1549 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1550 because the ] is interpreted as the first data character and the
1551 terminating ] is not found. PCRE has been made compatible with Perl in this
1552 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1553 cause memory overwriting.
1554
1555 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1556 string has been matched (to stop an infinite loop). It was not recognizing
1557 a conditional subpattern that could match an empty string if that
1558 subpattern was within another subpattern. For example, it looped when
1559 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1560 condition was not nested. This bug has been fixed.
1561
1562 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1563 past the start of the subject in the presence of bytes with the top bit
1564 set, for example "\x8aBCD".
1565
1566 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1567 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1568
1569 14. Optimized (?!) to (*FAIL).
1570
1571 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1572 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1573 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1574 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1575 does: it's just the validity check that is more restrictive.
1576
1577 16. Inserted checks for integer overflows during escape sequence (backslash)
1578 processing, and also fixed erroneous offset values for syntax errors during
1579 backslash processing.
1580
1581 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1582 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1583
1584 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1585 caused an overrun.
1586
1587 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1588 something other than just ASCII characters) inside a group that had an
1589 unlimited repeat caused a loop at compile time (while checking to see
1590 whether the group could match an empty string).
1591
1592 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1593 [\P{Any}] did so. (Error in the code for printing property names.)
1594
1595 21. An orphan \E inside a character class could cause a crash.
1596
1597 22. A repeated capturing bracket such as (A)? could cause a wild memory
1598 reference during compilation.
1599
1600 23. There are several functions in pcre_compile() that scan along a compiled
1601 expression for various reasons (e.g. to see if it's fixed length for look
1602 behind). There were bugs in these functions when a repeated \p or \P was
1603 present in the pattern. These operators have additional parameters compared
1604 with \d, etc, and these were not being taken into account when moving along
1605 the compiled data. Specifically:
1606
1607 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1608 length.
1609
1610 (b) An item such as \pL+ within a repeated group could cause crashes or
1611 loops.
1612
1613 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1614 "reference to non-existent subpattern" error.
1615
1616 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1617
1618 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1619 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1620
1621 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1622
1623 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1624 character were causing crashes (broken optimization).
1625
1626 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1627 \p or \P) caused a compile-time loop.
1628
1629 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1630 break. For example, the unstudied pattern [\r\n]A does not match the string
1631 "\r\nA" because change 7.0/46 below moves the current point on by two
1632 characters after failing to match at the start. However, the pattern \nA
1633 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1634 the same is true. There doesn't seem any very clean way out of this, but
1635 what I have chosen to do makes the common cases work: PCRE now takes note
1636 of whether there can be an explicit match for \r or \n anywhere in the
1637 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1638 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1639 pattern has explicit CR or LF references.
1640
1641 29. Added (*CR) etc for changing newline setting at start of pattern.
1642
1643
1644 Version 7.2 19-Jun-07
1645 ---------------------
1646
1647 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1648 which is apparently normally available under Windows.
1649
1650 2. Re-jig the pcregrep tests with different newline settings in an attempt
1651 to make them independent of the local environment's newline setting.
1652
1653 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1654
1655 4. Some of the "internals" tests were previously cut out when the link size
1656 was not 2, because the output contained actual offsets. The recent new
1657 "Z" feature of pcretest means that these can be cut out, making the tests
1658 usable with all link sizes.
1659
1660 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1661 stack recursion. This gives a massive performance boost under BSD, but just
1662 a small improvement under Linux. However, it saves one field in the frame
1663 in all cases.
1664
1665 6. Added more features from the forthcoming Perl 5.10:
1666
1667 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1668 recursion call. It refers to the nth most recently opened parentheses.
1669
1670 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1671 to be opened parentheses.
1672
1673 (c) Conditions that refer to capturing parentheses can be specified
1674 relatively, for example, (?(-2)... or (?(+3)...
1675
1676 (d) \K resets the start of the current match so that everything before
1677 is not part of it.
1678
1679 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1680
1681 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1682 reference syntax.
1683
1684 (g) (?| introduces a group in which the numbering of parentheses in each
1685 alternative starts with the same number.
1686
1687 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1688
1689 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1690 PCRE_INFO_JCHANGED.
1691
1692 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1693 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1694 for detecting groups that can match an empty string.
1695
1696 9. A pattern with a very large number of alternatives (more than several
1697 hundred) was running out of internal workspace during the pre-compile
1698 phase, where pcre_compile() figures out how much memory will be needed. A
1699 bit of new cunning has reduced the workspace needed for groups with
1700 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1701 workspace instead of running out of the 4096 that are available.
1702
1703 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1704
1705 11. Applied patch from Google to remove an optimization that didn't quite work.
1706 The report of the bug said:
1707
1708 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1709 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1710 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1711
1712 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1713 it matched the wrong number of bytes.
1714
1715
1716 Version 7.1 24-Apr-07
1717 ---------------------
1718
1719 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1720 that is more "standard", making use of automake and other Autotools. There
1721 is some re-arrangement of the files and adjustment of comments consequent
1722 on this.
1723
1724 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1725 for recursive directory scanning broke on some systems because the files
1726 are not scanned in any specific order and on different systems the order
1727 was different. A call to "sort" has been inserted into RunGrepTest for the
1728 approprate test as a short-term fix. In the longer term there may be an
1729 alternative.
1730
1731 3. I had an email from Eric Raymond about problems translating some of PCRE's
1732 man pages to HTML (despite the fact that I distribute HTML pages, some
1733 people do their own conversions for various reasons). The problems
1734 concerned the use of low-level troff macros .br and .in. I have therefore
1735 removed all such uses from the man pages (some were redundant, some could
1736 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1737 HTML has been updated to handle .nf/.fi and to complain if it encounters
1738 .br or .in.
1739
1740 4. Updated comments in configure.ac that get placed in config.h.in and also
1741 arranged for config.h to be included in the distribution, with the name
1742 config.h.generic, for the benefit of those who have to compile without
1743 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1744
1745 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1746 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1747 makevp.bat for the current PCRE, using the additional files
1748 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1749
1750 6. A Windows user reported a minor discrepancy with test 2, which turned out
1751 to be caused by a trailing space on an input line that had got lost in his
1752 copy. The trailing space was an accident, so I've just removed it.
1753
1754 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1755 that is needed.
1756
1757 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1758 as "const" (a) because they are and (b) because it helps the PHP
1759 maintainers who have recently made a script to detect big data structures
1760 in the php code that should be moved to the .rodata section. I remembered
1761 to update Builducptable as well, so it won't revert if ucptable.h is ever
1762 re-created.
1763
1764 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1765 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1766 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1767 support is not required. This saves 1.5-2K of code, which is important in
1768 some applications.
1769
1770 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1771 so as not to refer to the tables, even though these functions will never be
1772 called when UTF-8 support is disabled. Otherwise there are problems with a
1773 shared library.
1774
1775 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1776
1777 (a) It was defining its arguments as char * instead of void *.
1778
1779 (b) It was assuming that all moves were upwards in memory; this was true
1780 a long time ago when I wrote it, but is no longer the case.
1781
1782 The emulated memove() is provided for those environments that have neither
1783 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1784 is clearly not the case, as these two bugs were recently reported.
1785
1786 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1787 and Detrail to create the HTML documentation, the .txt form of the man
1788 pages, and it removes trailing spaces from listed files. It also creates
1789 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1790 case, it wraps all the #defines with #ifndefs. This script should be run
1791 before "make dist".
1792
1793 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1794 with Unicode property support.
1795
1796 (a) For a maximizing quantifier, if the two different cases of the
1797 character were of different lengths in their UTF-8 codings (there are
1798 some cases like this - I found 11), and the matching function had to
1799 back up over a mixture of the two cases, it incorrectly assumed they
1800 were both the same length.
1801
1802 (b) When PCRE was configured to use the heap rather than the stack for
1803 recursion during matching, it was not correctly preserving the data for
1804 the other case of a UTF-8 character when checking ahead for a match
1805 while processing a minimizing repeat. If the check also involved
1806 matching a wide character, but failed, corruption could cause an
1807 erroneous result when trying to check for a repeat of the original
1808 character.
1809
1810 13. Some tidying changes to the testing mechanism:
1811
1812 (a) The RunTest script now detects the internal link size and whether there
1813 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1814 values substituted by "configure". (The RunGrepTest script already did
1815 this for UTF-8.) The configure.ac script no longer substitutes the
1816 relevant variables.
1817
1818 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1819 with length and offset values. This means that the output is different
1820 for different internal link sizes. Test 2 is skipped for link sizes
1821 other than 2 because of this, bypassing the problem. Unfortunately,
1822 there was also a test in test 3 (the locale tests) that used /B and
1823 failed for link sizes other than 2. Rather than cut the whole test out,
1824 I have added a new /Z option to pcretest that replaces the length and
1825 offset values with spaces. This is now used to make test 3 independent
1826 of link size. (Test 2 will be tidied up later.)
1827
1828 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1829 segmentation fault instead of returning the appropriate error message.
1830
1831 15. In multiline mode when the newline sequence was set to "any", the pattern
1832 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1833 This doesn't seem right; it now treats the CRLF combination as the line
1834 ending, and so does not match in that case. It's only a pattern such as ^$
1835 that would hit this one: something like ^ABC$ would have failed after \r
1836 and then tried again after \r\n.
1837
1838 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1839 in an attempt to make files that differ only in their line terminators
1840 compare equal. This works on Linux.
1841
1842 17. Under certain error circumstances pcregrep might try to free random memory
1843 as it exited. This is now fixed, thanks to valgrind.
1844
1845 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1846 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1847 was because its rules for how to advance for /g after matching an empty
1848 string at the end of a line did not allow for this case. They now check for
1849 it specially.
1850
1851 20. pcretest is supposed to handle patterns and data of any length, by
1852 extending its buffers when necessary. It was getting this wrong when the
1853 buffer for a data line had to be extended.
1854
1855 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1856 CRLF as a newline sequence.
1857
1858 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1859 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1860 I have nevertheless tidied it up.
1861
1862 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1863
1864 24. Added a man page for pcre-config.
1865
1866
1867 Version 7.0 19-Dec-06
1868 ---------------------
1869
1870 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1871 moving to gcc 4.1.1.
1872
1873 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1874 sys/time.h, which is documented as needed for this function. It doesn't
1875 seem to matter on Linux, but it showed up on some releases of OS X.
1876
1877 3. It seems that there are systems where bytes whose values are greater than
1878 127 match isprint() in the "C" locale. The "C" locale should be the
1879 default when a C program starts up. In most systems, only ASCII printing
1880 characters match isprint(). This difference caused the output from pcretest
1881 to vary, making some of the tests fail. I have changed pcretest so that:
1882
1883 (a) When it is outputting text in the compiled version of a pattern, bytes
1884 other than 32-126 are always shown as hex escapes.
1885
1886 (b) When it is outputting text that is a matched part of a subject string,
1887 it does the same, unless a different locale has been set for the match
1888 (using the /L modifier). In this case, it uses isprint() to decide.
1889
1890 4. Fixed a major bug that caused incorrect computation of the amount of memory
1891 required for a compiled pattern when options that changed within the
1892 pattern affected the logic of the preliminary scan that determines the
1893 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1894 that the computed length was too small. The symptoms of this bug were
1895 either the PCRE error "internal error: code overflow" from pcre_compile(),
1896 or a glibc crash with a message such as "pcretest: free(): invalid next
1897 size (fast)". Examples of patterns that provoked this bug (shown in
1898 pcretest format) are:
1899
1900 /(?-x: )/x
1901 /(?x)(?-x: \s*#\s*)/
1902 /((?i)[\x{c0}])/8
1903 /(?i:[\x{c0}])/8
1904
1905 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1906 is now done differently.
1907
1908 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1909 wrapper classes; (b) implement a new function in the C++ scanner that is
1910 more efficient than the old way of doing things because it avoids levels of
1911 recursion in the regex matching; (c) add a paragraph to the documentation
1912 for the FullMatch() function.
1913
1914 6. The escape sequence \n was being treated as whatever was defined as
1915 "newline". Not only was this contrary to the documentation, which states
1916 that \n is character 10 (hex 0A), but it also went horribly wrong when
1917 "newline" was defined as CRLF. This has been fixed.
1918
1919 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1920 was being set to -1 for the "end of line" case (supposedly a value that no
1921 character can have). Though this value is never used (the check for end of
1922 line is "zero bytes in current character"), it caused compiler complaints.
1923 I've changed it to 0xffffffff.
1924
1925 8. In pcre_version.c, the version string was being built by a sequence of
1926 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1927 string (as it is for production releases) called a macro with an empty
1928 argument. The C standard says the result of this is undefined. The gcc
1929 compiler treats it as an empty string (which was what was wanted) but it is
1930 reported that Visual C gives an error. The source has been hacked around to
1931 avoid this problem.
1932
1933 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1934 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1935 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1936 of them did).
1937
1938 10. Originally, pcretest opened its input and output without "b"; then I was
1939 told that "b" was needed in some environments, so it was added for release
1940 5.0 to both the input and output. (It makes no difference on Unix-like
1941 systems.) Later I was told that it is wrong for the input on Windows. I've
1942 now abstracted the modes into two macros, to make it easier to fiddle with
1943 them, and removed "b" from the input mode under Windows.
1944
1945 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1946
1947 12. Added -help and --help to pcretest as an official way of being reminded
1948 of the options.
1949
1950 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1951 and pcrecpp.cc because they annoy compilers at high warning levels.
1952
1953 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1954
1955 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1956 scripts are not C programs :-) and which was not noticed because it works
1957 on Linux.
1958
1959 16. pcretest is supposed to handle any length of pattern and data line (as one
1960 line or as a continued sequence of lines) by extending its input buffer if
1961 necessary. This feature was broken for very long pattern lines, leading to
1962 a string of junk being passed to pcre_compile() if the pattern was longer
1963 than about 50K.
1964
1965 17. I have done a major re-factoring of the way pcre_compile() computes the
1966 amount of memory needed for a compiled pattern. Previously, there was code
1967 that made a preliminary scan of the pattern in order to do this. That was
1968 OK when PCRE was new, but as the facilities have expanded, it has become
1969 harder and harder to keep it in step with the real compile phase, and there
1970 have been a number of bugs (see for example, 4 above). I have now found a
1971 cunning way of running the real compile function in a "fake" mode that
1972 enables it to compute how much memory it would need, while actually only
1973 ever using a few hundred bytes of working memory and without too many
1974 tests of the mode. This should make future maintenance and development
1975 easier. A side effect of this work is that the limit of 200 on the nesting
1976 depth of parentheses has been removed (though this was never a serious
1977 limitation, I suspect). However, there is a downside: pcre_compile() now
1978 runs more slowly than before (30% or more, depending on the pattern). I
1979 hope this isn't a big issue. There is no effect on runtime performance.
1980
1981 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1982 newline (only possible for the last line of a file) and it was a
1983 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1984
1985 19. Added additional timing features to pcretest. (1) The -tm option now times
1986 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1987 separate command line item, by a number that specifies the number of
1988 repeats to use when timing. The default is 50000; this gives better
1989 precision, but takes uncomfortably long for very large patterns.
1990
1991 20. Extended pcre_study() to be more clever in cases where a branch of a
1992 subpattern has no definite first character. For example, (a*|b*)[cd] would
1993 previously give no result from pcre_study(). Now it recognizes that the
1994 first character must be a, b, c, or d.
1995
1996 21. There was an incorrect error "recursive call could loop indefinitely" if
1997 a subpattern (or the entire pattern) that was being tested for matching an
1998 empty string contained only one non-empty item after a nested subpattern.
1999 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2000 incorrectly, because the \d was being skipped in the check.
2001
2002 22. The pcretest program now has a new pattern option /B and a command line
2003 option -b, which is equivalent to adding /B to every pattern. This causes
2004 it to show the compiled bytecode, without the additional information that
2005 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2006 is the same as /B/I).
2007
2008 23. A new optimization is now able automatically to treat some sequences such
2009 as a*b as a*+b. More specifically, if something simple (such as a character
2010 or a simple class like \d) has an unlimited quantifier, and is followed by
2011 something that cannot possibly match the quantified thing, the quantifier
2012 is automatically "possessified".
2013
2014 24. A recursive reference to a subpattern whose number was greater than 39
2015 went wrong under certain circumstances in UTF-8 mode. This bug could also
2016 have affected the operation of pcre_study().
2017
2018 25. Realized that a little bit of performance could be had by replacing
2019 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2020
2021 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2022
2023 27. Possessive quantifiers such as a++ were previously implemented by turning
2024 them into atomic groups such as ($>a+). Now they have their own opcodes,
2025 which improves performance. This includes the automatically created ones
2026 from 23 above.
2027
2028 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2029 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2030 the first matched character to be a colon. This applied both to named and
2031 numbered groups.
2032
2033 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2034
2035 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2036 building PCRE on Windows easier, so I have included it in the distribution.
2037
2038 31. There is now a check in pcretest against a ridiculously large number being
2039 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2040 loop, the loop is abandoned.
2041
2042 32. Forward references to subpatterns in conditions such as (?(2)...) where
2043 subpattern 2 is defined later cause pcre_compile() to search forwards in
2044 the pattern for the relevant set of parentheses. This search went wrong
2045 when there were unescaped parentheses in a character class, parentheses
2046 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2047
2048 33. "Subroutine" calls and backreferences were previously restricted to
2049 referencing subpatterns earlier in the regex. This restriction has now
2050 been removed.
2051
2052 34. Added a number of extra features that are going to be in Perl 5.10. On the
2053 whole, these are just syntactic alternatives for features that PCRE had
2054 previously implemented using the Python syntax or my own invention. The
2055 other formats are all retained for compatibility.
2056
2057 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2058 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2059 also .NET compatible.
2060
2061 (b) A recursion or subroutine call to a named group can now be defined as
2062 (?&name) as well as (?P>name).
2063
2064 (c) A backreference to a named group can now be defined as \k<name> or
2065 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2066 5.10, are also .NET compatible.
2067
2068 (d) A conditional reference to a named group can now use the syntax
2069 (?(<name>) or (?('name') as well as (?(name).
2070
2071 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2072 groups (named and numbered) that are never evaluated inline, but can be
2073 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2074 is always false. There may be only one alternative in such a group.
2075
2076 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2077 as the simple (?(R). The condition is true only if the most recent
2078 recursion is that of the given number or name. It does not search out
2079 through the entire recursion stack.
2080
2081 (g) The escape \gN or \g{N} has been added, where N is a positive or
2082 negative number, specifying an absolute or relative reference.
2083
2084 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2085 some "unreachable code" warnings.
2086
2087 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2088 things, this adds five new scripts.
2089
2090 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2091 There were also incompatibilities regarding the handling of \Q..\E inside
2092 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2093 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2094
2095 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2096 matches an empty string, and forcibly breaks the loop. There were bugs in
2097 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2098 against aaaa the result was just "a" rather than "aaaa", for example. Two
2099 separate and independent bugs (that affected different cases) have been
2100 fixed.
2101
2102 39. Refactored the code to abolish the use of different opcodes for small
2103 capturing bracket numbers. This is a tidy that I avoided doing when I
2104 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2105 The new approach is not only tidier, it makes it possible to reduce the
2106 memory needed to fix the previous bug (38).
2107
2108 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2109 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2110 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2111 mode.
2112
2113 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2114 report.
2115
2116 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2117 copy construction and assignment in the C++ wrapper.
2118
2119 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2120 couple of bugs that could have given wrong results in the "--newline=crlf"
2121 case.
2122
2123 44. Added a number of casts and did some reorganization of signed/unsigned int
2124 variables following suggestions from Dair Grant. Also renamed the variable
2125 "this" as "item" because it is a C++ keyword.
2126
2127 45. Arranged for dftables to add
2128
2129 #include "pcre_internal.h"
2130
2131 to pcre_chartables.c because without it, gcc 4.x may remove the array
2132 definition from the final binary if PCRE is built into a static library and
2133 dead code stripping is activated.
2134
2135 46. For an unanchored pattern, if a match attempt fails at the start of a
2136 newline sequence, and the newline setting is CRLF or ANY, and the next two
2137 characters are CRLF, advance by two characters instead of one.
2138
2139
2140 Version 6.7 04-Jul-06
2141 ---------------------
2142
2143 1. In order to handle tests when input lines are enormously long, pcretest has
2144 been re-factored so that it automatically extends its buffers when
2145 necessary. The code is crude, but this _is_ just a test program. The
2146 default size has been increased from 32K to 50K.
2147
2148 2. The code in pcre_study() was using the value of the re argument before
2149 testing it for NULL. (Of course, in any sensible call of the function, it
2150 won't be NULL.)
2151
2152 3. The memmove() emulation function in pcre_internal.h, which is used on
2153 systems that lack both memmove() and bcopy() - that is, hardly ever -
2154 was missing a "static" storage class specifier.
2155
2156 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2157 containing an extended class (one that cannot be represented by a bitmap
2158 because it contains high-valued characters or Unicode property items, e.g.
2159 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2160 pattern, but PCRE should not loop if you do not (it no longer does).
2161 [Detail: two cases were found: (a) a repeated subpattern containing an
2162 extended class; (b) a recursive reference to a subpattern that followed a
2163 previous extended class. It wasn't skipping over the extended class
2164 correctly when UTF-8 mode was not set.]
2165
2166 5. A negated single-character class was not being recognized as fixed-length
2167 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2168 compile error "lookbehind assertion is not fixed length".
2169
2170 6. The RunPerlTest auxiliary script was showing an unexpected difference
2171 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2172 write a Perl script that can interpret lines of an input file either as
2173 byte characters or as UTF-8, which is what "perltest" was being required to
2174 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2175 can't do is switch easily at run time between having the "use utf8;" pragma
2176 or not. In the end, I fudged it by using the RunPerlTest script to insert
2177 "use utf8;" explicitly for the UTF-8 tests.
2178
2179 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2180 the end of the subject string, contrary to the documentation and to what
2181 Perl does. This was true of both matching functions. Now it matches only at
2182 the start of the subject and immediately after *internal* newlines.
2183
2184 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2185 a pointer to an int instead of a pointer to an unsigned long int. This
2186 caused problems on 64-bit systems.
2187
2188 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2189 instance of the 'standard' template library not being so standard".
2190
2191 10. There was no check on the number of named subpatterns nor the maximum
2192 length of a subpattern name. The product of these values is used to compute
2193 the size of the memory block for a compiled pattern. By supplying a very
2194 long subpattern name and a large number of named subpatterns, the size
2195 computation could be caused to overflow. This is now prevented by limiting
2196 the length of names to 32 characters, and the number of named subpatterns
2197 to 10,000.
2198
2199 11. Subpatterns that are repeated with specific counts have to be replicated in
2200 the compiled pattern. The size of memory for this was computed from the
2201 length of the subpattern and the repeat count. The latter is limited to
2202 65535, but there was no limit on the former, meaning that integer overflow
2203 could in principle occur. The compiled length of a repeated subpattern is
2204 now limited to 30,000 bytes in order to prevent this.
2205
2206 12. Added the optional facility to have named substrings with the same name.
2207
2208 13. Added the ability to use a named substring as a condition, using the
2209 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2210 are numbers (not recommended). Forward references are permitted.
2211
2212 14. Added forward references in named backreferences (if you see what I mean).
2213
2214 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2215 pattern could run off the end of the subject. For example, the pattern
2216 "(?s)(.{1,5})"8 did this with the subject "ab".
2217
2218 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2219 PCRE_CASELESS was set when matching characters that were quantified with ?
2220 or *.
2221
2222 17. A character class other than a single negated character that had a minimum
2223 but no maximum quantifier - for example [ab]{6,} - was not handled
2224 correctly by pce_dfa_exec(). It would match only one character.
2225
2226 18. A valid (though odd) pattern that looked like a POSIX character
2227 class but used an invalid character after [ (for example [[,abc,]]) caused
2228 pcre_compile() to give the error "Failed: internal error: code overflow" or
2229 in some cases to crash with a glibc free() error. This could even happen if
2230 the pattern terminated after [[ but there just happened to be a sequence of
2231 letters, a binary zero, and a closing ] in the memory that followed.
2232
2233 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2234 over the years. Originally (before any Unicode support), just the bottom 8
2235 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2236 output from "man perlunicode" includes this:
2237
2238 The regular expression compiler produces polymorphic opcodes. That
2239 is, the pattern adapts to the data and automatically switches to
2240 the Unicode character scheme when presented with Unicode data--or
2241 instead uses a traditional byte scheme when presented with byte
2242 data.
2243
2244 Sadly, a wide octal escape does not cause a switch, and in a string with
2245 no other multibyte characters, these octal escapes are treated as before.
2246 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2247 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2248 Unicode string.
2249
2250 I have not perpetrated such confusion in PCRE. Up till now, it took just
2251 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2252 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2253 translate to the appropriate multibyte character.
2254
2255 29. Applied some refactoring to reduce the number of warnings from Microsoft
2256 and Borland compilers. This has included removing the fudge introduced
2257 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2258 a warning about an unused variable.
2259
2260 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2261 characters since release 4.0, because Perl (from release 5.004) does not.
2262 [Or at least, is documented not to: some releases seem to be in conflict
2263 with the documentation.] However, when a pattern was studied with
2264 pcre_study() and all its branches started with \s, PCRE still included VT
2265 as a possible starting character. Of course, this did no harm; it just
2266 caused an unnecessary match attempt.
2267
2268 22. Removed a now-redundant internal flag bit that recorded the fact that case
2269 dependency changed within the pattern. This was once needed for "required
2270 byte" processing, but is no longer used. This recovers a now-scarce options
2271 bit. Also moved the least significant internal flag bit to the most-
2272 significant bit of the word, which was not previously used (hangover from
2273 the days when it was an int rather than a uint) to free up another bit for
2274 the future.
2275
2276 23. Added support for CRLF line endings as well as CR and LF. As well as the
2277 default being selectable at build time, it can now be changed at runtime
2278 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2279 specify that it is scanning data with non-default line endings.
2280
2281 24. Changed the definition of CXXLINK to make it agree with the definition of
2282 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2283
2284 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2285 recursions. This makes a big different to stack usage for some patterns.
2286
2287 26. If a subpattern containing a named recursion or subroutine reference such
2288 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2289 the space required for the compiled pattern went wrong and gave too small a
2290 value. Depending on the environment, this could lead to "Failed: internal
2291 error: code overflow at offset 49" or "glibc detected double free or
2292 corruption" errors.
2293
2294 27. Applied patches from Google (a) to support the new newline modes and (b) to
2295 advance over multibyte UTF-8 characters in GlobalReplace.
2296
2297 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2298 difference for some implementation of PCRE in some Windows version.
2299
2300 29. Added some extra testing facilities to pcretest:
2301
2302 \q<number> in a data line sets the "match limit" value
2303 \Q<number> in a data line sets the "match recursion limt" value
2304 -S <number> sets the stack size, where <number> is in megabytes
2305
2306 The -S option isn't available for Windows.
2307
2308
2309 Version 6.6 06-Feb-06
2310 ---------------------
2311
2312 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2313 in pcreposix.h. I have copied the definition from pcre.h.
2314
2315 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2316 because pcre.h is no longer a built file.
2317
2318 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2319 not normally included in the compiled code.
2320
2321
2322 Version 6.5 01-Feb-06
2323 ---------------------
2324
2325 1. When using the partial match feature with pcre_dfa_exec(), it was not
2326 anchoring the second and subsequent partial matches at the new starting
2327 point. This could lead to incorrect results. For example, with the pattern
2328 /1234/, partially matching against "123" and then "a4" gave a match.
2329
2330 2. Changes to pcregrep:
2331
2332 (a) All non-match returns from pcre_exec() were being treated as failures
2333 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2334 error message is output. Some extra information is given for the
2335 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2336 probably the only errors that are likely to be caused by users (by
2337 specifying a regex that has nested indefinite repeats, for instance).
2338 If there are more than 20 of these errors, pcregrep is abandoned.
2339
2340 (b) A binary zero was treated as data while matching, but terminated the
2341 output line if it was written out. This has been fixed: binary zeroes
2342 are now no different to any other data bytes.
2343
2344 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2345 used to set a locale for matching. The --locale=xxxx long option has
2346 been added (no short equivalent) to specify a locale explicitly on the
2347 pcregrep command, overriding the environment variables.
2348
2349 (d) When -B was used with -n, some line numbers in the output were one less
2350 than they should have been.
2351
2352 (e) Added the -o (--only-matching) option.
2353
2354 (f) If -A or -C was used with -c (count only), some lines of context were
2355 accidentally printed for the final match.
2356
2357 (g) Added the -H (--with-filename) option.
2358
2359 (h) The combination of options -rh failed to suppress file names for files
2360 that were found from directory arguments.
2361
2362 (i) Added the -D (--devices) and -d (--directories) options.
2363
2364 (j) Added the -F (--fixed-strings) option.
2365
2366 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2367
2368 (l) Added the --colo(u)r option.
2369
2370 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2371 is not present by default.
2372
2373 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2374 items such as (?R) or (?1), when the recursion could match a number of
2375 alternatives. If it matched one of the alternatives, but subsequently,
2376 outside the recursion, there was a failure, the code tried to back up into
2377 the recursion. However, because of the way PCRE is implemented, this is not
2378 possible, and the result was an incorrect result from the match.
2379
2380 In order to prevent this happening, the specification of recursion has
2381 been changed so that all such subpatterns are automatically treated as
2382 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2383
2384 4. I had overlooked the fact that, in some locales, there are characters for
2385 which isalpha() is true but neither isupper() nor islower() are true. In
2386 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2387 and ordfeminine) are like this. This affected the treatment of \w and \W
2388 when they appeared in character classes, but not when they appeared outside
2389 a character class. The bit map for "word" characters is now created
2390 separately from the results of isalnum() instead of just taking it from the
2391 upper, lower, and digit maps. (Plus the underscore character, of course.)
2392
2393 5. The above bug also affected the handling of POSIX character classes such as
2394 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2395 permanent tables. Instead, the bit maps for such a class were previously
2396 created as the appropriate unions of the upper, lower, and digit bitmaps.
2397 Now they are created by subtraction from the [[:word:]] class, which has
2398 its own bitmap.
2399
2400 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2401 It is created by subtracting the vertical space characters (\x09, \x0a,
2402 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2403 subtraction was done in the overall bitmap for a character class, meaning
2404 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2405 be recognized. This bug has been fixed.
2406
2407 7. Patches from the folks at Google:
2408
2409 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2410 real life, but is still worth protecting against".
2411
2412 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2413 regular expressions".
2414
2415 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2416 have it.
2417
2418 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2419 "configure" and the latter not, in order to fix a problem somebody had
2420 with compiling the Arg class on HP-UX.
2421
2422 (e) Improve the error-handling of the C++ wrapper a little bit.
2423
2424 (f) New tests for checking recursion limiting.
2425
2426 8. The pcre_memmove() function, which is used only if the environment does not
2427 have a standard memmove() function (and is therefore rarely compiled),
2428 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2429 returning a result (though PCRE never actually uses the result).
2430
2431 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2432 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2433 returned instead of calling malloc() with an overflowing number that would
2434 most likely cause subsequent chaos.
2435
2436 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2437
2438 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2439 with this option is matched, the nmatch and pmatch options of regexec() are
2440 ignored.
2441
2442 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2443 provided in case anyone wants to the the POSIX interface with UTF-8
2444 strings.
2445
2446 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2447 C++ linking (needed for some HP-UX environments).
2448
2449 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2450 (unused parameter) and in the pcre_printint() function (omitted "default"
2451 switch label when the default is to do nothing).
2452
2453 15. Added some code to make it possible, when PCRE is compiled as a C++
2454 library, to replace subject pointers for pcre_exec() with a smart pointer
2455 class, thus making it possible to process discontinuous strings.
2456
2457 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2458 much the same function. They were added by different people who were trying
2459 to make PCRE easy to compile on non-Unix systems. It has been suggested
2460 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2461 for compiling on Windows systems. I have therefore replaced it with
2462 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2463 defaults to "extern" for C or "extern C" for C++, which works fine on
2464 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2465 SCOPE with something explicit in config.h. In addition:
2466
2467 (a) pcreposix.h still had just "extern" instead of either of these macros;
2468 I have replaced it with PCRE_DATA_SCOPE.
2469
2470 (b) Functions such as _pcre_xclass(), which are internal to the library,
2471 but external in the C sense, all had PCRE_EXPORT in their definitions.
2472 This is apparently wrong for the Windows case, so I have removed it.
2473 (It makes no difference on Unix-like systems.)
2474
2475 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2476 of recursive calls to match(). This is different to MATCH_LIMIT because
2477 that limits the total number of calls to match(), not all of which increase
2478 the depth of recursion. Limiting the recursion depth limits the amount of
2479 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2480 when PCRE is compiled, and changed at run time. A patch from Google adds
2481 this functionality to the C++ interface.
2482
2483 18. Changes to the handling of Unicode character properties:
2484
2485 (a) Updated the table to Unicode 4.1.0.
2486
2487 (b) Recognize characters that are not in the table as "Cn" (undefined).
2488
2489 (c) I revised the way the table is implemented to a much improved format
2490 which includes recognition of ranges. It now supports the ranges that
2491 are defined in UnicodeData.txt, and it also amalgamates other
2492 characters into ranges. This has reduced the number of entries in the
2493 table from around 16,000 to around 3,000, thus reducing its size
2494 considerably. I realized I did not need to use a tree structure after
2495 all - a binary chop search is just as efficient. Having reduced the
2496 number of entries, I extended their size from 6 bytes to 8 bytes to
2497 allow for more data.
2498
2499 (d) Added support for Unicode script names via properties such as \p{Han}.
2500
2501 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2502 matching that character.
2503
2504 20. When matching a repeated Unicode property with a minimum greater than zero,
2505 (for example \pL{2,}), PCRE could look past the end of the subject if it
2506 reached it while seeking the minimum number of characters. This could
2507 happen only if some of the characters were more than one byte long, because
2508 there is a check for at least the minimum number of bytes.
2509
2510 21. Refactored the implementation of \p and \P so as to be more general, to
2511 allow for more different types of property in future. This has changed the
2512 compiled form incompatibly. Anybody with saved compiled patterns that use
2513 \p or \P will have to recompile them.
2514
2515 22. Added "Any" and "L&" to the supported property types.
2516
2517 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2518 but give a compile time error if the value is greater than 0xff.
2519
2520 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2521 accidentally not being installed or uninstalled.
2522
2523 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2524 made were to insert the current release number. This seemed silly, because
2525 it made things harder for people building PCRE on systems that don't run
2526 "configure". I have turned pcre.h into a distributed file, no longer built
2527 by "configure", with the version identification directly included. There is
2528 no longer a pcre.h.in file.
2529
2530 However, this change necessitated a change to the pcre-config script as
2531 well. It is built from pcre-config.in, and one of the substitutions was the
2532 release number. I have updated configure.ac so that ./configure now finds
2533 the release number by grepping pcre.h.
2534
2535 26. Added the ability to run the tests under valgrind.
2536
2537
2538 Version 6.4 05-Sep-05
2539 ---------------------
2540
2541 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2542 "--" to be printed when multiple files were scanned, even when none of the
2543 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2544 consider it to be a bug, and have restored the previous behaviour.
2545
2546 2. A couple of code tidies to get rid of compiler warnings.
2547
2548 3. The pcretest program used to cheat by referring to symbols in the library
2549 whose names begin with _pcre_. These are internal symbols that are not
2550 really supposed to be visible externally, and in some environments it is
2551 possible to suppress them. The cheating is now confined to including
2552 certain files from the library's source, which is a bit cleaner.
2553
2554 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2555 file's purpose clearer.
2556
2557 5. Reorganized pcre_ucp_findchar().
2558
2559
2560 Version 6.3 15-Aug-05
2561 ---------------------
2562
2563 1. The file libpcre.pc.in did not have general read permission in the tarball.
2564
2565 2. There were some problems when building without C++ support:
2566
2567 (a) If C++ support was not built, "make install" and "make test" still
2568 tried to test it.
2569
2570 (b) There were problems when the value of CXX was explicitly set. Some
2571 changes have been made to try to fix these, and ...
2572
2573 (c) --disable-cpp can now be used to explicitly disable C++ support.
2574
2575 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2576 backslash in a target when C++ was disabled. This confuses some
2577 versions of "make", apparently. Using an intermediate variable solves
2578 this. (Same for CPP_LOBJ.)
2579
2580 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2581 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2582 necessary on certain architectures.
2583
2584 4. Added a setting of -export-symbols-regex to the link command to remove
2585 those symbols that are exported in the C sense, but actually are local
2586 within the library, and not documented. Their names all begin with
2587 "_pcre_". This is not a perfect job, because (a) we have to except some
2588 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2589 available (and never for static libraries). I have made a note to try to
2590 find a way round (a) in the future.
2591
2592
2593 Version 6.2 01-Aug-05
2594 ---------------------
2595
2596 1. There was no test for integer overflow of quantifier values. A construction
2597 such as {1111111111111111} would give undefined results. What is worse, if
2598 a minimum quantifier for a parenthesized subpattern overflowed and became
2599 negative, the calculation of the memory size went wrong. This could have
2600 led to memory overwriting.
2601
2602 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2603
2604 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2605 operating environments where this matters.
2606
2607 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2608 PCRE options from within the C++ wrapper.
2609
2610 5. Named capturing subpatterns were not being correctly counted when a pattern
2611 was compiled. This caused two problems: (a) If there were more than 100
2612 such subpatterns, the calculation of the memory needed for the whole
2613 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2614 back references of the form \12, where the number was greater than 9, were
2615 not recognized as back references, even though there were sufficient
2616 previous subpatterns.
2617
2618 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2619 versions of gcc, e.g. 2.95.4.
2620
2621
2622 Version 6.1 21-Jun-05
2623 ---------------------
2624
2625 1. There was one reference to the variable "posix" in pcretest.c that was not
2626 surrounded by "#if !defined NOPOSIX".
2627
2628 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2629 the cross-check on the old pcre_info() function, for the benefit of the
2630 cut-down version of PCRE that is currently imported into Exim.
2631
2632 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2633 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2634 patterns that start (?i) but I don't think that matters. The use of (?i) is
2635 just an example; this all applies to the other options as well.
2636
2637 4. Since libtool seems to echo the compile commands it is issuing, the output
2638 from "make" can be reduced a bit by putting "@" in front of each libtool
2639 compile command.
2640
2641 5. Patch from the folks at Google for configure.in to be a bit more thorough
2642 in checking for a suitable C++ installation before trying to compile the
2643 C++ stuff. This should fix a reported problem when a compiler was present,
2644 but no suitable headers.
2645
2646 6. The man pages all had just "PCRE" as their title. I have changed them to
2647 be the relevant file name. I have also arranged that these names are
2648 retained in the file doc/pcre.txt, which is a concatenation in text format
2649 of all the man pages except the little individual ones for each function.
2650
2651 7. The NON-UNIX-USE file had not been updated for the different set of source
2652 files that come with release 6. I also added a few comments about the C++
2653 wrapper.
2654
2655
2656 Version 6.0 07-Jun-05
2657 ---------------------
2658
2659 1. Some minor internal re-organization to help with my DFA experiments.
2660
2661 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2662 didn't matter for the library itself when fully configured, but did matter
2663 when compiling without UCP support, or within Exim, where the ucp files are
2664 not imported.
2665
2666 3. Refactoring of the library code to split up the various functions into
2667 different source modules. The addition of the new DFA matching code (see
2668 below) to a single monolithic source would have made it really too
2669 unwieldy, quite apart from causing all the code to be include in a
2670 statically linked application, when only some functions are used. This is
2671 relevant even without the DFA addition now that patterns can be compiled in
2672 one application and matched in another.
2673
2674 The downside of splitting up is that there have to be some external
2675 functions and data tables that are used internally in different modules of
2676 the library but which are not part of the API. These have all had their
2677 names changed to start with "_pcre_" so that they are unlikely to clash
2678 with other external names.
2679
2680 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2681 a different (DFA) algorithm. Although it is slower than the original
2682 function, it does have some advantages for certain types of matching
2683 problem.
2684
2685 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2686 including restarting after a partial match.
2687
2688 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2689 defined when compiling for Windows was sent to me. I have put it into the
2690 code, though I have no means of testing or verifying it.
2691
2692 7. Added the pcre_refcount() auxiliary function.
2693
2694 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2695 match before or at the first newline in the subject string. In pcretest,
2696 the /f option on a pattern can be used to set this.
2697
2698 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2699 would behave wrongly. This has been present in PCRE since release 4.0.
2700
2701 10. A number of changes to the pcregrep command:
2702
2703 (a) Refactored how -x works; insert ^(...)$ instead of setting
2704 PCRE_ANCHORED and checking the length, in preparation for adding
2705 something similar for -w.
2706
2707 (b) Added the -w (match as a word) option.
2708
2709 (c) Refactored the way lines are read and buffered so as to have more
2710 than one at a time available.
2711
2712 (d) Implemented a pcregrep test script.
2713
2714 (e) Added the -M (multiline match) option. This allows patterns to match
2715 over several lines of the subject. The buffering ensures that at least
2716 8K, or the rest of the document (whichever is the shorter) is available
2717 for matching (and similarly the previous 8K for lookbehind assertions).
2718
2719 (f) Changed the --help output so that it now says
2720
2721 -w, --word-regex(p)
2722
2723 instead of two lines, one with "regex" and the other with "regexp"
2724 because that confused at least one person since the short forms are the
2725 same. (This required a bit of code, as the output is generated
2726 automatically from a table. It wasn't just a text change.)
2727
2728 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2729 option but starts with a hyphen. Could be a pattern or a path name
2730 starting with a hyphen, for instance.
2731
2732 (h) "-" can be given as a file name to represent stdin.
2733
2734 (i) When file names are being printed, "(standard input)" is used for
2735 the standard input, for compatibility with GNU grep. Previously
2736 "<stdin>" was used.
2737
2738 (j) The option --label=xxx can be used to supply a name to be used for
2739 stdin when file names are being printed. There is no short form.
2740
2741 (k) Re-factored the options decoding logic because we are going to add
2742 two more options that take data. Such options can now be given in four
2743 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2744
2745 (l) Added the -A, -B, and -C options for requesting that lines of context
2746 around matches be printed.
2747
2748 (m) Added the -L option to print the names of files that do not contain
2749 any matching lines, that is, the complement of -l.
2750
2751 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2752 continue to scan other files.
2753
2754 (o) The -s option was incorrectly implemented. For compatibility with other
2755 greps, it now suppresses the error message for a non-existent or non-
2756 accessible file (but not the return code). There is a new option called
2757 -q that suppresses the output of matching lines, which was what -s was
2758 previously doing.
2759
2760 (p) Added --include and --exclude options to specify files for inclusion
2761 and exclusion when recursing.
2762
2763 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2764 Hopefully, it now does.
2765
2766 12. Missing cast in pcre_study().
2767
2768 13. Added an "uninstall" target to the makefile.
2769
2770 14. Replaced "extern" in the function prototypes in Makefile.in with
2771 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2772 world, but is set differently for Windows.
2773
2774 15. Added a second compiling function called pcre_compile2(). The only
2775 difference is that it has an extra argument, which is a pointer to an
2776 integer error code. When there is a compile-time failure, this is set
2777 non-zero, in addition to the error test pointer being set to point to an
2778 error message. The new argument may be NULL if no error number is required
2779 (but then you may as well call pcre_compile(), which is now just a
2780 wrapper). This facility is provided because some applications need a
2781 numeric error indication, but it has also enabled me to tidy up the way
2782 compile-time errors are handled in the POSIX wrapper.
2783
2784 16. Added VPATH=.libs to the makefile; this should help when building with one
2785 prefix path and installing with another. (Or so I'm told by someone who
2786 knows more about this stuff than I do.)
2787
2788 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2789 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2790 match everything, including newlines. This is not POSIX-compatible, but
2791 somebody wanted the feature. From pcretest it can be activated by using
2792 both the P and the s flags.
2793
2794 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2795
2796 19. libpcre.pc was being incorrectly installed as executable.
2797
2798 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2799 it now also looks for '\r' so that it will work unmodified on Windows.
2800
2801 21. Added Google's contributed C++ wrapper to the distribution.
2802
2803 22. Added some untidy missing memory free() calls in pcretest, to keep
2804 Electric Fence happy when testing.
2805
2806
2807
2808 Version 5.0 13-Sep-04
2809 ---------------------
2810
2811 1. Internal change: literal characters are no longer packed up into items
2812 containing multiple characters in a single byte-string. Each character
2813 is now matched using a separate opcode. However, there may be more than one
2814 byte in the character in UTF-8 mode.
2815
2816 2. The pcre_callout_block structure has two new fields: pattern_position and
2817 next_item_length. These contain the offset in the pattern to the next match
2818 item, and its length, respectively.
2819
2820 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2821 insertion of callouts before each pattern item. Added the /C option to
2822 pcretest to make use of this.
2823
2824 4. On the advice of a Windows user, the lines
2825
2826 #if defined(_WIN32) || defined(WIN32)
2827 _setmode( _fileno( stdout ), 0x8000 );
2828 #endif /* defined(_WIN32) || defined(WIN32) */
2829
2830 have been added to the source of pcretest. This apparently does useful
2831 magic in relation to line terminators.
2832
2833 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2834 for the benefit of those environments where the "b" makes a difference.
2835
2836 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2837 to know about it. I have put a hack into configure.in that adds in code
2838 to set GCC=yes if CC=icc. This seems to end up at a point in the
2839 generated configure script that is early enough to affect the setting of
2840 compiler options, which is what is needed, but I have no means of testing
2841 whether it really works. (The user who reported this had patched the
2842 generated configure script, which of course I cannot do.)
2843
2844 LATER: After change 22 below (new libtool files), the configure script
2845 seems to know about icc (and also ecc). Therefore, I have commented out
2846 this hack in configure.in.
2847
2848 7. Added support for pkg-config (2 patches were sent in).
2849
2850 8. Negated POSIX character classes that used a combination of internal tables
2851 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2852 [[:^ascii]]. Typically, they would match almost any characters. The other
2853 POSIX classes were not broken in this way.
2854
2855 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2856 to find the match, as PCRE was deluded into thinking that the match had to
2857 start at the start point or following a newline. The same bug applied to
2858 patterns with negative forward assertions or any backward assertions
2859 preceding ".*" at the start, unless the pattern required a fixed first
2860 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2861
2862 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2863 starting at the last subject character, bytes beyond the end of the subject
2864 string were read.
2865
2866 11. Renamed the variable "class" as "classbits" to make life easier for C++
2867 users. (Previously there was a macro definition, but it apparently wasn't
2868 enough.)
2869
2870 12. Added the new field "tables" to the extra data so that tables can be passed
2871 in at exec time, or the internal tables can be re-selected. This allows
2872 a compiled regex to be saved and re-used at a later time by a different
2873 program that might have everything at different addresses.
2874
2875 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2876 -R library as well as a -L library.
2877
2878 14. The debugging options of pcretest (-d on the command line or D on a
2879 pattern) showed incorrect output for anything following an extended class
2880 that contained multibyte characters and which was followed by a quantifier.
2881
2882 15. Added optional support for general category Unicode character properties
2883 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2884 support. It adds about 90K to the size of the library. The meanings of the
2885 inbuilt class escapes such as \d and \s have NOT been changed.
2886
2887 16. Updated pcredemo.c to include calls to free() to release the memory for the
2888 compiled pattern.
2889
2890 17. The generated file chartables.c was being created in the source directory
2891 instead of in the building directory. This caused the build to fail if the
2892 source directory was different from the building directory, and was
2893 read-only.
2894
2895 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2896 file. No doubt somebody will tell me if they don't make sense... Also added
2897 Dan Mooney's comments about building on OpenVMS.
2898
2899 19. Added support for partial matching via the PCRE_PARTIAL option for
2900 pcre_exec() and the \P data escape in pcretest.
2901
2902 20. Extended pcretest with 3 new pattern features:
2903
2904 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2905 write the compiled pattern to the file whose name is "rest-of-line".
2906 This is a straight binary dump of the data, with the saved pointer to
2907 the character tables forced to be NULL. The study data, if any, is
2908 written too. After writing, pcretest reads a new pattern.
2909
2910 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2911 compiled pattern from the given file. There must not be any
2912 occurrences of "<" in the file name (pretty unlikely); if there are,
2913 pcretest will instead treat the initial "<" as a pattern delimiter.
2914 After reading in the pattern, pcretest goes on to read data lines as
2915 usual.
2916
2917 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2918 and 16-bit fields in a compiled pattern, to simulate a pattern that
2919 was compiled on a host of opposite endianness.
2920
2921 21. The pcre-exec() function can now cope with patterns that were compiled on
2922 hosts of opposite endianness, with this restriction:
2923
2924 As for any compiled expression that is saved and used later, the tables
2925 pointer field cannot be preserved; the extra_data field in the arguments
2926 to pcre_exec() should be used to pass in a tables address if a value
2927 other than the default internal tables were used at compile time.
2928
2929 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2930 now diagnosed as an error. Previously, most of the time, a negative number
2931 would have been treated as zero, but if in addition "ovector" was passed as
2932 NULL, a crash could occur.
2933
2934 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2935 new versions from the libtool 1.5 distribution (the last one is a copy of
2936 a file called libtool.m4). This seems to have fixed the need to patch
2937 "configure" to support Darwin 1.3 (which I used to do). However, I still
2938 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2939 workstation).
2940
2941 24. Changed the PCRE licence to be the more standard "BSD" licence.
2942
2943
2944 Version 4.5 01-Dec-03
2945 ---------------------
2946
2947 1. There has been some re-arrangement of the code for the match() function so
2948 that it can be compiled in a version that does not call itself recursively.
2949 Instead, it keeps those local variables that need separate instances for
2950 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2951 needs to "recurse". Keeping track of where control must go is done by means
2952 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2953 hide most of the details from the main code, and operates only if
2954 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2955 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2956 operating.
2957
2958 To make it easier for callers to provide specially tailored get/free
2959 functions for this usage, two new functions, pcre_stack_malloc, and
2960 pcre_stack_free, are used. They are always called in strict stacking order,
2961 and the size of block requested is always the same.
2962
2963 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2964 PCRE has been compiled to use the stack or the heap for recursion. The
2965 -C option of pcretest uses this to show which version is compiled.
2966
2967 A new data escape \S, is added to pcretest; it causes the amounts of store
2968 obtained and freed by both kinds of malloc/free at match time to be added
2969 to the output.
2970
2971 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2972 what's available on my current Linux desktop machine.
2973
2974 3. When matching a UTF-8 string, the test for a valid string at the start has
2975 been extended. If start_offset is not zero, PCRE now checks that it points
2976 to a byte that is the start of a UTF-8 character. If not, it returns
2977 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2978 this is necessary because there may be backward assertions in the pattern.
2979 When matching the same subject several times, it may save resources to use
2980 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2981
2982 4. The code for checking the validity of UTF-8 strings has been tightened so
2983 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2984 containing "overlong sequences".
2985
2986 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2987 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2988 should have been "&", but it just so happened that all the cases this let
2989 through by mistake were picked up later in the function.
2990
2991 6. I had used a variable called "isblank" - this is a C99 function, causing
2992 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2993
2994 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2995 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2996 so the version is shown; (c) stop "make test" echoing "./RunTest".
2997
2998 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2999
3000 9. The prototype for memmove() for systems that don't have it was using
3001 size_t, but the inclusion of the header that defines size_t was later. I've
3002 moved the #includes for the C headers earlier to avoid this.
3003
3004 10. Added some adjustments to the code to make it easier to compiler on certain
3005 special systems:
3006
3007 (a) Some "const" qualifiers were missing.
3008 (b) Added the macro EXPORT before all exported functions; by default this
3009 is defined to be empty.
3010 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3011 that it reads its output file name as an argument instead of writing
3012 to the standard output and assuming this can be redirected.
3013
3014 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3015 class containing characters with values greater than 255, PCRE compilation
3016 went into a loop.
3017
3018 12. A recursive reference to a subpattern that was within another subpattern
3019 that had a minimum quantifier of zero caused PCRE to crash. For example,
3020 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3021 recursion. If the recursively-called subpattern itself had a zero repeat,
3022 that was OK.
3023
3024 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3025 buffer into which it was copied (for escape processing) was still set at
3026 1024, so long lines caused crashes.
3027
3028 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3029 "internal error: code overflow...". This applied to any character class
3030 that was followed by a possessive quantifier.
3031
3032 15. Modified the Makefile to add libpcre.la as a prerequisite for
3033 libpcreposix.la because I was told this is needed for a parallel build to
3034 work.
3035
3036 16. If a pattern that contained .* following optional items at the start was
3037 studied, the wrong optimizing data was generated, leading to matching
3038 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3039 matching string must start with a or b or c. The correct conclusion for
3040 this pattern is that a match can start with any character.
3041
3042
3043 Version 4.4 13-Aug-03
3044 ---------------------
3045
3046 1. In UTF-8 mode, a character class containing characters with values between
3047 127 and 255 was not handled correctly if the compiled pattern was studied.
3048 In fixing this, I have also improved the studying algorithm for such
3049 classes (slightly).
3050
3051 2. Three internal functions had redundant arguments passed to them. Removal
3052 might give a very teeny performance improvement.
3053
3054 3. Documentation bug: the value of the capture_top field in a callout is *one
3055 more than* the number of the hightest numbered captured substring.
3056
3057 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3058 in incorrectly linking with a previously installed version. They now link
3059 explicitly with libpcre.la.
3060
3061 5. configure.in no longer needs to recognize Cygwin specially.
3062
3063 6. A problem in pcre.in for Windows platforms is fixed.
3064
3065 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3066 pcretest, it used to include the size of the study block as part of its
3067 output. Unfortunately, the structure contains a field that has a different
3068 size on different hardware architectures. This meant that the tests that
3069 showed this size failed. As the block is currently always of a fixed size,
3070 this information isn't actually particularly useful in pcretest output, so
3071 I have just removed it.
3072
3073 8. Three pre-processor statements accidentally did not start in column 1.
3074 Sadly, there are *still* compilers around that complain, even though
3075 standard C has not required this for well over a decade. Sigh.
3076
3077 9. In pcretest, the code for checking callouts passed small integers in the
3078 callout_data field, which is a void * field. However, some picky compilers
3079 complained about the casts involved for this on 64-bit systems. Now
3080 pcretest passes the address of the small integer instead, which should get
3081 rid of the warnings.
3082
3083 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3084 both compile and run time, and gives an error if an invalid UTF-8 sequence
3085 is found. There is a option for disabling this check in cases where the
3086 string is known to be correct and/or the maximum performance is wanted.
3087
3088 11. In response to a bug report, I changed one line in Makefile.in from
3089
3090 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3091 to
3092 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3093
3094 to look similar to other lines, but I have no way of telling whether this
3095 is the right thing to do, as I do not use Windows. No doubt I'll get told
3096 if it's wrong...
3097
3098
3099 Version 4.3 21-May-03
3100 ---------------------
3101
3102 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3103 Makefile.
3104
3105 2. Some refactoring to improve the quality of the code:
3106
3107 (i) The utf8_table... variables are now declared "const".
3108
3109 (ii) The code for \cx, which used the "case flipping" table to upper case
3110 lower case letters, now just substracts 32. This is ASCII-specific,
3111 but the whole concept of \cx is ASCII-specific, so it seems
3112 reasonable.
3113
3114 (iii) PCRE was using its character types table to recognize decimal and
3115 hexadecimal digits in the pattern. This is silly, because it handles
3116 only 0-9, a-f, and A-F, but the character types table is locale-
3117 specific, which means strange things might happen. A private
3118 table is now used for this - though it costs 256 bytes, a table is
3119 much faster than multiple explicit tests. Of course, the standard
3120 character types table is still used for matching digits in subject
3121 strings against \d.
3122
3123 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3124 ending in _t are). So I've renamed it as ESC_tee.
3125
3126 3. The first argument for regexec() in the POSIX wrapper should have been
3127 defined as "const".
3128
3129 4. Changed pcretest to use malloc() for its buffers so that they can be
3130 Electric Fenced for debugging.
3131
3132 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3133 to read one or more bytes before the start of the subject string. Often this
3134 had no effect on PCRE's behaviour, but in some circumstances it could
3135 provoke a segmentation fault.
3136
3137 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3138 to try to read one or more bytes before the start of the subject string.
3139
3140 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3141 UTF-8 support could misbehave in various ways if the subject string
3142 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3143 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3144 back over UTF-8 characters.)
3145
3146
3147 Version 4.2 14-Apr-03
3148 ---------------------
3149
3150 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3151
3152 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3153 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3154 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3155 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3156 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3157 and BUILD_EXEEXT
3158 Note: automatic setting of the BUILD variables is not yet working
3159 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3160 compile-time but not at link-time
3161 [LINK]: use for linking executables only
3162 make different versions for Windows and non-Windows
3163 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3164 libraries
3165 [LINK_FOR_BUILD]: new variable
3166 [OBJEXT]: use throughout
3167 [EXEEXT]: use throughout
3168 <winshared>: new target
3169 <wininstall>: new target
3170 <dftables.o>: use native compiler
3171 <dftables>: use native linker
3172 <install>: handle Windows platform correctly
3173 <clean>: ditto
3174 <check>: ditto
3175 copy DLL to top builddir before testing
3176
3177 As part of these changes, -no-undefined was removed again. This was reported
3178 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3179 in any case.
3180
3181 3. Some tidies to get rid of compiler warnings:
3182
3183 . In the match_data structure, match_limit was an unsigned long int, whereas
3184 match_call_count was an int. I've made them both unsigned long ints.
3185
3186 . In pcretest the fact that a const uschar * doesn't automatically cast to
3187 a void * provoked a warning.
3188
3189 . Turning on some more compiler warnings threw up some "shadow" variables
3190 and a few more missing casts.
3191
3192 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3193 option, a class that contained a single character with a value between 128
3194 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3195
3196 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3197 option, a class that contained several characters, but with at least one
3198 whose value was between 128 and 255 caused PCRE to crash.
3199
3200
3201 Version 4.1 12-Mar-03
3202 ---------------------
3203
3204 1. Compiling with gcc -pedantic found a couple of places where casts were
3205 needed, and a string in dftables.c that was longer than standard compilers are
3206 required to support.
3207
3208 2. Compiling with Sun's compiler found a few more places where the code could
3209 be tidied up in order to avoid warnings.
3210
3211 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3212 first of these names is deprecated in the latest Autoconf in favour of the name
3213 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3214 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3215 analogy I have changed it to CFLAGS_FOR_BUILD.
3216
3217 4. Added -no-undefined to the linking command in the Makefile, because this is
3218 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3219 linking step for the pcreposix library.
3220
3221 5. PCRE was failing to diagnose the case of two named groups with the same
3222 name.
3223
3224 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3225 literal character that is needed in the subject for a match, and scans along to
3226 ensure that it is present before embarking on the full matching process. This
3227 saves time in cases of nested unlimited repeats that are never going to match.
3228 Problem: the scan can take a lot of time if the subject is very long (e.g.
3229 megabytes), thus penalizing straightforward matches. It is now done only if the
3230 amount of subject to be scanned is less than 1000 bytes.
3231
3232 7. A lesser problem with the same optimization is that it was recording the
3233 first character of an anchored pattern as "needed", thus provoking a search
3234 right along the subject, even when the first match of the pattern was going to
3235 fail. The "needed" character is now not set for anchored patterns, unless it
3236 follows something in the pattern that is of non-fixed length. Thus, it still
3237 fulfils its original purpose of finding quick non-matches in cases of nested
3238 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3239
3240
3241 Version 4.0 17-Feb-03
3242 ---------------------
3243
3244 1. If a comment in an extended regex that started immediately after a meta-item
3245 extended to the end of string, PCRE compiled incorrect data. This could lead to
3246 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3247
3248 2. Moved to autoconf 2.53 and libtool 1.4.2.
3249
3250 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3251 the special perltest8 script is no longer needed - all the tests can be run
3252 from a single perltest script.
3253
3254 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3255 by \s. It has now been removed in PCRE. This means it isn't recognized as
3256 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3257 class [:space:] *does* include VT, thereby creating a mess.
3258
3259 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3260 space and tab.
3261
3262 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3263 its new features into the main test script, reducing the number of scripts.
3264
3265 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3266 were backward compatible, and made the (?i) apply to the whole pattern, as if
3267 /i were given. Now it behaves more logically, and applies the option setting
3268 only to what follows. PCRE has been changed to follow suit. However, if it
3269 finds options settings right at the start of the pattern, it extracts them into
3270 the global options, as before. Thus, they show up in the info data.
3271
3272 8. Added support for the \Q...\E escape sequence. Characters in between are
3273 treated as literals. This is slightly different from Perl in that $ and @ are
3274 also handled as literals inside the quotes. In Perl, they will cause variable
3275 interpolation. Note the following examples:
3276
3277 Pattern PCRE matches Perl matches
3278
3279 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3280 \Qabc\$xyz\E abc\$xyz abc\$xyz
3281 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3282
3283 For compatibility with Perl, \Q...\E sequences are recognized inside character
3284 classes as well as outside them.
3285
3286 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3287 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3288 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3289 signed/unsigned warnings.
3290
3291 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3292 option for pcretest, so I've replaced it by a simple function that does just
3293 that job.
3294
3295 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3296 "pcregrep -".
3297
3298 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3299 Java package. This provides some syntactic sugar for simple cases of what my
3300 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3301 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3302 item, you can use this simplified notation. Note that only makes sense with
3303 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3304 greediness, whatever the setting of the PCRE_UNGREEDY option.
3305
3306 13. A change of greediness default within a pattern was not taking effect at
3307 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3308 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3309 was abstracted outside.
3310
3311 14. PCRE now supports the \G assertion. It is true when the current matching
3312 position is at the start point of the match. This differs from \A when the
3313 starting offset is non-zero. Used with the /g option of pcretest (or similar
3314 code), it works in the same way as it does for Perl's /g option. If all
3315 alternatives of a regex begin with \G, the expression is anchored to the start
3316 match position, and the "anchored" flag is set in the compiled expression.
3317
3318 15. Some bugs concerning the handling of certain option changes within patterns
3319 have been fixed. These applied to options other than (?ims). For example,
3320 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3321 been the other way round. Some of this was related to change 7 above.
3322
3323 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3324 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3325 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3326 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3327
3328 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3329 mode. Unlike ".", it always matches newline, whatever the setting of
3330 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3331 assertions. Perl allows it, but it doesn't (in general) work because it can't
3332 calculate the length of the lookbehind. At least, that's the case for Perl
3333 5.8.0 - I've been told they are going to document that it doesn't work in
3334 future.
3335
3336 18. Added an error diagnosis for escapes that PCRE does not support: these are
3337 \L, \l, \N, \P, \p, \U, \u, and \X.
3338
3339 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3340 reading past the end of the pattern in cases such as /[abcd/.
3341
3342 20. PCRE was getting more memory than necessary for patterns with classes that
3343 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3344
3345 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3346 compiling PCRE for use with Virtual Pascal.
3347
3348 22. Small fix to the Makefile to make it work properly if the build is done
3349 outside the source tree.
3350
3351 23. Added a new extension: a condition to go with recursion. If a conditional
3352 subpattern starts with (?(R) the "true" branch is used if recursion has
3353 happened, whereas the "false" branch is used only at the top level.
3354
3355 24. When there was a very long string of literal characters (over 255 bytes
3356 without UTF support, over 250 bytes with UTF support), the computation of how
3357 much memory was required could be incorrect, leading to segfaults or other
3358 strange effects.
3359
3360 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3361 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3362 there was a subsequent back reference to those brackets. This meant that, for
3363 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3364 possible to check for precisely this case. All we can do is abandon the
3365 optimization if .* occurs inside capturing brackets when there are any back
3366 references whatsoever. (See below for a better fix that came later.)
3367
3368 26. The handling of the optimization for finding the first character of a
3369 non-anchored pattern, and for finding a character that is required later in the
3370 match were failing in some cases. This didn't break the matching; it just
3371 failed to optimize when it could. The way this is done has been re-implemented.
3372
3373 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3374
3375 28. Added a new feature that provides some of the functionality that Perl
3376 provides with (?{...}). The facility is termed a "callout". The way it is done
3377 in PCRE is for the caller to provide an optional function, by setting
3378 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3379 global variable. By default it is unset, which disables all calling out. To get
3380 the function called, the regex must include (?C) at appropriate points. This
3381 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3382 This provides a means of identifying different callout points. When PCRE
3383 reaches such a point in the regex, if pcre_callout has been set, the external
3384 function is called. It is provided with data in a structure called
3385 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3386 matching continues; if it returns a non-zero value, the match at the current
3387 point fails. However, backtracking will occur if possible. [This was changed
3388 later and other features added - see item 49 below.]
3389
3390 29. pcretest is upgraded to test the callout functionality. It provides a
3391 callout function that displays information. By default, it shows the start of
3392 the match and the current position in the text. There are some new data escapes
3393 to vary what happens:
3394
3395 \C+ in addition, show current contents of captured substrings
3396 \C- do not supply a callout function
3397 \C!n return 1 when callout number n is reached
3398 \C!n!m return 1 when callout number n is reached for the mth time
3399
3400 30. If pcregrep was called with the -l option and just a single file name, it
3401 output "<stdin>" if a match was found, instead of the file name.
3402
3403 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3404 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3405 pcre_exec(). This saves a malloc/free per call. The default value of
3406 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3407 when configuring.
3408
3409 32. The default maximum size of a compiled pattern is 64K. There have been a
3410 few cases of people hitting this limit. The code now uses macros to handle the
3411 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3412 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3413 configuring. Tests 2 and 5 work only with 2-byte links because they output
3414 debugging information about compiled patterns.
3415
3416 33. Internal code re-arrangements:
3417
3418 (a) Moved the debugging function for printing out a compiled regex into
3419 its own source file (printint.c) and used #include to pull it into
3420 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3421 separate copies.
3422
3423 (b) Defined the list of op-code names for debugging as a macro in
3424 internal.h so that it is next to the definition of the opcodes.
3425
3426 (c) Defined a table of op-code lengths for simpler skipping along compiled
3427 code. This is again a macro in internal.h so that it is next to the
3428 definition of the opcodes.
3429
3430 34. Added support for recursive calls to individual subpatterns, along the
3431 lines of Robin Houston's patch (but implemented somewhat differently).
3432
3433 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3434 allow it to read and process whole directories in Win32. This code was
3435 contributed by Lionel Fourquaux; it has not been tested by me.
3436
3437 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3438 used to name a group. Names consist of alphanumerics and underscores, and must
3439 be unique. Back references use the syntax (?P=name) and recursive calls use
3440 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3441 numbers. The function pcre_fullinfo() can be used after compilation to extract
3442 a name/number map. There are three relevant calls:
3443
3444 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3445 PCRE_INFO_NAMECOUNT yields the number of entries
3446 PCRE_INFO_NAMETABLE yields a pointer to the map.
3447
3448 The map is a vector of fixed-size entries. The size of each entry depends on
3449 the length of the longest name used. The first two bytes of each entry are the
3450 group number, most significant byte first. There follows the corresponding
3451 name, zero terminated. The names are in alphabetical order.
3452
3453 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3454 case instead of 255. Making it the same both with and without UTF-8 support
3455 means that the same test output works with both.
3456
3457 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3458 calling malloc() with a zero argument.
3459
3460 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3461 optimization. I've improved things by keeping a bitmap of backreferences with
3462 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3463 fact referenced, the optimization can be applied. It is unlikely that a
3464 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3465 the match to follow \n) will appear inside brackets with a number greater than
3466 31, but if it does, any back reference > 31 suppresses the optimization.
3467
3468 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3469 of disabling numbered capturing parentheses. Any opening parenthesis that is
3470 not followed by ? behaves as if it were followed by ?: but named parentheses
3471 can still be used for capturing (and they will acquire numbers in the usual
3472 way).
3473
3474 41. Redesigned the return codes from the match() function into yes/no/error so
3475 that errors can be passed back from deep inside the nested calls. A malloc
3476 failure while inside a recursive subpattern call now causes the
3477 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3478
3479 42. It is now possible to set a limit on the number of times the match()
3480 function is called in a call to pcre_exec(). This facility makes it possible to
3481 limit the amount of recursion and backtracking, though not in a directly
3482 obvious way, because the match() function is used in a number of different
3483 circumstances. The count starts from zero for each position in the subject
3484 string (for non-anchored patterns). The default limit is, for compatibility, a
3485 large number, namely 10 000 000. You can change this in two ways:
3486
3487 (a) When configuring PCRE before making, you can use --with-match-limit=n
3488 to set a default value for the compiled library.
3489
3490 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3491 a different value is set. See 45 below.
3492
3493 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3494
3495 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3496 of things that can be changed at compile time. The first argument specifies
3497 what is wanted and the second points to where the information is to be placed.
3498 The current list of available information is:
3499
3500 PCRE_CONFIG_UTF8
3501
3502 The output is an integer that is set to one if UTF-8 support is available;
3503 otherwise it is set to zero.
3504
3505 PCRE_CONFIG_NEWLINE
3506
3507 The output is an integer that it set to the value of the code that is used for
3508 newline. It is either LF (10) or CR (13).
3509
3510 PCRE_CONFIG_LINK_SIZE
3511
3512 The output is an integer that contains the number of bytes used for internal
3513 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3514
3515 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3516
3517 The output is an integer that contains the threshold above which the POSIX
3518 interface uses malloc() for output vectors. See item 31 above.
3519
3520 PCRE_CONFIG_MATCH_LIMIT
3521
3522 The output is an unsigned integer that contains the default limit of the number
3523 of match() calls in a pcre_exec() execution. See 42 above.
3524
3525 44. pcretest has been upgraded by the addition of the -C option. This causes it
3526 to extract all the available output from the new pcre_config() function, and to
3527 output it. The program then exits immediately.
3528
3529 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3530 order to support additional features. One way would have been to define
3531 pcre_exec2() (for example) with extra arguments, but this would not have been
3532 extensible, and would also have required all calls to the original function to
3533 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3534 is used for passing in "extra" data from pcre_study().
3535
3536 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3537 contains the following fields:
3538
3539 flags a bitmap indicating which of the following fields are set
3540 study_data opaque data from pcre_study()
3541 match_limit a way of specifying a limit on match() calls for a specific
3542 call to pcre_exec()
3543 callout_data data for callouts (see 49 below)
3544
3545 The flag bits are also defined in pcre.h, and are
3546
3547 PCRE_EXTRA_STUDY_DATA
3548 PCRE_EXTRA_MATCH_LIMIT
3549 PCRE_EXTRA_CALLOUT_DATA
3550
3551 The pcre_study() function now returns one of these new pcre_extra blocks, with
3552 the actual study data pointed to by the study_data field, and the
3553 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3554 before. That is, this change is entirely upwards-compatible and requires no
3555 change to existing code.
3556
3557 If you want to pass in additional data to pcre_exec(), you can either place it
3558 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3559 block.
3560
3561 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3562 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3563 times with different match limits, until it finds the minimum value needed for
3564 pcre_exec() to complete. The value is then output. This can be instructive; for
3565 most simple matches the number is quite small, but for pathological cases it
3566 gets very large very quickly.
3567
3568 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3569 returns the size of the data block pointed to by the study_data field in a
3570 pcre_extra block, that is, the value that was passed as the argument to
3571 pcre_malloc() when PCRE was getting memory in which to place the information
3572 created by pcre_study(). The fourth argument should point to a size_t variable.
3573 pcretest has been extended so that this information is shown after a successful
3574 pcre_study() call when information about the compiled regex is being displayed.
3575
3576 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3577 because what follows is always an absolute path. (Later: it turns out that this
3578 is more than cosmetic for MinGW, because it doesn't like empty path
3579 components.)
3580
3581 49. Some changes have been made to the callout feature (see 28 above):
3582
3583 (i) A callout function now has three choices for what it returns:
3584
3585 0 => success, carry on matching
3586 > 0 => failure at this point, but backtrack if possible
3587 < 0 => serious error, return this value from pcre_exec()
3588
3589 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3590 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3591 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3592 use by callout functions. It will never be used by PCRE itself.
3593
3594 (ii) The pcre_extra structure (see 45 above) has a void * field called
3595 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3596 pcre_callout_block structure has a field of the same name. The contents of
3597 the field passed in the pcre_extra structure are passed to the callout
3598 function in the corresponding field in the callout block. This makes it
3599 easier to use the same callout-containing regex from multiple threads. For
3600 testing, the pcretest program has a new data escape
3601
3602 \C*n pass the number n (may be negative) as callout_data
3603
3604 If the callout function in pcretest receives a non-zero value as
3605 callout_data, it returns that value.
3606
3607 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3608 there were some redundant $(CFLAGS) in commands that are now specified as
3609 $(LINK), which already includes $(CFLAGS).
3610
3611 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3612 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3613 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3614 one-byte characters throughout. Note that case-insensitive matching applies
3615 only to characters whose values are less than 256. PCRE doesn't support the
3616 notion of cases for higher-valued characters.
3617
3618 (i) A character class whose characters are all within 0-255 is handled as
3619 a bit map, and the map is inverted for negative classes. Previously, a
3620 character > 255 always failed to match such a class; however it should
3621 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3622
3623 (ii) A negated character class with a single character < 255 is coded as
3624 "not this character" (OP_NOT). This wasn't working properly when the test
3625 character was multibyte, either singly or repeated.
3626
3627 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3628 mode, for example: \x{100}{2,3}.
3629
3630 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3631 singly or repeated) now correctly test multibyte characters. However,
3632 PCRE doesn't recognize any characters with values greater than 255 as
3633 digits, spaces, or word characters. Such characters always match \D, \S,
3634 and \W, and never match \d, \s, or \w.
3635
3636 (v) Classes may now contain characters and character ranges with values
3637 greater than 255. For example: [ab\x{100}-\x{400}].
3638
3639 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3640 PCRE in UTF-8 mode.
3641
3642 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3643 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3644 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3645 value.)
3646
3647 53. The single man page has become too large. I have therefore split it up into
3648 a number of separate man pages. These also give rise to individual HTML pages;
3649 these are now put in a separate directory, and there is an index.html page that
3650 lists them all. Some hyperlinking between the pages has been installed.
3651
3652 54. Added convenience functions for handling named capturing parentheses.
3653
3654 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3655 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3656 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3657 are faulted.
3658
3659 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3660 calling configure. These values are used when compiling the dftables.c program
3661 which is run to generate the source of the default character tables. They
3662 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3663 you will need to set these values.
3664
3665 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3666
3667
3668 Version 3.9 02-Jan-02
3669 ---------------------
3670
3671 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3672
3673 2. If --disable-static was given, the building process failed when trying to
3674 build pcretest and pcregrep. (For some reason it was using libtool to compile
3675 them, which is not right, as they aren't part of the library.)
3676
3677
3678 Version 3.8 18-Dec-01
3679 ---------------------
3680
3681 1. The experimental UTF-8 code was completely screwed up. It was packing the
3682 bytes in the wrong order. How dumb can you get?
3683
3684
3685 Version 3.7 29-Oct-01
3686 ---------------------
3687
3688 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3689 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3690 this didn't happen under Solaris 8, where I normally test things.
3691
3692 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3693 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3694 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3695 this fix an hour or so after the initial 3.7 release.)
3696
3697
3698 Version 3.6 23-Oct-01
3699 ---------------------
3700
3701 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3702 offsets passed as NULL with zero offset count.
3703
3704 2. The config.guess and config.sub files had not been updated when I moved to
3705 the latest autoconf.
3706
3707
3708 Version 3.5 15-Aug-01
3709 ---------------------
3710
3711 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3712 had been forgotten.
3713
3714 2. By using declared but undefined structures, we can avoid using "void"
3715 definitions in pcre.h while keeping the internal definitions of the structures
3716 private.
3717
3718 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3719 user point of view, this means that both static and shared libraries are built
3720 by default, but this can be individually controlled. More of the work of
3721 handling this static/shared cases is now inside libtool instead of PCRE's make
3722 file.
3723
3724 4. The pcretest utility is now installed along with pcregrep because it is
3725 useful for users (to test regexs) and by doing this, it automatically gets
3726 relinked by libtool. The documentation has been turned into a man page, so
3727 there are now .1, .txt, and .html versions in /doc.
3728
3729 5. Upgrades to pcregrep:
3730 (i) Added long-form option names like gnu grep.
3731 (ii) Added --help to list all options with an explanatory phrase.
3732 (iii) Added -r, --recursive to recurse into sub-directories.
3733 (iv) Added -f, --file to read patterns from a file.
3734
3735 6. pcre_exec() was referring to its "code" argument before testing that
3736 argument for NULL (and giving an error if it was NULL).
3737
3738 7. Upgraded Makefile.in to allow for compiling in a different directory from
3739 the source directory.
3740
3741 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3742 options bits, the pointer it was passed was to an int instead of to an unsigned
3743 long int. This mattered only on 64-bit systems.
3744
3745 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3746 generated) instead of pcre.in, which it its source. Also made the same change
3747 in several of the .c files.
3748
3749 10. A new release of gcc defines printf() as a macro, which broke pcretest
3750 because it had an ifdef in the middle of a string argument for printf(). Fixed
3751 by using separate calls to printf().
3752
3753 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3754 script, to force use of CR or LF instead of \n in the source. On non-Unix
3755 systems, the value can be set in config.h.
3756
3757 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3758 absolute limit. Changed the text of the error message to make this clear, and
3759 likewise updated the man page.
3760
3761 13. The limit of 99 on the number of capturing subpatterns has been removed.
3762 The new limit is 65535, which I hope will not be a "real" limit.
3763
3764
3765 Version 3.4 22-Aug-00
3766 ---------------------
3767
3768 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3769
3770 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3771
3772
3773 Version 3.3 01-Aug-00
3774 ---------------------
3775
3776 1. If an octal character was given, but the value was greater than \377, it
3777 was not getting masked to the least significant bits, as documented. This could
3778 lead to crashes in some systems.
3779
3780 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3781 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3782
3783 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3784 These just pass their arguments on to (pcre_free)(), but they are provided
3785 because some uses of PCRE bind it to non-C systems that can call its functions,
3786 but cannot call free() or pcre_free() directly.
3787
3788 4. Add "make test" as a synonym for "make check". Corrected some comments in
3789 the Makefile.
3790
3791 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3792 Makefile.
3793
3794 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3795 command called pgrep for grepping around the active processes.
3796
3797 7. Added the beginnings of support for UTF-8 character strings.
3798
3799 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3800 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3801 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3802 out for the ar command.)
3803
3804
3805 Version 3.2 12-May-00
3806 ---------------------
3807
3808 This is purely a bug fixing release.
3809
3810 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3811 of ZA. This was just one example of several cases that could provoke this bug,
3812 which was introduced by change 9 of version 2.00. The code for breaking
3813 infinite loops after an iteration that matches an empty string was't working
3814 correctly.
3815
3816 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3817 when matched against abbab (for example). After matching an empty string, it
3818 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3819 caused it to match further down the string than it should.
3820
3821 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3822 was there because it doesn't seem to be needed, and it causes trouble on some
3823 systems, as it is not a Standard C header. It has been removed.
3824
3825 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3826 were reported on the Macintosh. The changes were from
3827
3828 while ((c = *(++ptr)) != 0 && c != '\n');
3829 to
3830 while ((c = *(++ptr)) != 0 && c != '\n') ;
3831
3832 Totally extraordinary, but if that's what it takes...
3833
3834 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3835 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3836 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3837 assumes the way PCRE uses memmove() (always moving upwards).
3838
3839 6. PCRE is being used in one environment where strchr() is not available. There
3840 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3841 faster code anyway.
3842
3843
3844 Version 3.1 09-Feb-00
3845 ---------------------
3846
3847 The only change in this release is the fixing of some bugs in Makefile.in for
3848 the "install" target:
3849
3850 (1) It was failing to install pcreposix.h.
3851
3852 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3853
3854
3855 Version 3.0 01-Feb-00
3856 ---------------------
3857
3858 1. Add support for the /+ modifier to perltest (to output $` like it does in
3859 pcretest).
3860
3861 2. Add support for the /g modifier to perltest.
3862
3863 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3864 matches null strings.
3865
3866 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3867 pattern. Perl treats empty patterns specially - it reuses the most recent
3868 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3869 effect.
3870
3871 5. The POSIX interface was broken in that it was just handing over the POSIX
3872 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3873 required a bigger vector, with some working space on the end. This means that
3874 the POSIX wrapper now has to get and free some memory, and copy the results.
3875
3876 6. Added some simple autoconf support, placing the test data and the
3877 documentation in separate directories, re-organizing some of the
3878 information files, and making it build pcre-config (a GNU standard). Also added
3879 libtool support for building PCRE as a shared library, which is now the
3880 default.
3881
3882 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3883 09 are not valid octal constants. Single digits will be used for minor values
3884 less than 10.
3885
3886 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3887 existing programs that set these in the POSIX interface can use PCRE without
3888 modification.
3889
3890 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3891 return all that pcre_info() returns, plus additional data. The pcre_info()
3892 function is retained for compatibility, but is considered to be obsolete.
3893
3894 10. Added experimental recursion feature (?R) to handle one common case that
3895 Perl 5.6 will be able to do with (?p{...}).
3896
3897 11. Added support for POSIX character classes like [:alpha:], which Perl is
3898 adopting.
3899
3900
3901 Version 2.08 31-Aug-99
3902 ----------------------
3903
3904 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3905 trying to match at the startoffset position, but instead was moving forward to
3906 the next newline as if a previous match had failed.
3907
3908 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3909 and could get into a loop if a null string was matched other than at the start
3910 of the subject.
3911
3912 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3913 be distinguished at compile time, and for completeness also added PCRE_DATE.
3914
3915 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3916 in GnuWin32 environments.
3917
3918
3919 Version 2.07 29-Jul-99
3920 ----------------------
3921
3922 1. The documentation is now supplied in plain text form and HTML as well as in
3923 the form of man page sources.
3924
3925 2. C++ compilers don't like assigning (void *) values to other pointer types.
3926 In particular this affects malloc(). Although there is no problem in Standard
3927 C, I've put in casts to keep C++ compilers happy.
3928
3929 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3930 should be (const char *).
3931
3932 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3933 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3934 However, I haven't made this a standard facility. The documentation doesn't
3935 mention it, and the Makefile doesn't support it.
3936
3937 5. The Makefile now contains an "install" target, with editable destinations at
3938 the top of the file. The pcretest program is not installed.
3939
3940 6. pgrep -V now gives the PCRE version number and date.
3941
3942 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3943 causing the entire string to be ignored, instead of just the last character.
3944
3945 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3946 non-matching string, it can take a very, very long time, even for strings of
3947 quite modest length, because of the nested recursion. PCRE now does better in
3948 some of these cases. It does this by remembering the last required literal
3949 character in the pattern, and pre-searching the subject to ensure it is present
3950 before running the real match. In other words, it applies a heuristic to detect
3951 some types of certain failure quickly, and in the above example, if presented
3952 with a string that has no trailing " it gives "no match" very quickly.
3953
3954 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3955 other alternatives are tried instead.
3956
3957
3958 Version 2.06 09-Jun-99
3959 ----------------------
3960
3961 1. Change pcretest's output for amount of store used to show just the code
3962 space, because the remainder (the data block) varies in size between 32-bit and
3963 64-bit systems.
3964
3965 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3966 start matching at. This allows lookbehinds to work when searching for multiple
3967 occurrences in a string.
3968
3969 3. Added additional options to pcretest for testing multiple occurrences:
3970
3971 /+ outputs the rest of the string that follows a match
3972 /g loops for multiple occurrences, using the new startoffset argument
3973 /G loops for multiple occurrences by passing an incremented pointer
3974
3975 4. PCRE wasn't doing the "first character" optimization for patterns starting
3976 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3977 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3978 the letter 'x'. On long subject strings, this gives a significant speed-up.
3979
3980
3981 Version 2.05 21-Apr-99
3982 ----------------------
3983
3984 1. Changed the type of magic_number from int to long int so that it works
3985 properly on 16-bit systems.
3986
3987 2. Fixed a bug which caused patterns starting with .* not to work correctly
3988 when the subject string contained newline characters. PCRE was assuming
3989 anchoring for such patterns in all cases, which is not correct because .* will
3990 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3991 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3992 must be retried after every newline in the subject.
3993
3994
3995 Version 2.04 18-Feb-99
3996 ----------------------
3997
3998 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3999 computation of the store needed to hold the pattern was incorrect (too large).
4000 If such patterns were nested a few deep, this could multiply and become a real
4001 problem.
4002
4003 2. Added /M option to pcretest to show the memory requirement of a specific
4004 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4005
4006 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4007 compiled in such a way that the backtracking after subsequent failure was
4008 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4009 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4010
4011
4012 Version 2.03 02-Feb-99
4013 ----------------------
4014
4015 1. Fixed typo and small mistake in man page.
4016
4017 2. Added 4th condition (GPL supersedes if conflict) and created separate
4018 LICENCE file containing the conditions.
4019
4020 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4021 Perl, that is the internal \ allows the delimiter to be included in the
4022 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4023 the final delimiter, add \ to the end of the pattern (to test the error).
4024
4025 4. Added the convenience functions for extracting substrings after a successful
4026 match. Updated pcretest to make it able to test these functions.
4027
4028
4029 Version 2.02 14-Jan-99
4030 ----------------------
4031
4032 1. Initialized the working variables associated with each extraction so that
4033 their saving and restoring doesn't refer to uninitialized store.
4034
4035 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4036 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4037 fix the problem.
4038
4039 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4040 calls, and wasn't printing the correct value for compiling calls. Increased the
4041 default value of LOOPREPEAT, and the number of significant figures in the
4042 times.
4043
4044 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4045
4046 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4047 a building problem on Windows NT with a FAT file system.
4048
4049
4050 Version 2.01 21-Oct-98
4051 ----------------------
4052
4053 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4054 to character tables built by pcre_maketables() in the current locale. If NULL
4055 is passed, the default tables are used.
4056
4057
4058 Version 2.00 24-Sep-98
4059 ----------------------
4060
4061 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4062 it any more.
4063
4064 2. Allow quantification of (?>) groups, and make it work correctly.
4065
4066 3. The first character computation wasn't working for (?>) groups.
4067
4068 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4069 end of the subject) and add 5.005's \z, which really does match only at the
4070 very end of the subject.
4071
4072 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4073
4074 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4075 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4076 localized options. All options to pcre_study() were also removed.
4077
4078 7. Add other new features from 5.005:
4079
4080 $(?<= positive lookbehind
4081 $(?<! negative lookbehind
4082 (?imsx-imsx) added the unsetting capability
4083 such a setting is global if at outer level; local otherwise
4084 (?imsx-imsx:) non-capturing groups with option setting
4085 (?(cond)re|re) conditional pattern matching
4086
4087 A backreference to itself in a repeated group matches the previous
4088 captured string.
4089
4090 8. General tidying up of studying (both automatic and via "study")
4091 consequential on the addition of new assertions.
4092
4093 9. As in 5.005, unlimited repeated groups that could match an empty substring
4094 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4095 runtime if any iteration does actually match an empty substring.
4096
4097 10. Include the RunTest script in the distribution.
4098
4099 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4100 discrepancies, some of which were old and were also with respect to 5.004. They
4101 have now been fixed.
4102
4103
4104 Version 1.09 28-Apr-98
4105 ----------------------
4106
4107 1. A negated single character class followed by a quantifier with a minimum
4108 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4109 program crashes, or just wrong answers. This did not apply to negated classes
4110 containing more than one character, or to minima other than one.
4111
4112
4113 Version 1.08 27-Mar-98
4114 ----------------------
4115
4116 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4117
4118 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4119 latter must appear before anything that relies on it in the pattern.
4120
4121
4122 Version 1.07 16-Feb-98
4123 ----------------------
4124
4125 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4126 repeat of a potentially empty string).
4127
4128
4129 Version 1.06 23-Jan-98
4130 ----------------------
4131
4132 1. Added Markus Oberhumer's little patches for C++.
4133
4134 2. Literal strings longer than 255 characters were broken.
4135
4136
4137 Version 1.05 23-Dec-97
4138 ----------------------
4139
4140 1. Negated character classes containing more than one character were failing if
4141 PCRE_CASELESS was set at run time.
4142
4143
4144 Version 1.04 19-Dec-97
4145 ----------------------
4146
4147 1. Corrected the man page, where some "const" qualifiers had been omitted.
4148
4149 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4150 input syntax.
4151
4152 3. Fixed memory leak which occurred when a regex with back references was
4153 matched with an offsets vector that wasn't big enough. The temporary memory
4154 that is used in this case wasn't being freed if the match failed.
4155
4156 4. Tidied pcretest to ensure it frees memory that it gets.
4157
4158 5. Temporary memory was being obtained in the case where the passed offsets
4159 vector was exactly big enough.
4160
4161 6. Corrected definition of offsetof() from change 5 below.
4162
4163 7. I had screwed up change 6 below and broken the rules for the use of
4164 setjmp(). Now fixed.
4165
4166
4167 Version 1.03 18-Dec-97
4168 ----------------------
4169
4170 1. A erroneous regex with a missing opening parenthesis was correctly
4171 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4172 on some systems.
4173
4174 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4175 it was reported that one broken compiler failed on the former because "code" is
4176 also an independent variable.
4177
4178 3. The erroneous regex a[]b caused an array overrun reference.
4179
4180 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4181 fail on data ending with that character. (It was going on too far, and checking
4182 the next character, typically a binary zero.) This was specific to the
4183 optimized code for single-character negative classes.
4184
4185 5. Added a contributed patch from the TIN world which does the following:
4186
4187 + Add an undef for memmove, in case the the system defines a macro for it.
4188
4189 + Add a definition of offsetof(), in case there isn't one. (I don't know
4190 the reason behind this - offsetof() is part of the ANSI standard - but
4191 it does no harm).
4192
4193 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4194 most of the places where whitespace preceded '#'. I have given up and
4195 allowed the remaining 2 cases to be at the margin.
4196
4197 + Rename some variables in pcre to eliminate shadowing. This seems very
4198 pedantic, but does no harm, of course.
4199
4200 6. Moved the call to setjmp() into its own function, to get rid of warnings
4201 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4202
4203 7. Constructs such as \d{8,} were compiling into the equivalent of
4204 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4205 outcome, but in this particular case used more store than had been allocated,
4206 which caused the bug to be discovered because it threw up an internal error.
4207
4208 8. The debugging code in both pcre and pcretest for outputting the compiled
4209 form of a regex was going wrong in the case of back references followed by
4210 curly-bracketed repeats.
4211
4212
4213 Version 1.02 12-Dec-97
4214 ----------------------
4215
4216 1. Typos in pcre.3 and comments in the source fixed.
4217
4218 2. Applied a contributed patch to get rid of places where it used to remove
4219 'const' from variables, and fixed some signed/unsigned and uninitialized
4220 variable warnings.
4221
4222 3. Added the "runtest" target to Makefile.
4223
4224 4. Set default compiler flag to -O2 rather than just -O.
4225
4226
4227 Version 1.01 19-Nov-97
4228 ----------------------
4229
4230 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4231 like /([ab]*)*/, that is, for classes with more than one character in them.
4232
4233 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4234 as /((?>a*))*/ (a PCRE_EXTRA facility).
4235
4236
4237 Version 1.00 18-Nov-97
4238 ----------------------
4239
4240 1. Added compile-time macros to support systems such as SunOS4 which don't have
4241 memmove() or strerror() but have other things that can be used instead.
4242
4243 2. Arranged that "make clean" removes the executables.
4244
4245
4246 Version 0.99 27-Oct-97
4247 ----------------------
4248
4249 1. Fixed bug in code for optimizing classes with only one character. It was
4250 initializing a 32-byte map regardless, which could cause it to run off the end
4251 of the memory it had got.
4252
4253 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4254
4255
4256 Version 0.98 22-Oct-97
4257 ----------------------
4258
4259 1. Fixed bug in code for handling temporary memory usage when there are more
4260 back references than supplied space in the ovector. This could cause segfaults.
4261
4262
4263 Version 0.97 21-Oct-97
4264 ----------------------
4265
4266 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4267
4268 2. Optimized negated single characters not to use a bit map.
4269
4270 3. Brought error texts together as macro definitions; clarified some of them;
4271 fixed one that was wrong - it said "range out of order" when it meant "invalid
4272 escape sequence".
4273
4274 4. Changed some char * arguments to const char *.
4275
4276 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4277
4278 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4279 pcretest.
4280
4281
4282 Version 0.96 16-Oct-97
4283 ----------------------
4284
4285 1. Added a simple "pgrep" utility to the distribution.
4286
4287 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4288 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4289 where "ddd" means "one or more decimal digits".
4290
4291 3. Fixed serious bug. If a pattern had a back reference, but the call to
4292 pcre_exec() didn't supply a large enough ovector to record the related
4293 identifying subpattern, the match always failed. PCRE now remembers the number
4294 of the largest back reference, and gets some temporary memory in which to save
4295 the offsets during matching if necessary, in order to ensure that
4296 backreferences always work.
4297
4298 4. Increased the compatibility with Perl in a number of ways:
4299
4300 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4301 to request this handling. The option can be set at compile or exec time.
4302
4303 (b) $ matches before a terminating newline by default; an option
4304 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4305 mode). The option can be set at compile or exec time.
4306
4307 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4308 the same as Perl's. If the decimal number it represents is less than 10
4309 or there aren't that many previous left capturing parentheses, an octal
4310 escape is read. Inside a character class, it's always an octal escape,
4311 even if it is a single digit.
4312
4313 (d) An escaped but undefined alphabetic character is taken as a literal,
4314 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4315 escapes.
4316
4317 (e) {0} is now permitted. (The previous item is removed from the compiled
4318 pattern).
4319
4320 5. Changed all the names of code files so that the basic parts are no longer
4321 than 10 characters, and abolished the teeny "globals.c" file.
4322
4323 6. Changed the handling of character classes; they are now done with a 32-byte
4324 bit map always.
4325
4326 7. Added the -d and /D options to pcretest to make it possible to look at the
4327 internals of compilation without having to recompile pcre.
4328
4329
4330 Version 0.95 23-Sep-97
4331 ----------------------
4332
4333 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4334 \x20 at the start of a run of normal characters. These were being treated as
4335 real characters, instead of the source characters being re-checked.
4336
4337
4338 Version 0.94 18-Sep-97
4339 ----------------------
4340
4341 1. The functions are now thread-safe, with the caveat that the global variables
4342 containing pointers to malloc() and free() or alternative functions are the
4343 same for all threads.
4344
4345 2. Get pcre_study() to generate a bitmap of initial characters for non-
4346 anchored patterns when this is possible, and use it if passed to pcre_exec().
4347
4348
4349 Version 0.93 15-Sep-97
4350 ----------------------
4351
4352 1. /(b)|(:+)/ was computing an incorrect first character.
4353
4354 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4355 but not actually doing anything yet.
4356
4357 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4358 as Perl does (e.g. [-az] or [az-]).
4359
4360 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4361 all possible positions.
4362
4363 5. Split up into different modules to avoid including unneeded functions in a
4364 compiled binary. However, compile and exec are still in one module. The "study"
4365 function is split off.
4366
4367 6. The character tables are now in a separate module whose source is generated
4368 by an auxiliary program - but can then be edited by hand if required. There are
4369 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4370 toupper() in the code.
4371
4372 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4373 make them global. Abolish the function for setting them, as the caller can now
4374 set them directly.
4375
4376
4377 Version 0.92 11-Sep-97
4378 ----------------------
4379
4380 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4381 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4382
4383 2. Caseless matching was not working in character classes if the characters in
4384 the pattern were in upper case.
4385
4386 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4387
4388 4. Make PCRE_ANCHORED public and accept as a compile option.
4389
4390 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4391 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4392 pass them.
4393
4394 6. Give an error if bad option bits passed at compile or run time.
4395
4396 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4397 pcretest to cause it to pass that flag.
4398
4399 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4400 options, and the first character, if set.
4401
4402 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4403
4404
4405 Version 0.91 10-Sep-97
4406 ----------------------
4407
4408 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4409 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4410
4411 2. PCRE was looping on encountering an indefinitely repeated back reference to
4412 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4413 Perl does - treats the match as successful.
4414
4415 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5