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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 856 - (show annotations) (download)
Sat Jan 7 17:02:55 2012 UTC (2 years, 8 months ago) by ph10
File size: 224497 byte(s)
Documentation updates.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.30
5 ------------
6
7 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
8 name is defined in ctype.h.
9
10 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
11 only in quite long subpatterns.
12
13 3. Removed the function pcre_info(), which has been obsolete and deprecated
14 since it was replaced by pcre_fullinfo() in February 2000.
15
16 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
17 match a (*MARK), and the match failed at the start of the subject, a
18 reference to memory before the start of the subject could occur. This bug
19 was introduced by fix 17 of release 8.21.
20
21 5. A reference to an unset group with zero minimum repetition was giving
22 totally wrong answers (in non-JavaScript-compatibility mode). For example,
23 /(another)?(\1?)test/ matched against "hello world test". This bug was
24 introduced in release 8.13.
25
26 6. Add support for 16-bit character strings (a large amount of work involving
27 many changes and refactorings).
28
29 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
30 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
31 from a file.
32
33
34 Version 8.21 12-Dec-2011
35 ------------------------
36
37 1. Updating the JIT compiler.
38
39 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
40 are added as well.
41
42 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
43 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
44 calling _pcre_jit_exec. Some extra comments are added.
45
46 4. (*MARK) settings inside atomic groups that do not contain any capturing
47 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
48 was introduced by change 18 for 8.20.
49
50 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
51 ECMA-262 standard.
52
53 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
54 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
55 This bug was probably introduced by change 9 of 8.13.
56
57 7. While fixing 6 above, I noticed that a number of other items were being
58 incorrectly rejected as "not fixed length". This arose partly because newer
59 opcodes had not been added to the fixed-length checking code. I have (a)
60 corrected the bug and added tests for these items, and (b) arranged for an
61 error to occur if an unknown opcode is encountered while checking for fixed
62 length instead of just assuming "not fixed length". The items that were
63 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
64 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
65 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
66
67 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
68 being incorrectly compiled and would have given unpredicatble results.
69
70 9. A possessively repeated subpattern with minimum repeat count greater than
71 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
72 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
73 the first (A) could occur when it should not.
74
75 10. Add a cast and remove a redundant test from the code.
76
77 11. JIT should use pcre_malloc/pcre_free for allocation.
78
79 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
80 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
81 is anything other than /usr, -L is still shown).
82
83 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
84
85 14. Perl does not support \N without a following name in a [] class; PCRE now
86 also gives an error.
87
88 15. If a forward reference was repeated with an upper limit of around 2000,
89 it caused the error "internal error: overran compiling workspace". The
90 maximum number of forward references (including repeats) was limited by the
91 internal workspace, and dependent on the LINK_SIZE. The code has been
92 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
93 the default depends on LINK_SIZE. There is a new upper limit (for safety)
94 of around 200,000 forward references. While doing this, I also speeded up
95 the filling in of repeated forward references.
96
97 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
98 incorrectly expecting the subject to contain another "a" after the start.
99
100 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
101 in the match, the SKIP should be ignored. This was not happening; instead
102 the SKIP was being treated as NOMATCH. For patterns such as
103 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
104 tested.
105
106 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
107 now much more compatible with Perl, in particular in cases where the result
108 is a non-match for a non-anchored pattern. For example, if
109 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
110 "m", where previously it did not return a name. A side effect of this
111 change is that for partial matches, the last encountered mark name is
112 returned, as for non matches. A number of tests that were previously not
113 Perl-compatible have been moved into the Perl-compatible test files. The
114 refactoring has had the pleasing side effect of removing one argument from
115 the match() function, thus reducing its stack requirements.
116
117 19. If the /S+ option was used in pcretest to study a pattern using JIT,
118 subsequent uses of /S (without +) incorrectly behaved like /S+.
119
120 21. Retrieve executable code size support for the JIT compiler and fixing
121 some warnings.
122
123 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
124 not work when the shorter character appeared right at the end of the
125 subject string.
126
127 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
128 systems.
129
130 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
131 output it when the /M option is used in pcretest.
132
133 25. The CheckMan script was not being included in the distribution. Also, added
134 an explicit "perl" to run Perl scripts from the PrepareRelease script
135 because this is reportedly needed in Windows.
136
137 26. If study data was being save in a file and studying had not found a set of
138 "starts with" bytes for the pattern, the data written to the file (though
139 never used) was taken from uninitialized memory and so caused valgrind to
140 complain.
141
142 27. Updated RunTest.bat as provided by Sheri Pierce.
143
144 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
145
146 29. Computation of memory usage for the table of capturing group names was
147 giving an unnecessarily large value.
148
149
150 Version 8.20 21-Oct-2011
151 ------------------------
152
153 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
154 a POSIX class. After further experiments with Perl, which convinced me that
155 Perl has bugs and confusions, a closing square bracket is no longer allowed
156 in a POSIX name. This bug also affected patterns with classes that started
157 with full stops.
158
159 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
160 captured substring, but while checking the failing first alternative,
161 substring 1 is temporarily captured. If the output vector supplied to
162 pcre_exec() was not big enough for this capture, the yield of the function
163 was still zero ("insufficient space for captured substrings"). This cannot
164 be totally fixed without adding another stack variable, which seems a lot
165 of expense for a edge case. However, I have improved the situation in cases
166 such as /(a)(b)x|abc/ matched against "abc", where the return code
167 indicates that fewer than the maximum number of slots in the ovector have
168 been set.
169
170 3. Related to (2) above: when there are more back references in a pattern than
171 slots in the output vector, pcre_exec() uses temporary memory during
172 matching, and copies in the captures as far as possible afterwards. It was
173 using the entire output vector, but this conflicts with the specification
174 that only 2/3 is used for passing back captured substrings. Now it uses
175 only the first 2/3, for compatibility. This is, of course, another edge
176 case.
177
178 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
179 main code base, and can be used by building with --enable-jit. When this is
180 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
181 runtime --no-jit option is given.
182
183 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
184 ovector, the return from the function was zero, implying that there were
185 other matches that did not fit. The correct "exactly full" value is now
186 returned.
187
188 6. If a subpattern that was called recursively or as a subroutine contained
189 (*PRUNE) or any other control that caused it to give a non-standard return,
190 invalid errors such as "Error -26 (nested recursion at the same subject
191 position)" or even infinite loops could occur.
192
193 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
194 computing the minimum length on reaching *ACCEPT, and so ended up with the
195 wrong value of 1 rather than 0. Further investigation indicates that
196 computing a minimum subject length in the presence of *ACCEPT is difficult
197 (think back references, subroutine calls), and so I have changed the code
198 so that no minimum is registered for a pattern that contains *ACCEPT.
199
200 8. If (*THEN) was present in the first (true) branch of a conditional group,
201 it was not handled as intended. [But see 16 below.]
202
203 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
204 Sheri Pierce.
205
206 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
207 the first byte in a match must be "a".
208
209 11. Change 17 for 8.13 increased the recursion depth for patterns like
210 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
211 pattern contains any instances of (*THEN). If it does not, the old
212 optimizations are restored. It would be nice to do this on a per-group
213 basis, but at the moment that is not feasible.
214
215 12. In some environments, the output of pcretest -C is CRLF terminated. This
216 broke RunTest's code that checks for the link size. A single white space
217 character after the value is now allowed for.
218
219 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
220 For "fr", it uses the Windows-specific input and output files.
221
222 14. If (*THEN) appeared in a group that was called recursively or as a
223 subroutine, it did not work as intended. [But see next item.]
224
225 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
226 pattern fragments (but not containing any | characters). If A and B are
227 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
228 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
229 D. In other words, Perl considers parentheses that do not contain any |
230 characters to be part of a surrounding alternative, whereas PCRE was
231 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
232 differently. PCRE now behaves in the same way as Perl, except in the case
233 of subroutine/recursion calls such as (?1) which have in any case always
234 been different (but PCRE had them first :-).
235
236 16. Related to 15 above: Perl does not treat the | in a conditional group as
237 creating alternatives. Such a group is treated in the same way as an
238 ordinary group without any | characters when processing (*THEN). PCRE has
239 been changed to match Perl's behaviour.
240
241 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
242 RunGrepTest script failed.
243
244 18. Change 22 for version 13 caused atomic groups to use more stack. This is
245 inevitable for groups that contain captures, but it can lead to a lot of
246 stack use in large patterns. The old behaviour has been restored for atomic
247 groups that do not contain any capturing parentheses.
248
249 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
250 suppress the check for a minimum subject length at run time. (If it was
251 given to pcre_exec() or pcre_dfa_exec() it did work.)
252
253 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
254 fail to work when decoding hex characters in data strings in EBCDIC
255 environments.
256
257 21. It appears that in at least one Mac OS environment, the isxdigit() function
258 is implemented as a macro that evaluates to its argument more than once,
259 contravening the C 90 Standard (I haven't checked a later standard). There
260 was an instance in pcretest which caused it to go wrong when processing
261 \x{...} escapes in subject strings. The has been rewritten to avoid using
262 things like p++ in the argument of isxdigit().
263
264
265 Version 8.13 16-Aug-2011
266 ------------------------
267
268 1. The Unicode data tables have been updated to Unicode 6.0.0.
269
270 2. Two minor typos in pcre_internal.h have been fixed.
271
272 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
273 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
274 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
275
276 4. There were a number of related bugs in the code for matching backrefences
277 caselessly in UTF-8 mode when codes for the characters concerned were
278 different numbers of bytes. For example, U+023A and U+2C65 are an upper
279 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
280 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
281 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
282 2-byte code at the end of the subject (it thought there wasn't enough data
283 left).
284
285 5. Comprehensive information about what went wrong is now returned by
286 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
287 as the output vector has at least 2 elements. The offset of the start of
288 the failing character and a reason code are placed in the vector.
289
290 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
291 now returned is for the first byte of the failing character, instead of the
292 last byte inspected. This is an incompatible change, but I hope it is small
293 enough not to be a problem. It makes the returned offset consistent with
294 pcre_exec() and pcre_dfa_exec().
295
296 7. pcretest now gives a text phrase as well as the error number when
297 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
298 failure, the offset and reason code are output.
299
300 8. When \R was used with a maximizing quantifier it failed to skip backwards
301 over a \r\n pair if the subsequent match failed. Instead, it just skipped
302 back over a single character (\n). This seems wrong (because it treated the
303 two characters as a single entity when going forwards), conflicts with the
304 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
305 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
306 has been changed.
307
308 9. Some internal refactoring has changed the processing so that the handling
309 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
310 time (the PCRE_DOTALL option was changed this way some time ago: version
311 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
312 which was always a bit of a fudge. It also means that there is one less
313 argument for the match() function, which reduces its stack requirements
314 slightly. This change also fixes an incompatibility with Perl: the pattern
315 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
316
317 10. More internal refactoring has drastically reduced the number of recursive
318 calls to match() for possessively repeated groups such as (abc)++ when
319 using pcre_exec().
320
321 11. While implementing 10, a number of bugs in the handling of groups were
322 discovered and fixed:
323
324 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
325 (a|)*(?1) gave a compile-time internal error.
326 ((a|)+)+ did not notice that the outer group could match an empty string.
327 (^a|^)+ was not marked as anchored.
328 (.*a|.*)+ was not marked as matching at start or after a newline.
329
330 12. Yet more internal refactoring has removed another argument from the match()
331 function. Special calls to this function are now indicated by setting a
332 value in a variable in the "match data" data block.
333
334 13. Be more explicit in pcre_study() instead of relying on "default" for
335 opcodes that mean there is no starting character; this means that when new
336 ones are added and accidentally left out of pcre_study(), testing should
337 pick them up.
338
339 14. The -s option of pcretest has been documented for ages as being an old
340 synonym of -m (show memory usage). I have changed it to mean "force study
341 for every regex", that is, assume /S for every regex. This is similar to -i
342 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
343 using it. It makes it easier to run collections of tests with and without
344 study enabled, and thereby test pcre_study() more easily. All the standard
345 tests are now run with and without -s (but some patterns can be marked as
346 "never study" - see 20 below).
347
348 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
349 restoration of the capturing data to the outer values was not happening
350 correctly.
351
352 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
353 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
354 pattern had matched an empty string, and so incorrectly returned a no
355 match.
356
357 17. There was optimizing code for the last branch of non-capturing parentheses,
358 and also for the obeyed branch of a conditional subexpression, which used
359 tail recursion to cut down on stack usage. Unfortunately, now that there is
360 the possibility of (*THEN) occurring in these branches, tail recursion is
361 no longer possible because the return has to be checked for (*THEN). These
362 two optimizations have therefore been removed. [But see 8.20/11 above.]
363
364 18. If a pattern containing \R was studied, it was assumed that \R always
365 matched two bytes, thus causing the minimum subject length to be
366 incorrectly computed because \R can also match just one byte.
367
368 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
369 was incorrectly computed.
370
371 20. If /S is present twice on a test pattern in pcretest input, it now
372 *disables* studying, thereby overriding the use of -s on the command line
373 (see 14 above). This is necessary for one or two tests to keep the output
374 identical in both cases.
375
376 21. When (*ACCEPT) was used in an assertion that matched an empty string and
377 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
378
379 22. When an atomic group that contained a capturing parenthesis was
380 successfully matched, but the branch in which it appeared failed, the
381 capturing was not being forgotten if a higher numbered group was later
382 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
383 group 1 to "a", when in fact it should be unset. This applied to multi-
384 branched capturing and non-capturing groups, repeated or not, and also to
385 positive assertions (capturing in negative assertions does not happen
386 in PCRE) and also to nested atomic groups.
387
388 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
389 subject after a captured substring, to make it easier to tell which of a
390 number of identical substrings has been captured.
391
392 24. The way atomic groups are processed by pcre_exec() has been changed so that
393 if they are repeated, backtracking one repetition now resets captured
394 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
395 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
396 "aaa". Previously, it would have been "a". As part of this code
397 refactoring, the way recursive calls are handled has also been changed.
398
399 25. If an assertion condition captured any substrings, they were not passed
400 back unless some other capturing happened later. For example, if
401 (?(?=(a))a) was matched against "a", no capturing was returned.
402
403 26. When studying a pattern that contained subroutine calls or assertions,
404 the code for finding the minimum length of a possible match was handling
405 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
406 group 1 called group 2 while simultaneously a separate group 2 called group
407 1). A stack overflow occurred in this case. I have fixed this by limiting
408 the recursion depth to 10.
409
410 27. Updated RunTest.bat in the distribution to the version supplied by Tom
411 Fortmann. This supports explicit test numbers on the command line, and has
412 argument validation and error reporting.
413
414 28. An instance of \X with an unlimited repeat could fail if at any point the
415 first character it looked at was a mark character.
416
417 29. Some minor code refactoring concerning Unicode properties and scripts
418 should reduce the stack requirement of match() slightly.
419
420 30. Added the '=' option to pcretest to check the setting of unused capturing
421 slots at the end of the pattern, which are documented as being -1, but are
422 not included in the return count.
423
424 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
425 compiled something random. Now it gives a compile-time error (as does
426 Perl).
427
428 32. A *MARK encountered during the processing of a positive assertion is now
429 recorded and passed back (compatible with Perl).
430
431 33. If --only-matching or --colour was set on a pcregrep call whose pattern
432 had alternative anchored branches, the search for a second match in a line
433 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
434 matched the line "0102" twice. The same bug affected patterns that started
435 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
436 twice.
437
438 34. Previously, PCRE did not allow quantification of assertions. However, Perl
439 does, and because of capturing effects, quantifying parenthesized
440 assertions may at times be useful. Quantifiers are now allowed for
441 parenthesized assertions.
442
443 35. A minor code tidy in pcre_compile() when checking options for \R usage.
444
445 36. \g was being checked for fancy things in a character class, when it should
446 just be a literal "g".
447
448 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
449 appearance of a nested POSIX class supersedes an apparent external class.
450 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
451 unescaped square brackets may also appear as part of class names. For
452 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
453 more like Perl. (But see 8.20/1 above.)
454
455 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
456 was because it thought it was \N{name}, which is not supported).
457
458 39. Add minix to OS list not supporting the -S option in pcretest.
459
460 40. PCRE tries to detect cases of infinite recursion at compile time, but it
461 cannot analyze patterns in sufficient detail to catch mutual recursions
462 such as ((?1))((?2)). There is now a runtime test that gives an error if a
463 subgroup is called recursively as a subpattern for a second time at the
464 same position in the subject string. In previous releases this might have
465 been caught by the recursion limit, or it might have run out of stack.
466
467 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
468 happen only once. PCRE was, however incorrectly giving a compile time error
469 "recursive call could loop indefinitely" because it cannot analyze the
470 pattern in sufficient detail. The compile time test no longer happens when
471 PCRE is compiling a conditional subpattern, but actual runaway loops are
472 now caught at runtime (see 40 above).
473
474 42. It seems that Perl allows any characters other than a closing parenthesis
475 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
476 has been changed to be the same.
477
478 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
479 as not to get warnings when autogen.sh is called. Also changed
480 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
481
482 44. To help people who use pcregrep to scan files containing exceedingly long
483 lines, the following changes have been made:
484
485 (a) The default value of the buffer size parameter has been increased from
486 8K to 20K. (The actual buffer used is three times this size.)
487
488 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
489 PCRE is built.
490
491 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
492 to be set at run time.
493
494 (d) Numerical values in pcregrep options can be followed by K or M, for
495 example --buffer-size=50K.
496
497 (e) If a line being scanned overflows pcregrep's buffer, an error is now
498 given and the return code is set to 2.
499
500 45. Add a pointer to the latest mark to the callout data block.
501
502 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
503 partial match of an empty string instead of no match. This was specific to
504 the use of ".".
505
506 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
507 complete match instead of a partial match. This bug was dependent on both
508 the PCRE_UTF8 and PCRE_DOTALL options being set.
509
510 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
511 starting byte set, because \b was not being ignored.
512
513
514 Version 8.12 15-Jan-2011
515 ------------------------
516
517 1. Fixed some typos in the markup of the man pages, and wrote a script that
518 checks for such things as part of the documentation building process.
519
520 2. On a big-endian 64-bit system, pcregrep did not correctly process the
521 --match-limit and --recursion-limit options (added for 8.11). In
522 particular, this made one of the standard tests fail. (The integer value
523 went into the wrong half of a long int.)
524
525 3. If the --colour option was given to pcregrep with -v (invert match), it
526 did strange things, either producing crazy output, or crashing. It should,
527 of course, ignore a request for colour when reporting lines that do not
528 match.
529
530 4. Another pcregrep bug caused similar problems if --colour was specified with
531 -M (multiline) and the pattern match finished with a line ending.
532
533 5. In pcregrep, when a pattern that ended with a literal newline sequence was
534 matched in multiline mode, the following line was shown as part of the
535 match. This seems wrong, so I have changed it.
536
537 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
538 the check for further matches in the same line (so they could be coloured)
539 to overrun the end of the current line. If another match was found, it was
540 incorrectly shown (and then shown again when found in the next line).
541
542 7. If pcregrep was compiled under Windows, there was a reference to the
543 function pcregrep_exit() before it was defined. I am assuming this was
544 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
545 reported by a user. I've moved the definition above the reference.
546
547
548 Version 8.11 10-Dec-2010
549 ------------------------
550
551 1. (*THEN) was not working properly if there were untried alternatives prior
552 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
553 backtracked to try for "b" instead of moving to the next alternative branch
554 at the same level (in this case, to look for "c"). The Perl documentation
555 is clear that when (*THEN) is backtracked onto, it goes to the "next
556 alternative in the innermost enclosing group".
557
558 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
559 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
560 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
561 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
562 (*THEN).
563
564 3. If \s appeared in a character class, it removed the VT character from
565 the class, even if it had been included by some previous item, for example
566 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
567 of \s, but is part of the POSIX "space" class.)
568
569 4. A partial match never returns an empty string (because you can always
570 match an empty string at the end of the subject); however the checking for
571 an empty string was starting at the "start of match" point. This has been
572 changed to the "earliest inspected character" point, because the returned
573 data for a partial match starts at this character. This means that, for
574 example, /(?<=abc)def/ gives a partial match for the subject "abc"
575 (previously it gave "no match").
576
577 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
578 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
579 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
580 has an implication that the given string is incomplete (because a partial
581 match is preferred over a full match). For this reason, these items now
582 give a partial match in this situation. [Aside: previously, the one case
583 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
584 match rather than a full match, which was wrong by the old rules, but is
585 now correct.]
586
587 6. There was a bug in the handling of #-introduced comments, recognized when
588 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
589 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
590 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
591 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
592 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
593 places in pcre_compile().
594
595 7. Related to (6) above, when pcre_compile() was skipping #-introduced
596 comments when looking ahead for named forward references to subpatterns,
597 the only newline sequence it recognized was NL. It now handles newlines
598 according to the set newline convention.
599
600 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
601 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
602 cater for a lack of strerror(). These oversights have been fixed.
603
604 9. Added --match-limit and --recursion-limit to pcregrep.
605
606 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
607
608 11. When the -o option was used, pcregrep was setting a return code of 1, even
609 when matches were found, and --line-buffered was not being honoured.
610
611 12. Added an optional parentheses number to the -o and --only-matching options
612 of pcregrep.
613
614 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
615 can match an empty string. The code to do it in pcretest and pcredemo
616 needed fixing:
617
618 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
619 only one byte after an empty string match just before CRLF (this case
620 just got forgotten; "any" and "anycrlf" were OK).
621
622 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
623 mode when an empty string match preceded an ASCII character followed by
624 a non-ASCII character. (The code for advancing by one character rather
625 than one byte was nonsense.)
626
627 (c) The pcredemo.c sample program did not have any code at all to handle
628 the cases when CRLF is a valid newline sequence.
629
630 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
631 as a starting offset was within the subject string. There is now a new
632 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
633 negative or greater than the length of the string. In order to test this,
634 pcretest is extended to allow the setting of negative starting offsets.
635
636 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
637 starting offset points to the beginning of a UTF-8 character was
638 unnecessarily clumsy. I tidied it up.
639
640 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
641 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
642
643 17. Nobody had reported that the --include_dir option, which was added in
644 release 7.7 should have been called --include-dir (hyphen, not underscore)
645 for compatibility with GNU grep. I have changed it to --include-dir, but
646 left --include_dir as an undocumented synonym, and the same for
647 --exclude-dir, though that is not available in GNU grep, at least as of
648 release 2.5.4.
649
650 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
651 characters from a string of bytes) have been redefined so as not to use
652 loops, in order to improve performance in some environments. At the same
653 time, I abstracted some of the common code into auxiliary macros to save
654 repetition (this should not affect the compiled code).
655
656 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
657 compile-time error is now given if \c is not followed by an ASCII
658 character, that is, a byte less than 128. (In EBCDIC mode, the code is
659 different, and any byte value is allowed.)
660
661 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
662 START_OPTIMIZE option, which is now allowed at compile time - but just
663 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
664 to pcregrep and other applications that have no direct access to PCRE
665 options. The new /Y option in pcretest sets this option when calling
666 pcre_compile().
667
668 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
669 back references. Groups containing recursive back references were forced to
670 be atomic by that change, but in the case of named groups, the amount of
671 memory required was incorrectly computed, leading to "Failed: internal
672 error: code overflow". This has been fixed.
673
674 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
675 pcretest.c, to avoid build problems in some Borland environments.
676
677
678 Version 8.10 25-Jun-2010
679 ------------------------
680
681 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
682 THEN.
683
684 2. (*ACCEPT) was not working when inside an atomic group.
685
686 3. Inside a character class, \B is treated as a literal by default, but
687 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
688 causes the error). The code is unchanged, but I tidied the documentation.
689
690 4. Inside a character class, PCRE always treated \R and \X as literals,
691 whereas Perl faults them if its -w option is set. I have changed PCRE so
692 that it faults them when PCRE_EXTRA is set.
693
694 5. Added support for \N, which always matches any character other than
695 newline. (It is the same as "." when PCRE_DOTALL is not set.)
696
697 6. When compiling pcregrep with newer versions of gcc which may have
698 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
699 declared with attribute warn_unused_result" were given. Just casting the
700 result to (void) does not stop the warnings; a more elaborate fudge is
701 needed. I've used a macro to implement this.
702
703 7. Minor change to pcretest.c to avoid a compiler warning.
704
705 8. Added four artifical Unicode properties to help with an option to make
706 \s etc use properties (see next item). The new properties are: Xan
707 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
708
709 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
710 use Unicode properties. (*UCP) at the start of a pattern can be used to set
711 this option. Modified pcretest to add /W to test this facility. Added
712 REG_UCP to make it available via the POSIX interface.
713
714 10. Added --line-buffered to pcregrep.
715
716 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
717 studied, and the match started with a letter with a code point greater than
718 127 whose first byte was different to the first byte of the other case of
719 the letter, the other case of this starting letter was not recognized
720 (#976).
721
722 12. If a pattern that was studied started with a repeated Unicode property
723 test, for example, \p{Nd}+, there was the theoretical possibility of
724 setting up an incorrect bitmap of starting bytes, but fortunately it could
725 not have actually happened in practice until change 8 above was made (it
726 added property types that matched character-matching opcodes).
727
728 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
729 possible starting bytes for non-anchored patterns.
730
731 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
732 \R, and also a number of cases that involve Unicode properties, both
733 explicit and implicit when PCRE_UCP is set.
734
735 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
736 input, it could crash or give wrong results if characters with values
737 greater than 0xc0 were present in the subject string. (Detail: it assumed
738 UTF-8 input when processing these items.)
739
740 16. Added a lot of (int) casts to avoid compiler warnings in systems where
741 size_t is 64-bit (#991).
742
743 17. Added a check for running out of memory when PCRE is compiled with
744 --disable-stack-for-recursion (#990).
745
746 18. If the last data line in a file for pcretest does not have a newline on
747 the end, a newline was missing in the output.
748
749 19. The default pcre_chartables.c file recognizes only ASCII characters (values
750 less than 128) in its various bitmaps. However, there is a facility for
751 generating tables according to the current locale when PCRE is compiled. It
752 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
753 characters, are recognized by isspace() and therefore were getting set in
754 these tables, and indeed these tables seem to approximate to ISO 8859. This
755 caused a problem in UTF-8 mode when pcre_study() was used to create a list
756 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
757 which of course cannot start UTF-8 characters. I have changed the code so
758 that only real ASCII characters (less than 128) and the correct starting
759 bytes for UTF-8 encodings are set for characters greater than 127 when in
760 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
761 altogether.)
762
763 20. Added the /T option to pcretest so as to be able to run tests with non-
764 standard character tables, thus making it possible to include the tests
765 used for 19 above in the standard set of tests.
766
767 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
768 reference to a subpattern the other side of a comment that contains an
769 opening parenthesis caused either an internal compiling error, or a
770 reference to the wrong subpattern.
771
772
773 Version 8.02 19-Mar-2010
774 ------------------------
775
776 1. The Unicode data tables have been updated to Unicode 5.2.0.
777
778 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
779 configured.
780
781 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
782 original author of that file, following a query about its status.
783
784 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
785 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
786
787 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
788 quantifier applied to a forward-referencing subroutine call, could compile
789 incorrect code or give the error "internal error: previously-checked
790 referenced subpattern not found".
791
792 6. Both MS Visual Studio and Symbian OS have problems with initializing
793 variables to point to external functions. For these systems, therefore,
794 pcre_malloc etc. are now initialized to local functions that call the
795 relevant global functions.
796
797 7. There were two entries missing in the vectors called coptable and poptable
798 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
799 I've fixed the data, and added a kludgy way of testing at compile time that
800 the lengths are correct (equal to the number of opcodes).
801
802 8. Following on from 7, I added a similar kludge to check the length of the
803 eint vector in pcreposix.c.
804
805 9. Error texts for pcre_compile() are held as one long string to avoid too
806 much relocation at load time. To find a text, the string is searched,
807 counting zeros. There was no check for running off the end of the string,
808 which could happen if a new error number was added without updating the
809 string.
810
811 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
812
813 11. \K was not working if it appeared in an atomic group or in a group that
814 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
815 \K is "not well defined" if used in an assertion. PCRE now accepts it if
816 the assertion is positive, but not if it is negative.
817
818 12. Change 11 fortuitously reduced the size of the stack frame used in the
819 "match()" function of pcre_exec.c by one pointer. Forthcoming
820 implementation of support for (*MARK) will need an extra pointer on the
821 stack; I have reserved it now, so that the stack frame size does not
822 decrease.
823
824 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
825 item in branch that calls a recursion is a subroutine call - as in the
826 second branch in the above example - was incorrectly given the compile-
827 time error "recursive call could loop indefinitely" because pcre_compile()
828 was not correctly checking the subroutine for matching a non-empty string.
829
830 14. The checks for overrunning compiling workspace could trigger after an
831 overrun had occurred. This is a "should never occur" error, but it can be
832 triggered by pathological patterns such as hundreds of nested parentheses.
833 The checks now trigger 100 bytes before the end of the workspace.
834
835 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
836
837
838 Version 8.01 19-Jan-2010
839 ------------------------
840
841 1. If a pattern contained a conditional subpattern with only one branch (in
842 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
843 computed the wrong minimum data length (which is of course zero for such
844 subpatterns). This could cause incorrect "no match" results.
845
846 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
847 the pattern is reset in the first branch, pcre_compile() failed with
848 "internal error: code overflow at offset...". This happened only when
849 the reset was to the original external option setting. (An optimization
850 abstracts leading options settings into an external setting, which was the
851 cause of this.)
852
853 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
854 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
855 assertion pattern did not match (meaning that the assertion was true), it
856 was incorrectly treated as false if the SKIP had been reached during the
857 matching. This also applied to assertions used as conditions.
858
859 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
860 assertion subpattern, including such a pattern used as a condition,
861 unpredictable results occurred, instead of the error return
862 PCRE_ERROR_DFA_UITEM.
863
864 5. The C++ GlobalReplace function was not working like Perl for the special
865 situation when an empty string is matched. It now does the fancy magic
866 stuff that is necessary.
867
868 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
869 removed. (These were left over from very, very early versions of PCRE.)
870
871 7. Some cosmetic changes to the code to make life easier when compiling it
872 as part of something else:
873
874 (a) Change DEBUG to PCRE_DEBUG.
875
876 (b) In pcre_compile(), rename the member of the "branch_chain" structure
877 called "current" as "current_branch", to prevent a collision with the
878 Linux macro when compiled as a kernel module.
879
880 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
881 prevent a collision with the Linux macro when compiled as a kernel
882 module.
883
884 8. In pcre_compile() there are some checks for integer overflows that used to
885 cast potentially large values to (double). This has been changed to that
886 when building, a check for int64_t is made, and if it is found, it is used
887 instead, thus avoiding the use of floating point arithmetic. (There is no
888 other use of FP in PCRE.) If int64_t is not found, the fallback is to
889 double.
890
891 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
892 2005 (difference between two addresses compared to an unsigned value).
893
894 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
895 custom one, because of the following reported problem in Windows:
896
897 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
898 under Win32.
899 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
900 therefore missing the function definition.
901 - The compiler thus generates a "C" signature for the test function.
902 - The linker fails to find the "C" function.
903 - PCRE fails to configure if asked to do so against libbz2.
904
905 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
906 messages were output:
907
908 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
909 rerunning libtoolize, to keep the correct libtool macros in-tree.
910 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
911
912 I have done both of these things.
913
914 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
915 most of the time, it *can* run out if it is given a pattern that contains a
916 runaway infinite recursion. I updated the discussion in the pcrestack man
917 page.
918
919 13. Now that we have gone to the x.xx style of version numbers, the minor
920 version may start with zero. Using 08 or 09 is a bad idea because users
921 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
922 interpreted as invalid octal numbers. I've updated the previous comment in
923 configure.ac, and also added a check that gives an error if 08 or 09 are
924 used.
925
926 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
927 causing partial matching to fail when the end of the subject matched \W
928 in a UTF-8 pattern where \W was quantified with a minimum of 3.
929
930 15. There were some discrepancies between the declarations in pcre_internal.h
931 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
932 their definitions. The declarations used "const uschar *" and the
933 definitions used USPTR. Even though USPTR is normally defined as "const
934 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
935 reported that: "This difference in casting confuses some C++ compilers, for
936 example, SunCC recognizes above declarations as different functions and
937 generates broken code for hbpcre." I have changed the declarations to use
938 USPTR.
939
940 16. GNU libtool is named differently on some systems. The autogen.sh script now
941 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
942 (FreeBSD).
943
944 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
945 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
946 comment: "Figure out how to create a longlong from a string: strtoll and
947 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
948 instance, but it only takes 2 args instead of 3!"
949
950 18. A subtle bug concerned with back references has been fixed by a change of
951 specification, with a corresponding code fix. A pattern such as
952 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
953 refers, was giving matches when it shouldn't. For example, xa=xaaa would
954 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
955 same bug. Such groups have to be quantified to be useful, or contained
956 inside another quantified group. (If there's no repetition, the reference
957 can never match.) The problem arises because, having left the group and
958 moved on to the rest of the pattern, a later failure that backtracks into
959 the group uses the captured value from the final iteration of the group
960 rather than the correct earlier one. I have fixed this in PCRE by forcing
961 any group that contains a reference to itself to be an atomic group; that
962 is, there cannot be any backtracking into it once it has completed. This is
963 similar to recursive and subroutine calls.
964
965
966 Version 8.00 19-Oct-09
967 ----------------------
968
969 1. The table for translating pcre_compile() error codes into POSIX error codes
970 was out-of-date, and there was no check on the pcre_compile() error code
971 being within the table. This could lead to an OK return being given in
972 error.
973
974 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
975 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
976 in a Windows environment.
977
978 3. The pcregrep --count option prints the count for each file even when it is
979 zero, as does GNU grep. However, pcregrep was also printing all files when
980 --files-with-matches was added. Now, when both options are given, it prints
981 counts only for those files that have at least one match. (GNU grep just
982 prints the file name in this circumstance, but including the count seems
983 more useful - otherwise, why use --count?) Also ensured that the
984 combination -clh just lists non-zero counts, with no names.
985
986 4. The long form of the pcregrep -F option was incorrectly implemented as
987 --fixed_strings instead of --fixed-strings. This is an incompatible change,
988 but it seems right to fix it, and I didn't think it was worth preserving
989 the old behaviour.
990
991 5. The command line items --regex=pattern and --regexp=pattern were not
992 recognized by pcregrep, which required --regex pattern or --regexp pattern
993 (with a space rather than an '='). The man page documented the '=' forms,
994 which are compatible with GNU grep; these now work.
995
996 6. No libpcreposix.pc file was created for pkg-config; there was just
997 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
998
999 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1000 when UCP support is not needed, by modifying the Python script that
1001 generates it from Unicode data files. This should not matter if the module
1002 is correctly used as a library, but I received one complaint about 50K of
1003 unwanted data. My guess is that the person linked everything into his
1004 program rather than using a library. Anyway, it does no harm.
1005
1006 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1007 was a minimum greater than 1 for a wide character in a possessive
1008 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1009 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1010 character. Chaos in the form of incorrect output or a compiling loop could
1011 result.
1012
1013 9. The restrictions on what a pattern can contain when partial matching is
1014 requested for pcre_exec() have been removed. All patterns can now be
1015 partially matched by this function. In addition, if there are at least two
1016 slots in the offset vector, the offset of the earliest inspected character
1017 for the match and the offset of the end of the subject are set in them when
1018 PCRE_ERROR_PARTIAL is returned.
1019
1020 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1021 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1022 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1023 and may be more useful for multi-segment matching.
1024
1025 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1026 used to be given if ever the end of the subject was reached; now it is
1027 given only if matching could not proceed because another character was
1028 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1029 string "Z", which now yields "no match" instead of "partial match". In the
1030 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1031 final character ended with (*FAIL).
1032
1033 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1034 if the pattern had a "must contain" character that was already found in the
1035 earlier partial match, unless partial matching was again requested. For
1036 example, with the pattern /dog.(body)?/, the "must contain" character is
1037 "g". If the first part-match was for the string "dog", restarting with
1038 "sbody" failed. This bug has been fixed.
1039
1040 13. The string returned by pcre_dfa_exec() after a partial match has been
1041 changed so that it starts at the first inspected character rather than the
1042 first character of the match. This makes a difference only if the pattern
1043 starts with a lookbehind assertion or \b or \B (\K is not supported by
1044 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1045 matching functions compatible, and I think it's the right thing to do.
1046
1047 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1048 so that the demonstration program is easily available in environments where
1049 PCRE has not been installed from source.
1050
1051 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1052 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1053 library.
1054
1055 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1056 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1057 is not the first non-POSIX option to be added. Clearly some people find
1058 these options useful.
1059
1060 17. If a caller to the POSIX matching function regexec() passes a non-zero
1061 value for nmatch with a NULL value for pmatch, the value of
1062 nmatch is forced to zero.
1063
1064 18. RunGrepTest did not have a test for the availability of the -u option of
1065 the diff command, as RunTest does. It now checks in the same way as
1066 RunTest, and also checks for the -b option.
1067
1068 19. If an odd number of negated classes containing just a single character
1069 interposed, within parentheses, between a forward reference to a named
1070 subpattern and the definition of the subpattern, compilation crashed with
1071 an internal error, complaining that it could not find the referenced
1072 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1073 [The bug was that it was starting one character too far in when skipping
1074 over the character class, thus treating the ] as data rather than
1075 terminating the class. This meant it could skip too much.]
1076
1077 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1078 /g option in pcretest when the pattern contains \K, which makes it possible
1079 to have an empty string match not at the start, even when the pattern is
1080 anchored. Updated pcretest and pcredemo to use this option.
1081
1082 21. If the maximum number of capturing subpatterns in a recursion was greater
1083 than the maximum at the outer level, the higher number was returned, but
1084 with unset values at the outer level. The correct (outer level) value is
1085 now given.
1086
1087 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1088 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1089 make it do so. The string so far is captured, making this feature
1090 compatible with Perl.
1091
1092 23. The tests have been re-organized, adding tests 11 and 12, to make it
1093 possible to check the Perl 5.10 features against Perl 5.10.
1094
1095 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1096 pattern matches a fixed length string. PCRE did not allow this; now it
1097 does. Neither allows recursion.
1098
1099 25. I finally figured out how to implement a request to provide the minimum
1100 length of subject string that was needed in order to match a given pattern.
1101 (It was back references and recursion that I had previously got hung up
1102 on.) This code has now been added to pcre_study(); it finds a lower bound
1103 to the length of subject needed. It is not necessarily the greatest lower
1104 bound, but using it to avoid searching strings that are too short does give
1105 some useful speed-ups. The value is available to calling programs via
1106 pcre_fullinfo().
1107
1108 26. While implementing 25, I discovered to my embarrassment that pcretest had
1109 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1110 study optimizations had never been tested with that matching function.
1111 Oops. What is worse, even when it was passed study data, there was a bug in
1112 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1113 were also very few tests of studied patterns with pcre_dfa_exec().
1114
1115 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1116 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1117 on the other side of the coin, they are no longer allowed to have different
1118 names, because these cannot be distinguished in PCRE, and this has caused
1119 confusion. (This is a difference from Perl.)
1120
1121 28. When duplicate subpattern names are present (necessarily with different
1122 numbers, as required by 27 above), and a test is made by name in a
1123 conditional pattern, either for a subpattern having been matched, or for
1124 recursion in such a pattern, all the associated numbered subpatterns are
1125 tested, and the overall condition is true if the condition is true for any
1126 one of them. This is the way Perl works, and is also more like the way
1127 testing by number works.
1128
1129
1130 Version 7.9 11-Apr-09
1131 ---------------------
1132
1133 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1134 (pcretest), all targets were linked against these libraries. This included
1135 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1136 libraries. This caused unwanted dependencies to be created. This problem
1137 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1138 pcretest is linked with readline.
1139
1140 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1141 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1142 moved outside it again, because FALSE and TRUE are already defined in AIX,
1143 but BOOL is not.
1144
1145 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1146 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1147
1148 4. The pcregrep documentation said spaces were inserted as well as colons (or
1149 hyphens) following file names and line numbers when outputting matching
1150 lines. This is not true; no spaces are inserted. I have also clarified the
1151 wording for the --colour (or --color) option.
1152
1153 5. In pcregrep, when --colour was used with -o, the list of matching strings
1154 was not coloured; this is different to GNU grep, so I have changed it to be
1155 the same.
1156
1157 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1158 each matching line was coloured. Now it goes on to look for further matches
1159 of any of the test patterns, which is the same behaviour as GNU grep.
1160
1161 7. A pattern that could match an empty string could cause pcregrep to loop; it
1162 doesn't make sense to accept an empty string match in pcregrep, so I have
1163 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1164 seems to be how GNU grep behaves.
1165
1166 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1167 start or after a newline", because the conditional assertion was not being
1168 correctly handled. The rule now is that both the assertion and what follows
1169 in the first alternative must satisfy the test.
1170
1171 9. If auto-callout was enabled in a pattern with a conditional group whose
1172 condition was an assertion, PCRE could crash during matching, both with
1173 pcre_exec() and pcre_dfa_exec().
1174
1175 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1176 used for matching.
1177
1178 11. Unicode property support in character classes was not working for
1179 characters (bytes) greater than 127 when not in UTF-8 mode.
1180
1181 12. Added the -M command line option to pcretest.
1182
1183 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1184
1185 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1186
1187 16. Added comments and documentation about mis-use of no_arg in the C++
1188 wrapper.
1189
1190 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1191 from Martin Jerabek that uses macro names for all relevant character and
1192 string constants.
1193
1194 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1195 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1196 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1197 these, but not everybody uses configure.
1198
1199 19. A conditional group that had only one branch was not being correctly
1200 recognized as an item that could match an empty string. This meant that an
1201 enclosing group might also not be so recognized, causing infinite looping
1202 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1203 with the subject "ab", where knowledge that the repeated group can match
1204 nothing is needed in order to break the loop.
1205
1206 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1207 exec(), but without supplying a callout function, matching went wrong.
1208
1209 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1210 leak if the size of the offset vector was greater than 30. When the vector
1211 is smaller, the saved offsets during recursion go onto a local stack
1212 vector, but for larger vectors malloc() is used. It was failing to free
1213 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1214 error, in fact).
1215
1216 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1217 heapframe that is used only when UTF-8 support is enabled. This caused no
1218 problem, but was untidy.
1219
1220 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1221 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1222 included within another project.
1223
1224 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1225 slightly modified by me:
1226
1227 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1228 not building pcregrep.
1229
1230 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1231 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1232
1233 25. Forward references, both numeric and by name, in patterns that made use of
1234 duplicate group numbers, could behave incorrectly or give incorrect errors,
1235 because when scanning forward to find the reference group, PCRE was not
1236 taking into account the duplicate group numbers. A pattern such as
1237 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1238
1239 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1240 the feature of a custom pointer more persuasive (as requested by a user).
1241
1242 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1243 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1244 pre-defined.
1245
1246 28. Added support for (*UTF8) at the start of a pattern.
1247
1248 29. Arrange for flags added by the "release type" setting in CMake to be shown
1249 in the configuration summary.
1250
1251
1252 Version 7.8 05-Sep-08
1253 ---------------------
1254
1255 1. Replaced UCP searching code with optimized version as implemented for Ad
1256 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1257 stage table and inline lookup instead of a function, giving speed ups of 2
1258 to 5 times on some simple patterns that I tested. Permission was given to
1259 distribute the MultiStage2.py script that generates the tables (it's not in
1260 the tarball, but is in the Subversion repository).
1261
1262 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1263 scripts.
1264
1265 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1266 a group with a zero qualifier. The result of the study could be incorrect,
1267 or the function might crash, depending on the pattern.
1268
1269 4. Caseless matching was not working for non-ASCII characters in back
1270 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1271 It now works when Unicode Property Support is available.
1272
1273 5. In pcretest, an escape such as \x{de} in the data was always generating
1274 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1275 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1276 truncation.
1277
1278 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1279
1280 7. Added two (int) casts to pcregrep when printing the difference of two
1281 pointers, in case they are 64-bit values.
1282
1283 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1284 test 2 if it fails.
1285
1286 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1287 and a #define of that name to empty if it is not externally set. This is to
1288 allow users of MSVC to set it if necessary.
1289
1290 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1291 the convenience functions in the pcre_get.c source file.
1292
1293 11. An option change at the start of a pattern that had top-level alternatives
1294 could cause overwriting and/or a crash. This command provoked a crash in
1295 some environments:
1296
1297 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1298
1299 This potential security problem was recorded as CVE-2008-2371.
1300
1301 12. For a pattern where the match had to start at the beginning or immediately
1302 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1303 pcre_dfa_exec() could read past the end of the passed subject if there was
1304 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1305 pcretest so that it places the subject at the end of its malloc-ed buffer.
1306
1307 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1308 exec() might read past the end of the data buffer in UTF-8 mode.
1309
1310 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1311 the data contained the byte 0x85 as part of a UTF-8 character within its
1312 first line. This applied both to normal and DFA matching.
1313
1314 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1315 /^[^d]*?$/8 failed to match "abc".
1316
1317 16. Added a missing copyright notice to pcrecpp_internal.h.
1318
1319 17. Make it more clear in the documentation that values returned from
1320 pcre_exec() in ovector are byte offsets, not character counts.
1321
1322 18. Tidied a few places to stop certain compilers from issuing warnings.
1323
1324 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1325 supplied by Stefan Weber. I made a further small update for 7.8 because
1326 there is a change of source arrangements: the pcre_searchfuncs.c module is
1327 replaced by pcre_ucd.c.
1328
1329
1330 Version 7.7 07-May-08
1331 ---------------------
1332
1333 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1334 a string to a long long, pretend we don't even have a long long." This is
1335 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1336
1337 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1338 pre-7.6 versions, which defined a global no_arg variable instead of putting
1339 it in the RE class. (See also #8 below.)
1340
1341 3. Remove a line of dead code, identified by coverity and reported by Nuno
1342 Lopes.
1343
1344 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1345
1346 (1) The include/exclude patterns were being applied to the whole pathnames
1347 of files, instead of just to the final components.
1348
1349 (2) If there was more than one level of directory, the subdirectories were
1350 skipped unless they satisfied the include/exclude conditions. This is
1351 inconsistent with GNU grep (and could even be seen as contrary to the
1352 pcregrep specification - which I improved to make it absolutely clear).
1353 The action now is always to scan all levels of directory, and just
1354 apply the include/exclude patterns to regular files.
1355
1356 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1357 --exclude_dir in the tests to avoid scanning .svn directories.
1358
1359 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1360 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1361 doesn't support NULs in patterns.
1362
1363 7. Added some missing "const"s to declarations of static tables in
1364 pcre_compile.c and pcre_dfa_exec.c.
1365
1366 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1367 caused by fix #2 above. (Subsequently also a second patch to fix the
1368 first patch. And a third patch - this was a messy problem.)
1369
1370 9. Applied Craig's patch to remove the use of push_back().
1371
1372 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1373 matching function regexec().
1374
1375 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1376 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1377 references. PCRE supports relative numbers with this syntax (I don't think
1378 Oniguruma does).
1379
1380 12. Previously, a group with a zero repeat such as (...){0} was completely
1381 omitted from the compiled regex. However, this means that if the group
1382 was called as a subroutine from elsewhere in the pattern, things went wrong
1383 (an internal error was given). Such groups are now left in the compiled
1384 pattern, with a new opcode that causes them to be skipped at execution
1385 time.
1386
1387 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1388 to the way PCRE behaves:
1389
1390 (a) A lone ] character is dis-allowed (Perl treats it as data).
1391
1392 (b) A back reference to an unmatched subpattern matches an empty string
1393 (Perl fails the current match path).
1394
1395 (c) A data ] in a character class must be notated as \] because if the
1396 first data character in a class is ], it defines an empty class. (In
1397 Perl it is not possible to have an empty class.) The empty class []
1398 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1399 The negative empty class [^] matches any one character, independently
1400 of the DOTALL setting.
1401
1402 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1403 non-existent subpattern following a character class starting with ']' and
1404 containing () gave an internal compiling error instead of "reference to
1405 non-existent subpattern". Fortunately, when the pattern did exist, the
1406 compiled code was correct. (When scanning forwards to check for the
1407 existencd of the subpattern, it was treating the data ']' as terminating
1408 the class, so got the count wrong. When actually compiling, the reference
1409 was subsequently set up correctly.)
1410
1411 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1412 it was being rejected as not supported by pcre_dfa_exec(), even though
1413 other assertions are supported. I have made pcre_dfa_exec() support
1414 (*FAIL).
1415
1416 16. The implementation of 13c above involved the invention of a new opcode,
1417 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1418 cannot be changed at match time, I realized I could make a small
1419 improvement to matching performance by compiling OP_ALLANY instead of
1420 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1421 on the OP_ANY path.
1422
1423 17. Compiling pcretest on Windows with readline support failed without the
1424 following two fixes: (1) Make the unistd.h include conditional on
1425 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1426
1427 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1428 ncurses library to be included for pcretest when ReadLine support is
1429 requested, but also to allow for it to be overridden. This patch came from
1430 Daniel Bergström.
1431
1432 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1433 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1434 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1435 spotting this.
1436
1437
1438 Version 7.6 28-Jan-08
1439 ---------------------
1440
1441 1. A character class containing a very large number of characters with
1442 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1443 overflow.
1444
1445 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1446 HAVE_LONG_LONG is not defined.
1447
1448 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1449 bring them up to date and include new features. This patch includes:
1450
1451 - Fixed PH's badly added libz and libbz2 support.
1452 - Fixed a problem with static linking.
1453 - Added pcredemo. [But later removed - see 7 below.]
1454 - Fixed dftables problem and added an option.
1455 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1456 HAVE_LONG_LONG.
1457 - Added readline support for pcretest.
1458 - Added an listing of the option settings after cmake has run.
1459
1460 4. A user submitted a patch to Makefile that makes it easy to create
1461 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1462 Makefile.am that cause it to include this special target, without
1463 affecting anything else. Note that the same mingw target plus all
1464 the other distribution libraries and programs are now supported
1465 when configuring with CMake (see 6 below) instead of with
1466 Configure/Make.
1467
1468 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1469 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1470 exported in the Windows port". It has not yet been confirmed that the patch
1471 solves the problem, but it does no harm.
1472
1473 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1474 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1475 with CMake, and also correct the comment about stack recursion.
1476
1477 7. Remove the automatic building of pcredemo from the ./configure system and
1478 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1479 of a program that users should build themselves after PCRE is installed, so
1480 building it automatically is not really right. What is more, it gave
1481 trouble in some build environments.
1482
1483 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1484
1485
1486 Version 7.5 10-Jan-08
1487 ---------------------
1488
1489 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1490 values in parens when parsing an RE using the C++ wrapper."
1491
1492 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1493 Characters greater than 255 were excluded from the class instead of being
1494 included.
1495
1496 3. The same bug as (2) above applied to negated POSIX classes such as
1497 [:^space:].
1498
1499 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1500 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1501 I have changed it.
1502
1503 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1504 first named subpattern) and a construct such as (?&a) would reference the
1505 first named subpattern whose name started with "a" (in other words, the
1506 length check was missing). Both these problems are fixed. "Subpattern name
1507 expected" is now given for (?&) (a zero-length name), and this patch also
1508 makes it give the same error for \k'' (previously it complained that that
1509 was a reference to a non-existent subpattern).
1510
1511 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1512 this is right because (?- can be followed by option settings as well as by
1513 digits. I have, however, made the messages clearer.
1514
1515 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1516 than the number used in the conditional) now cause a compile-time error.
1517 This is actually not compatible with Perl, which accepts such patterns, but
1518 treats the conditional as always being FALSE (as PCRE used to), but it
1519 seems to me that giving a diagnostic is better.
1520
1521 8. Change "alphameric" to the more common word "alphanumeric" in comments
1522 and messages.
1523
1524 9. Fix two occurrences of "backslash" in comments that should have been
1525 "backspace".
1526
1527 10. Remove two redundant lines of code that can never be obeyed (their function
1528 was moved elsewhere).
1529
1530 11. The program that makes PCRE's Unicode character property table had a bug
1531 which caused it to generate incorrect table entries for sequences of
1532 characters that have the same character type, but are in different scripts.
1533 It amalgamated them into a single range, with the script of the first of
1534 them. In other words, some characters were in the wrong script. There were
1535 thirteen such cases, affecting characters in the following ranges:
1536
1537 U+002b0 - U+002c1
1538 U+0060c - U+0060d
1539 U+0061e - U+00612
1540 U+0064b - U+0065e
1541 U+0074d - U+0076d
1542 U+01800 - U+01805
1543 U+01d00 - U+01d77
1544 U+01d9b - U+01dbf
1545 U+0200b - U+0200f
1546 U+030fc - U+030fe
1547 U+03260 - U+0327f
1548 U+0fb46 - U+0fbb1
1549 U+10450 - U+1049d
1550
1551 12. The -o option (show only the matching part of a line) for pcregrep was not
1552 compatible with GNU grep in that, if there was more than one match in a
1553 line, it showed only the first of them. It now behaves in the same way as
1554 GNU grep.
1555
1556 13. If the -o and -v options were combined for pcregrep, it printed a blank
1557 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1558 does the same. The return code can be used to tell if there were any
1559 non-matching lines.
1560
1561 14. Added --file-offsets and --line-offsets to pcregrep.
1562
1563 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1564 infinitely looping recursion. The bug was that positive lookaheads were not
1565 being skipped when checking for a possible empty match (negative lookaheads
1566 and both kinds of lookbehind were skipped).
1567
1568 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1569 inclusion of <windows.h> to before rather than after the definition of
1570 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1571
1572 17. Specifying a possessive quantifier with a specific limit for a Unicode
1573 character property caused pcre_compile() to compile bad code, which led at
1574 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1575 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1576 caused the error; without that there was no problem.
1577
1578 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1579
1580 19. Added --enable-pcretest-libreadline.
1581
1582 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1583 RE::GlobalReplace(). As a result, the number of replacements returned was
1584 double what it should be. I removed one of the increments, but Craig sent a
1585 later patch that removed the other one (the right fix) and added unit tests
1586 that check the return values (which was not done before).
1587
1588 21. Several CMake things:
1589
1590 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1591 the names libpcre and libpcreposix, not just pcre and pcreposix.
1592
1593 (2) The above change means that pcretest and pcregrep are now correctly
1594 linked with the newly-built libraries, not previously installed ones.
1595
1596 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1597
1598 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1599 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1600 UTF-8 newline character). The key issue is that the pattern starts .*;
1601 this means that the match must be either at the beginning, or after a
1602 newline. The bug was in the code for advancing after a failed match and
1603 checking that the new position followed a newline. It was not taking
1604 account of UTF-8 characters correctly.
1605
1606 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1607 character classes. PCRE was not treating the sequence [:...:] as a
1608 character class unless the ... were all letters. Perl, however, seems to
1609 allow any characters between [: and :], though of course it rejects as
1610 unknown any "names" that contain non-letters, because all the known class
1611 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1612 for example, whereas PCRE did not - it did not recognize a POSIX character
1613 class. This seemed a bit dangerous, so the code has been changed to be
1614 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1615 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1616 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1617 Perl does, and where it didn't before.
1618
1619 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1620 Windows environments %n is disabled by default.
1621
1622
1623 Version 7.4 21-Sep-07
1624 ---------------------
1625
1626 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1627 means that a class such as [\s] counted as "explicit reference to CR or
1628 LF". That isn't really right - the whole point of the change was to try to
1629 help when there was an actual mention of one of the two characters. So now
1630 the change happens only if \r or \n (or a literal CR or LF) character is
1631 encountered.
1632
1633 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1634 of both had grown to the point where there were only 3 bits left.
1635 Fortunately, there was spare space in the data structure, and so I have
1636 moved the internal flags into a new 16-bit field to free up more option
1637 bits.
1638
1639 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1640 but did not set the internal JCHANGED flag - either of these is enough to
1641 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1642 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1643 start sets both bits.
1644
1645 4. Added options (at build time, compile time, exec time) to change \R from
1646 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1647
1648 5. doc/pcresyntax.html was missing from the distribution.
1649
1650 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1651 compatibility, even though it is no longer used.
1652
1653 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1654 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1655 windows.h file is present (where different names are used). [This was
1656 reversed later after testing - see 16 below.]
1657
1658 8. Changed all #include <config.h> to #include "config.h". There were also
1659 some further <pcre.h> cases that I changed to "pcre.h".
1660
1661 9. When pcregrep was used with the --colour option, it missed the line ending
1662 sequence off the lines that it output.
1663
1664 10. It was pointed out to me that arrays of string pointers cause lots of
1665 relocations when a shared library is dynamically loaded. A technique of
1666 using a single long string with a table of offsets can drastically reduce
1667 these. I have refactored PCRE in four places to do this. The result is
1668 dramatic:
1669
1670 Originally: 290
1671 After changing UCP table: 187
1672 After changing error message table: 43
1673 After changing table of "verbs" 36
1674 After changing table of Posix names 22
1675
1676 Thanks to the folks working on Gregex for glib for this insight.
1677
1678 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1679 unicode-properties was also set.
1680
1681 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1682
1683 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1684 checked only for CRLF.
1685
1686 14. Added casts to pcretest.c to avoid compiler warnings.
1687
1688 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1689
1690 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1691 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1692 entirely. This removes changes made in 7 above.
1693
1694 17. The CMake files have been updated, and there is now more information about
1695 building with CMake in the NON-UNIX-USE document.
1696
1697
1698 Version 7.3 28-Aug-07
1699 ---------------------
1700
1701 1. In the rejigging of the build system that eventually resulted in 7.1, the
1702 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1703 brackets there is not right, since it causes compilers to look for an
1704 installed pcre.h, not the version that is in the source that is being
1705 compiled (which of course may be different). I have changed it back to:
1706
1707 #include "pcre.h"
1708
1709 I have a vague recollection that the change was concerned with compiling in
1710 different directories, but in the new build system, that is taken care of
1711 by the VPATH setting the Makefile.
1712
1713 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1714 when the subject happened to end in the byte 0x85 (e.g. if the last
1715 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1716 characters but of course it shouldn't be taken as a newline when it is part
1717 of another character. The bug was that, for an unlimited repeat of . in
1718 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1719 characters when looking for a newline.
1720
1721 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1722
1723 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1724 in debug output.
1725
1726 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1727 long printing in the pcrecpp unittest when running under MinGW.
1728
1729 6. ESC_K was left out of the EBCDIC table.
1730
1731 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1732 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1733 limit also applies to "virtual nesting" when a pattern is recursive, and in
1734 this case 1000 isn't so big. I have been able to remove this limit at the
1735 expense of backing off one optimization in certain circumstances. Normally,
1736 when pcre_exec() would call its internal match() function recursively and
1737 immediately return the result unconditionally, it uses a "tail recursion"
1738 feature to save stack. However, when a subpattern that can match an empty
1739 string has an unlimited repetition quantifier, it no longer makes this
1740 optimization. That gives it a stack frame in which to save the data for
1741 checking that an empty string has been matched. Previously this was taken
1742 from the 1000-entry workspace that had been reserved. So now there is no
1743 explicit limit, but more stack is used.
1744
1745 8. Applied Daniel's patches to solve problems with the import/export magic
1746 syntax that is required for Windows, and which was going wrong for the
1747 pcreposix and pcrecpp parts of the library. These were overlooked when this
1748 problem was solved for the main library.
1749
1750 9. There were some crude static tests to avoid integer overflow when computing
1751 the size of patterns that contain repeated groups with explicit upper
1752 limits. As the maximum quantifier is 65535, the maximum group length was
1753 set at 30,000 so that the product of these two numbers did not overflow a
1754 32-bit integer. However, it turns out that people want to use groups that
1755 are longer than 30,000 bytes (though not repeat them that many times).
1756 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1757 made it possible to implement the integer overflow checks in a much more
1758 dynamic way, which I have now done. The artificial limitation on group
1759 length has been removed - we now have only the limit on the total length of
1760 the compiled pattern, which depends on the LINK_SIZE setting.
1761
1762 10. Fixed a bug in the documentation for get/copy named substring when
1763 duplicate names are permitted. If none of the named substrings are set, the
1764 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1765 empty string.
1766
1767 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1768 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1769 because the ] is interpreted as the first data character and the
1770 terminating ] is not found. PCRE has been made compatible with Perl in this
1771 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1772 cause memory overwriting.
1773
1774 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1775 string has been matched (to stop an infinite loop). It was not recognizing
1776 a conditional subpattern that could match an empty string if that
1777 subpattern was within another subpattern. For example, it looped when
1778 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1779 condition was not nested. This bug has been fixed.
1780
1781 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1782 past the start of the subject in the presence of bytes with the top bit
1783 set, for example "\x8aBCD".
1784
1785 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1786 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1787
1788 14. Optimized (?!) to (*FAIL).
1789
1790 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1791 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1792 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1793 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1794 does: it's just the validity check that is more restrictive.
1795
1796 16. Inserted checks for integer overflows during escape sequence (backslash)
1797 processing, and also fixed erroneous offset values for syntax errors during
1798 backslash processing.
1799
1800 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1801 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1802
1803 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1804 caused an overrun.
1805
1806 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1807 something other than just ASCII characters) inside a group that had an
1808 unlimited repeat caused a loop at compile time (while checking to see
1809 whether the group could match an empty string).
1810
1811 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1812 [\P{Any}] did so. (Error in the code for printing property names.)
1813
1814 21. An orphan \E inside a character class could cause a crash.
1815
1816 22. A repeated capturing bracket such as (A)? could cause a wild memory
1817 reference during compilation.
1818
1819 23. There are several functions in pcre_compile() that scan along a compiled
1820 expression for various reasons (e.g. to see if it's fixed length for look
1821 behind). There were bugs in these functions when a repeated \p or \P was
1822 present in the pattern. These operators have additional parameters compared
1823 with \d, etc, and these were not being taken into account when moving along
1824 the compiled data. Specifically:
1825
1826 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1827 length.
1828
1829 (b) An item such as \pL+ within a repeated group could cause crashes or
1830 loops.
1831
1832 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1833 "reference to non-existent subpattern" error.
1834
1835 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1836
1837 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1838 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1839
1840 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1841
1842 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1843 character were causing crashes (broken optimization).
1844
1845 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1846 \p or \P) caused a compile-time loop.
1847
1848 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1849 break. For example, the unstudied pattern [\r\n]A does not match the string
1850 "\r\nA" because change 7.0/46 below moves the current point on by two
1851 characters after failing to match at the start. However, the pattern \nA
1852 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1853 the same is true. There doesn't seem any very clean way out of this, but
1854 what I have chosen to do makes the common cases work: PCRE now takes note
1855 of whether there can be an explicit match for \r or \n anywhere in the
1856 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1857 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1858 pattern has explicit CR or LF references.
1859
1860 29. Added (*CR) etc for changing newline setting at start of pattern.
1861
1862
1863 Version 7.2 19-Jun-07
1864 ---------------------
1865
1866 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1867 which is apparently normally available under Windows.
1868
1869 2. Re-jig the pcregrep tests with different newline settings in an attempt
1870 to make them independent of the local environment's newline setting.
1871
1872 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1873
1874 4. Some of the "internals" tests were previously cut out when the link size
1875 was not 2, because the output contained actual offsets. The recent new
1876 "Z" feature of pcretest means that these can be cut out, making the tests
1877 usable with all link sizes.
1878
1879 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1880 stack recursion. This gives a massive performance boost under BSD, but just
1881 a small improvement under Linux. However, it saves one field in the frame
1882 in all cases.
1883
1884 6. Added more features from the forthcoming Perl 5.10:
1885
1886 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1887 recursion call. It refers to the nth most recently opened parentheses.
1888
1889 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1890 to be opened parentheses.
1891
1892 (c) Conditions that refer to capturing parentheses can be specified
1893 relatively, for example, (?(-2)... or (?(+3)...
1894
1895 (d) \K resets the start of the current match so that everything before
1896 is not part of it.
1897
1898 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1899
1900 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1901 reference syntax.
1902
1903 (g) (?| introduces a group in which the numbering of parentheses in each
1904 alternative starts with the same number.
1905
1906 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1907
1908 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1909 PCRE_INFO_JCHANGED.
1910
1911 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1912 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1913 for detecting groups that can match an empty string.
1914
1915 9. A pattern with a very large number of alternatives (more than several
1916 hundred) was running out of internal workspace during the pre-compile
1917 phase, where pcre_compile() figures out how much memory will be needed. A
1918 bit of new cunning has reduced the workspace needed for groups with
1919 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1920 workspace instead of running out of the 4096 that are available.
1921
1922 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1923
1924 11. Applied patch from Google to remove an optimization that didn't quite work.
1925 The report of the bug said:
1926
1927 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1928 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1929 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1930
1931 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1932 it matched the wrong number of bytes.
1933
1934
1935 Version 7.1 24-Apr-07
1936 ---------------------
1937
1938 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1939 that is more "standard", making use of automake and other Autotools. There
1940 is some re-arrangement of the files and adjustment of comments consequent
1941 on this.
1942
1943 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1944 for recursive directory scanning broke on some systems because the files
1945 are not scanned in any specific order and on different systems the order
1946 was different. A call to "sort" has been inserted into RunGrepTest for the
1947 approprate test as a short-term fix. In the longer term there may be an
1948 alternative.
1949
1950 3. I had an email from Eric Raymond about problems translating some of PCRE's
1951 man pages to HTML (despite the fact that I distribute HTML pages, some
1952 people do their own conversions for various reasons). The problems
1953 concerned the use of low-level troff macros .br and .in. I have therefore
1954 removed all such uses from the man pages (some were redundant, some could
1955 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1956 HTML has been updated to handle .nf/.fi and to complain if it encounters
1957 .br or .in.
1958
1959 4. Updated comments in configure.ac that get placed in config.h.in and also
1960 arranged for config.h to be included in the distribution, with the name
1961 config.h.generic, for the benefit of those who have to compile without
1962 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1963
1964 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1965 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1966 makevp.bat for the current PCRE, using the additional files
1967 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1968
1969 6. A Windows user reported a minor discrepancy with test 2, which turned out
1970 to be caused by a trailing space on an input line that had got lost in his
1971 copy. The trailing space was an accident, so I've just removed it.
1972
1973 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1974 that is needed.
1975
1976 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1977 as "const" (a) because they are and (b) because it helps the PHP
1978 maintainers who have recently made a script to detect big data structures
1979 in the php code that should be moved to the .rodata section. I remembered
1980 to update Builducptable as well, so it won't revert if ucptable.h is ever
1981 re-created.
1982
1983 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1984 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1985 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1986 support is not required. This saves 1.5-2K of code, which is important in
1987 some applications.
1988
1989 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1990 so as not to refer to the tables, even though these functions will never be
1991 called when UTF-8 support is disabled. Otherwise there are problems with a
1992 shared library.
1993
1994 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1995
1996 (a) It was defining its arguments as char * instead of void *.
1997
1998 (b) It was assuming that all moves were upwards in memory; this was true
1999 a long time ago when I wrote it, but is no longer the case.
2000
2001 The emulated memove() is provided for those environments that have neither
2002 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2003 is clearly not the case, as these two bugs were recently reported.
2004
2005 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2006 and Detrail to create the HTML documentation, the .txt form of the man
2007 pages, and it removes trailing spaces from listed files. It also creates
2008 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2009 case, it wraps all the #defines with #ifndefs. This script should be run
2010 before "make dist".
2011
2012 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2013 with Unicode property support.
2014
2015 (a) For a maximizing quantifier, if the two different cases of the
2016 character were of different lengths in their UTF-8 codings (there are
2017 some cases like this - I found 11), and the matching function had to
2018 back up over a mixture of the two cases, it incorrectly assumed they
2019 were both the same length.
2020
2021 (b) When PCRE was configured to use the heap rather than the stack for
2022 recursion during matching, it was not correctly preserving the data for
2023 the other case of a UTF-8 character when checking ahead for a match
2024 while processing a minimizing repeat. If the check also involved
2025 matching a wide character, but failed, corruption could cause an
2026 erroneous result when trying to check for a repeat of the original
2027 character.
2028
2029 13. Some tidying changes to the testing mechanism:
2030
2031 (a) The RunTest script now detects the internal link size and whether there
2032 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2033 values substituted by "configure". (The RunGrepTest script already did
2034 this for UTF-8.) The configure.ac script no longer substitutes the
2035 relevant variables.
2036
2037 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2038 with length and offset values. This means that the output is different
2039 for different internal link sizes. Test 2 is skipped for link sizes
2040 other than 2 because of this, bypassing the problem. Unfortunately,
2041 there was also a test in test 3 (the locale tests) that used /B and
2042 failed for link sizes other than 2. Rather than cut the whole test out,
2043 I have added a new /Z option to pcretest that replaces the length and
2044 offset values with spaces. This is now used to make test 3 independent
2045 of link size. (Test 2 will be tidied up later.)
2046
2047 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2048 segmentation fault instead of returning the appropriate error message.
2049
2050 15. In multiline mode when the newline sequence was set to "any", the pattern
2051 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2052 This doesn't seem right; it now treats the CRLF combination as the line
2053 ending, and so does not match in that case. It's only a pattern such as ^$
2054 that would hit this one: something like ^ABC$ would have failed after \r
2055 and then tried again after \r\n.
2056
2057 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2058 in an attempt to make files that differ only in their line terminators
2059 compare equal. This works on Linux.
2060
2061 17. Under certain error circumstances pcregrep might try to free random memory
2062 as it exited. This is now fixed, thanks to valgrind.
2063
2064 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2065 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2066 was because its rules for how to advance for /g after matching an empty
2067 string at the end of a line did not allow for this case. They now check for
2068 it specially.
2069
2070 20. pcretest is supposed to handle patterns and data of any length, by
2071 extending its buffers when necessary. It was getting this wrong when the
2072 buffer for a data line had to be extended.
2073
2074 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2075 CRLF as a newline sequence.
2076
2077 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2078 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2079 I have nevertheless tidied it up.
2080
2081 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2082
2083 24. Added a man page for pcre-config.
2084
2085
2086 Version 7.0 19-Dec-06
2087 ---------------------
2088
2089 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2090 moving to gcc 4.1.1.
2091
2092 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2093 sys/time.h, which is documented as needed for this function. It doesn't
2094 seem to matter on Linux, but it showed up on some releases of OS X.
2095
2096 3. It seems that there are systems where bytes whose values are greater than
2097 127 match isprint() in the "C" locale. The "C" locale should be the
2098 default when a C program starts up. In most systems, only ASCII printing
2099 characters match isprint(). This difference caused the output from pcretest
2100 to vary, making some of the tests fail. I have changed pcretest so that:
2101
2102 (a) When it is outputting text in the compiled version of a pattern, bytes
2103 other than 32-126 are always shown as hex escapes.
2104
2105 (b) When it is outputting text that is a matched part of a subject string,
2106 it does the same, unless a different locale has been set for the match
2107 (using the /L modifier). In this case, it uses isprint() to decide.
2108
2109 4. Fixed a major bug that caused incorrect computation of the amount of memory
2110 required for a compiled pattern when options that changed within the
2111 pattern affected the logic of the preliminary scan that determines the
2112 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2113 that the computed length was too small. The symptoms of this bug were
2114 either the PCRE error "internal error: code overflow" from pcre_compile(),
2115 or a glibc crash with a message such as "pcretest: free(): invalid next
2116 size (fast)". Examples of patterns that provoked this bug (shown in
2117 pcretest format) are:
2118
2119 /(?-x: )/x
2120 /(?x)(?-x: \s*#\s*)/
2121 /((?i)[\x{c0}])/8
2122 /(?i:[\x{c0}])/8
2123
2124 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2125 is now done differently.
2126
2127 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2128 wrapper classes; (b) implement a new function in the C++ scanner that is
2129 more efficient than the old way of doing things because it avoids levels of
2130 recursion in the regex matching; (c) add a paragraph to the documentation
2131 for the FullMatch() function.
2132
2133 6. The escape sequence \n was being treated as whatever was defined as
2134 "newline". Not only was this contrary to the documentation, which states
2135 that \n is character 10 (hex 0A), but it also went horribly wrong when
2136 "newline" was defined as CRLF. This has been fixed.
2137
2138 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2139 was being set to -1 for the "end of line" case (supposedly a value that no
2140 character can have). Though this value is never used (the check for end of
2141 line is "zero bytes in current character"), it caused compiler complaints.
2142 I've changed it to 0xffffffff.
2143
2144 8. In pcre_version.c, the version string was being built by a sequence of
2145 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2146 string (as it is for production releases) called a macro with an empty
2147 argument. The C standard says the result of this is undefined. The gcc
2148 compiler treats it as an empty string (which was what was wanted) but it is
2149 reported that Visual C gives an error. The source has been hacked around to
2150 avoid this problem.
2151
2152 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2153 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2154 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2155 of them did).
2156
2157 10. Originally, pcretest opened its input and output without "b"; then I was
2158 told that "b" was needed in some environments, so it was added for release
2159 5.0 to both the input and output. (It makes no difference on Unix-like
2160 systems.) Later I was told that it is wrong for the input on Windows. I've
2161 now abstracted the modes into two macros, to make it easier to fiddle with
2162 them, and removed "b" from the input mode under Windows.
2163
2164 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2165
2166 12. Added -help and --help to pcretest as an official way of being reminded
2167 of the options.
2168
2169 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2170 and pcrecpp.cc because they annoy compilers at high warning levels.
2171
2172 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2173
2174 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2175 scripts are not C programs :-) and which was not noticed because it works
2176 on Linux.
2177
2178 16. pcretest is supposed to handle any length of pattern and data line (as one
2179 line or as a continued sequence of lines) by extending its input buffer if
2180 necessary. This feature was broken for very long pattern lines, leading to
2181 a string of junk being passed to pcre_compile() if the pattern was longer
2182 than about 50K.
2183
2184 17. I have done a major re-factoring of the way pcre_compile() computes the
2185 amount of memory needed for a compiled pattern. Previously, there was code
2186 that made a preliminary scan of the pattern in order to do this. That was
2187 OK when PCRE was new, but as the facilities have expanded, it has become
2188 harder and harder to keep it in step with the real compile phase, and there
2189 have been a number of bugs (see for example, 4 above). I have now found a
2190 cunning way of running the real compile function in a "fake" mode that
2191 enables it to compute how much memory it would need, while actually only
2192 ever using a few hundred bytes of working memory and without too many
2193 tests of the mode. This should make future maintenance and development
2194 easier. A side effect of this work is that the limit of 200 on the nesting
2195 depth of parentheses has been removed (though this was never a serious
2196 limitation, I suspect). However, there is a downside: pcre_compile() now
2197 runs more slowly than before (30% or more, depending on the pattern). I
2198 hope this isn't a big issue. There is no effect on runtime performance.
2199
2200 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2201 newline (only possible for the last line of a file) and it was a
2202 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2203
2204 19. Added additional timing features to pcretest. (1) The -tm option now times
2205 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2206 separate command line item, by a number that specifies the number of
2207 repeats to use when timing. The default is 50000; this gives better
2208 precision, but takes uncomfortably long for very large patterns.
2209
2210 20. Extended pcre_study() to be more clever in cases where a branch of a
2211 subpattern has no definite first character. For example, (a*|b*)[cd] would
2212 previously give no result from pcre_study(). Now it recognizes that the
2213 first character must be a, b, c, or d.
2214
2215 21. There was an incorrect error "recursive call could loop indefinitely" if
2216 a subpattern (or the entire pattern) that was being tested for matching an
2217 empty string contained only one non-empty item after a nested subpattern.
2218 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2219 incorrectly, because the \d was being skipped in the check.
2220
2221 22. The pcretest program now has a new pattern option /B and a command line
2222 option -b, which is equivalent to adding /B to every pattern. This causes
2223 it to show the compiled bytecode, without the additional information that
2224 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2225 is the same as /B/I).
2226
2227 23. A new optimization is now able automatically to treat some sequences such
2228 as a*b as a*+b. More specifically, if something simple (such as a character
2229 or a simple class like \d) has an unlimited quantifier, and is followed by
2230 something that cannot possibly match the quantified thing, the quantifier
2231 is automatically "possessified".
2232
2233 24. A recursive reference to a subpattern whose number was greater than 39
2234 went wrong under certain circumstances in UTF-8 mode. This bug could also
2235 have affected the operation of pcre_study().
2236
2237 25. Realized that a little bit of performance could be had by replacing
2238 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2239
2240 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2241
2242 27. Possessive quantifiers such as a++ were previously implemented by turning
2243 them into atomic groups such as ($>a+). Now they have their own opcodes,
2244 which improves performance. This includes the automatically created ones
2245 from 23 above.
2246
2247 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2248 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2249 the first matched character to be a colon. This applied both to named and
2250 numbered groups.
2251
2252 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2253
2254 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2255 building PCRE on Windows easier, so I have included it in the distribution.
2256
2257 31. There is now a check in pcretest against a ridiculously large number being
2258 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2259 loop, the loop is abandoned.
2260
2261 32. Forward references to subpatterns in conditions such as (?(2)...) where
2262 subpattern 2 is defined later cause pcre_compile() to search forwards in
2263 the pattern for the relevant set of parentheses. This search went wrong
2264 when there were unescaped parentheses in a character class, parentheses
2265 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2266
2267 33. "Subroutine" calls and backreferences were previously restricted to
2268 referencing subpatterns earlier in the regex. This restriction has now
2269 been removed.
2270
2271 34. Added a number of extra features that are going to be in Perl 5.10. On the
2272 whole, these are just syntactic alternatives for features that PCRE had
2273 previously implemented using the Python syntax or my own invention. The
2274 other formats are all retained for compatibility.
2275
2276 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2277 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2278 also .NET compatible.
2279
2280 (b) A recursion or subroutine call to a named group can now be defined as
2281 (?&name) as well as (?P>name).
2282
2283 (c) A backreference to a named group can now be defined as \k<name> or
2284 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2285 5.10, are also .NET compatible.
2286
2287 (d) A conditional reference to a named group can now use the syntax
2288 (?(<name>) or (?('name') as well as (?(name).
2289
2290 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2291 groups (named and numbered) that are never evaluated inline, but can be
2292 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2293 is always false. There may be only one alternative in such a group.
2294
2295 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2296 as the simple (?(R). The condition is true only if the most recent
2297 recursion is that of the given number or name. It does not search out
2298 through the entire recursion stack.
2299
2300 (g) The escape \gN or \g{N} has been added, where N is a positive or
2301 negative number, specifying an absolute or relative reference.
2302
2303 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2304 some "unreachable code" warnings.
2305
2306 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2307 things, this adds five new scripts.
2308
2309 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2310 There were also incompatibilities regarding the handling of \Q..\E inside
2311 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2312 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2313
2314 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2315 matches an empty string, and forcibly breaks the loop. There were bugs in
2316 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2317 against aaaa the result was just "a" rather than "aaaa", for example. Two
2318 separate and independent bugs (that affected different cases) have been
2319 fixed.
2320
2321 39. Refactored the code to abolish the use of different opcodes for small
2322 capturing bracket numbers. This is a tidy that I avoided doing when I
2323 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2324 The new approach is not only tidier, it makes it possible to reduce the
2325 memory needed to fix the previous bug (38).
2326
2327 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2328 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2329 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2330 mode.
2331
2332 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2333 report.
2334
2335 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2336 copy construction and assignment in the C++ wrapper.
2337
2338 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2339 couple of bugs that could have given wrong results in the "--newline=crlf"
2340 case.
2341
2342 44. Added a number of casts and did some reorganization of signed/unsigned int
2343 variables following suggestions from Dair Grant. Also renamed the variable
2344 "this" as "item" because it is a C++ keyword.
2345
2346 45. Arranged for dftables to add
2347
2348 #include "pcre_internal.h"
2349
2350 to pcre_chartables.c because without it, gcc 4.x may remove the array
2351 definition from the final binary if PCRE is built into a static library and
2352 dead code stripping is activated.
2353
2354 46. For an unanchored pattern, if a match attempt fails at the start of a
2355 newline sequence, and the newline setting is CRLF or ANY, and the next two
2356 characters are CRLF, advance by two characters instead of one.
2357
2358
2359 Version 6.7 04-Jul-06
2360 ---------------------
2361
2362 1. In order to handle tests when input lines are enormously long, pcretest has
2363 been re-factored so that it automatically extends its buffers when
2364 necessary. The code is crude, but this _is_ just a test program. The
2365 default size has been increased from 32K to 50K.
2366
2367 2. The code in pcre_study() was using the value of the re argument before
2368 testing it for NULL. (Of course, in any sensible call of the function, it
2369 won't be NULL.)
2370
2371 3. The memmove() emulation function in pcre_internal.h, which is used on
2372 systems that lack both memmove() and bcopy() - that is, hardly ever -
2373 was missing a "static" storage class specifier.
2374
2375 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2376 containing an extended class (one that cannot be represented by a bitmap
2377 because it contains high-valued characters or Unicode property items, e.g.
2378 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2379 pattern, but PCRE should not loop if you do not (it no longer does).
2380 [Detail: two cases were found: (a) a repeated subpattern containing an
2381 extended class; (b) a recursive reference to a subpattern that followed a
2382 previous extended class. It wasn't skipping over the extended class
2383 correctly when UTF-8 mode was not set.]
2384
2385 5. A negated single-character class was not being recognized as fixed-length
2386 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2387 compile error "lookbehind assertion is not fixed length".
2388
2389 6. The RunPerlTest auxiliary script was showing an unexpected difference
2390 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2391 write a Perl script that can interpret lines of an input file either as
2392 byte characters or as UTF-8, which is what "perltest" was being required to
2393 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2394 can't do is switch easily at run time between having the "use utf8;" pragma
2395 or not. In the end, I fudged it by using the RunPerlTest script to insert
2396 "use utf8;" explicitly for the UTF-8 tests.
2397
2398 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2399 the end of the subject string, contrary to the documentation and to what
2400 Perl does. This was true of both matching functions. Now it matches only at
2401 the start of the subject and immediately after *internal* newlines.
2402
2403 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2404 a pointer to an int instead of a pointer to an unsigned long int. This
2405 caused problems on 64-bit systems.
2406
2407 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2408 instance of the 'standard' template library not being so standard".
2409
2410 10. There was no check on the number of named subpatterns nor the maximum
2411 length of a subpattern name. The product of these values is used to compute
2412 the size of the memory block for a compiled pattern. By supplying a very
2413 long subpattern name and a large number of named subpatterns, the size
2414 computation could be caused to overflow. This is now prevented by limiting
2415 the length of names to 32 characters, and the number of named subpatterns
2416 to 10,000.
2417
2418 11. Subpatterns that are repeated with specific counts have to be replicated in
2419 the compiled pattern. The size of memory for this was computed from the
2420 length of the subpattern and the repeat count. The latter is limited to
2421 65535, but there was no limit on the former, meaning that integer overflow
2422 could in principle occur. The compiled length of a repeated subpattern is
2423 now limited to 30,000 bytes in order to prevent this.
2424
2425 12. Added the optional facility to have named substrings with the same name.
2426
2427 13. Added the ability to use a named substring as a condition, using the
2428 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2429 are numbers (not recommended). Forward references are permitted.
2430
2431 14. Added forward references in named backreferences (if you see what I mean).
2432
2433 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2434 pattern could run off the end of the subject. For example, the pattern
2435 "(?s)(.{1,5})"8 did this with the subject "ab".
2436
2437 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2438 PCRE_CASELESS was set when matching characters that were quantified with ?
2439 or *.
2440
2441 17. A character class other than a single negated character that had a minimum
2442 but no maximum quantifier - for example [ab]{6,} - was not handled
2443 correctly by pce_dfa_exec(). It would match only one character.
2444
2445 18. A valid (though odd) pattern that looked like a POSIX character
2446 class but used an invalid character after [ (for example [[,abc,]]) caused
2447 pcre_compile() to give the error "Failed: internal error: code overflow" or
2448 in some cases to crash with a glibc free() error. This could even happen if
2449 the pattern terminated after [[ but there just happened to be a sequence of
2450 letters, a binary zero, and a closing ] in the memory that followed.
2451
2452 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2453 over the years. Originally (before any Unicode support), just the bottom 8
2454 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2455 output from "man perlunicode" includes this:
2456
2457 The regular expression compiler produces polymorphic opcodes. That
2458 is, the pattern adapts to the data and automatically switches to
2459 the Unicode character scheme when presented with Unicode data--or
2460 instead uses a traditional byte scheme when presented with byte
2461 data.
2462
2463 Sadly, a wide octal escape does not cause a switch, and in a string with
2464 no other multibyte characters, these octal escapes are treated as before.
2465 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2466 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2467 Unicode string.
2468
2469 I have not perpetrated such confusion in PCRE. Up till now, it took just
2470 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2471 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2472 translate to the appropriate multibyte character.
2473
2474 29. Applied some refactoring to reduce the number of warnings from Microsoft
2475 and Borland compilers. This has included removing the fudge introduced
2476 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2477 a warning about an unused variable.
2478
2479 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2480 characters since release 4.0, because Perl (from release 5.004) does not.
2481 [Or at least, is documented not to: some releases seem to be in conflict
2482 with the documentation.] However, when a pattern was studied with
2483 pcre_study() and all its branches started with \s, PCRE still included VT
2484 as a possible starting character. Of course, this did no harm; it just
2485 caused an unnecessary match attempt.
2486
2487 22. Removed a now-redundant internal flag bit that recorded the fact that case
2488 dependency changed within the pattern. This was once needed for "required
2489 byte" processing, but is no longer used. This recovers a now-scarce options
2490 bit. Also moved the least significant internal flag bit to the most-
2491 significant bit of the word, which was not previously used (hangover from
2492 the days when it was an int rather than a uint) to free up another bit for
2493 the future.
2494
2495 23. Added support for CRLF line endings as well as CR and LF. As well as the
2496 default being selectable at build time, it can now be changed at runtime
2497 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2498 specify that it is scanning data with non-default line endings.
2499
2500 24. Changed the definition of CXXLINK to make it agree with the definition of
2501 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2502
2503 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2504 recursions. This makes a big different to stack usage for some patterns.
2505
2506 26. If a subpattern containing a named recursion or subroutine reference such
2507 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2508 the space required for the compiled pattern went wrong and gave too small a
2509 value. Depending on the environment, this could lead to "Failed: internal
2510 error: code overflow at offset 49" or "glibc detected double free or
2511 corruption" errors.
2512
2513 27. Applied patches from Google (a) to support the new newline modes and (b) to
2514 advance over multibyte UTF-8 characters in GlobalReplace.
2515
2516 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2517 difference for some implementation of PCRE in some Windows version.
2518
2519 29. Added some extra testing facilities to pcretest:
2520
2521 \q<number> in a data line sets the "match limit" value
2522 \Q<number> in a data line sets the "match recursion limt" value
2523 -S <number> sets the stack size, where <number> is in megabytes
2524
2525 The -S option isn't available for Windows.
2526
2527
2528 Version 6.6 06-Feb-06
2529 ---------------------
2530
2531 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2532 in pcreposix.h. I have copied the definition from pcre.h.
2533
2534 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2535 because pcre.h is no longer a built file.
2536
2537 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2538 not normally included in the compiled code.
2539
2540
2541 Version 6.5 01-Feb-06
2542 ---------------------
2543
2544 1. When using the partial match feature with pcre_dfa_exec(), it was not
2545 anchoring the second and subsequent partial matches at the new starting
2546 point. This could lead to incorrect results. For example, with the pattern
2547 /1234/, partially matching against "123" and then "a4" gave a match.
2548
2549 2. Changes to pcregrep:
2550
2551 (a) All non-match returns from pcre_exec() were being treated as failures
2552 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2553 error message is output. Some extra information is given for the
2554 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2555 probably the only errors that are likely to be caused by users (by
2556 specifying a regex that has nested indefinite repeats, for instance).
2557 If there are more than 20 of these errors, pcregrep is abandoned.
2558
2559 (b) A binary zero was treated as data while matching, but terminated the
2560 output line if it was written out. This has been fixed: binary zeroes
2561 are now no different to any other data bytes.
2562
2563 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2564 used to set a locale for matching. The --locale=xxxx long option has
2565 been added (no short equivalent) to specify a locale explicitly on the
2566 pcregrep command, overriding the environment variables.
2567
2568 (d) When -B was used with -n, some line numbers in the output were one less
2569 than they should have been.
2570
2571 (e) Added the -o (--only-matching) option.
2572
2573 (f) If -A or -C was used with -c (count only), some lines of context were
2574 accidentally printed for the final match.
2575
2576 (g) Added the -H (--with-filename) option.
2577
2578 (h) The combination of options -rh failed to suppress file names for files
2579 that were found from directory arguments.
2580
2581 (i) Added the -D (--devices) and -d (--directories) options.
2582
2583 (j) Added the -F (--fixed-strings) option.
2584
2585 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2586
2587 (l) Added the --colo(u)r option.
2588
2589 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2590 is not present by default.
2591
2592 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2593 items such as (?R) or (?1), when the recursion could match a number of
2594 alternatives. If it matched one of the alternatives, but subsequently,
2595 outside the recursion, there was a failure, the code tried to back up into
2596 the recursion. However, because of the way PCRE is implemented, this is not
2597 possible, and the result was an incorrect result from the match.
2598
2599 In order to prevent this happening, the specification of recursion has
2600 been changed so that all such subpatterns are automatically treated as
2601 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2602
2603 4. I had overlooked the fact that, in some locales, there are characters for
2604 which isalpha() is true but neither isupper() nor islower() are true. In
2605 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2606 and ordfeminine) are like this. This affected the treatment of \w and \W
2607 when they appeared in character classes, but not when they appeared outside
2608 a character class. The bit map for "word" characters is now created
2609 separately from the results of isalnum() instead of just taking it from the
2610 upper, lower, and digit maps. (Plus the underscore character, of course.)
2611
2612 5. The above bug also affected the handling of POSIX character classes such as
2613 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2614 permanent tables. Instead, the bit maps for such a class were previously
2615 created as the appropriate unions of the upper, lower, and digit bitmaps.
2616 Now they are created by subtraction from the [[:word:]] class, which has
2617 its own bitmap.
2618
2619 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2620 It is created by subtracting the vertical space characters (\x09, \x0a,
2621 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2622 subtraction was done in the overall bitmap for a character class, meaning
2623 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2624 be recognized. This bug has been fixed.
2625
2626 7. Patches from the folks at Google:
2627
2628 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2629 real life, but is still worth protecting against".
2630
2631 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2632 regular expressions".
2633
2634 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2635 have it.
2636
2637 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2638 "configure" and the latter not, in order to fix a problem somebody had
2639 with compiling the Arg class on HP-UX.
2640
2641 (e) Improve the error-handling of the C++ wrapper a little bit.
2642
2643 (f) New tests for checking recursion limiting.
2644
2645 8. The pcre_memmove() function, which is used only if the environment does not
2646 have a standard memmove() function (and is therefore rarely compiled),
2647 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2648 returning a result (though PCRE never actually uses the result).
2649
2650 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2651 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2652 returned instead of calling malloc() with an overflowing number that would
2653 most likely cause subsequent chaos.
2654
2655 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2656
2657 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2658 with this option is matched, the nmatch and pmatch options of regexec() are
2659 ignored.
2660
2661 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2662 provided in case anyone wants to the the POSIX interface with UTF-8
2663 strings.
2664
2665 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2666 C++ linking (needed for some HP-UX environments).
2667
2668 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2669 (unused parameter) and in the pcre_printint() function (omitted "default"
2670 switch label when the default is to do nothing).
2671
2672 15. Added some code to make it possible, when PCRE is compiled as a C++
2673 library, to replace subject pointers for pcre_exec() with a smart pointer
2674 class, thus making it possible to process discontinuous strings.
2675
2676 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2677 much the same function. They were added by different people who were trying
2678 to make PCRE easy to compile on non-Unix systems. It has been suggested
2679 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2680 for compiling on Windows systems. I have therefore replaced it with
2681 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2682 defaults to "extern" for C or "extern C" for C++, which works fine on
2683 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2684 SCOPE with something explicit in config.h. In addition:
2685
2686 (a) pcreposix.h still had just "extern" instead of either of these macros;
2687 I have replaced it with PCRE_DATA_SCOPE.
2688
2689 (b) Functions such as _pcre_xclass(), which are internal to the library,
2690 but external in the C sense, all had PCRE_EXPORT in their definitions.
2691 This is apparently wrong for the Windows case, so I have removed it.
2692 (It makes no difference on Unix-like systems.)
2693
2694 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2695 of recursive calls to match(). This is different to MATCH_LIMIT because
2696 that limits the total number of calls to match(), not all of which increase
2697 the depth of recursion. Limiting the recursion depth limits the amount of
2698 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2699 when PCRE is compiled, and changed at run time. A patch from Google adds
2700 this functionality to the C++ interface.
2701
2702 18. Changes to the handling of Unicode character properties:
2703
2704 (a) Updated the table to Unicode 4.1.0.
2705
2706 (b) Recognize characters that are not in the table as "Cn" (undefined).
2707
2708 (c) I revised the way the table is implemented to a much improved format
2709 which includes recognition of ranges. It now supports the ranges that
2710 are defined in UnicodeData.txt, and it also amalgamates other
2711 characters into ranges. This has reduced the number of entries in the
2712 table from around 16,000 to around 3,000, thus reducing its size
2713 considerably. I realized I did not need to use a tree structure after
2714 all - a binary chop search is just as efficient. Having reduced the
2715 number of entries, I extended their size from 6 bytes to 8 bytes to
2716 allow for more data.
2717
2718 (d) Added support for Unicode script names via properties such as \p{Han}.
2719
2720 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2721 matching that character.
2722
2723 20. When matching a repeated Unicode property with a minimum greater than zero,
2724 (for example \pL{2,}), PCRE could look past the end of the subject if it
2725 reached it while seeking the minimum number of characters. This could
2726 happen only if some of the characters were more than one byte long, because
2727 there is a check for at least the minimum number of bytes.
2728
2729 21. Refactored the implementation of \p and \P so as to be more general, to
2730 allow for more different types of property in future. This has changed the
2731 compiled form incompatibly. Anybody with saved compiled patterns that use
2732 \p or \P will have to recompile them.
2733
2734 22. Added "Any" and "L&" to the supported property types.
2735
2736 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2737 but give a compile time error if the value is greater than 0xff.
2738
2739 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2740 accidentally not being installed or uninstalled.
2741
2742 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2743 made were to insert the current release number. This seemed silly, because
2744 it made things harder for people building PCRE on systems that don't run
2745 "configure". I have turned pcre.h into a distributed file, no longer built
2746 by "configure", with the version identification directly included. There is
2747 no longer a pcre.h.in file.
2748
2749 However, this change necessitated a change to the pcre-config script as
2750 well. It is built from pcre-config.in, and one of the substitutions was the
2751 release number. I have updated configure.ac so that ./configure now finds
2752 the release number by grepping pcre.h.
2753
2754 26. Added the ability to run the tests under valgrind.
2755
2756
2757 Version 6.4 05-Sep-05
2758 ---------------------
2759
2760 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2761 "--" to be printed when multiple files were scanned, even when none of the
2762 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2763 consider it to be a bug, and have restored the previous behaviour.
2764
2765 2. A couple of code tidies to get rid of compiler warnings.
2766
2767 3. The pcretest program used to cheat by referring to symbols in the library
2768 whose names begin with _pcre_. These are internal symbols that are not
2769 really supposed to be visible externally, and in some environments it is
2770 possible to suppress them. The cheating is now confined to including
2771 certain files from the library's source, which is a bit cleaner.
2772
2773 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2774 file's purpose clearer.
2775
2776 5. Reorganized pcre_ucp_findchar().
2777
2778
2779 Version 6.3 15-Aug-05
2780 ---------------------
2781
2782 1. The file libpcre.pc.in did not have general read permission in the tarball.
2783
2784 2. There were some problems when building without C++ support:
2785
2786 (a) If C++ support was not built, "make install" and "make test" still
2787 tried to test it.
2788
2789 (b) There were problems when the value of CXX was explicitly set. Some
2790 changes have been made to try to fix these, and ...
2791
2792 (c) --disable-cpp can now be used to explicitly disable C++ support.
2793
2794 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2795 backslash in a target when C++ was disabled. This confuses some
2796 versions of "make", apparently. Using an intermediate variable solves
2797 this. (Same for CPP_LOBJ.)
2798
2799 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2800 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2801 necessary on certain architectures.
2802
2803 4. Added a setting of -export-symbols-regex to the link command to remove
2804 those symbols that are exported in the C sense, but actually are local
2805 within the library, and not documented. Their names all begin with
2806 "_pcre_". This is not a perfect job, because (a) we have to except some
2807 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2808 available (and never for static libraries). I have made a note to try to
2809 find a way round (a) in the future.
2810
2811
2812 Version 6.2 01-Aug-05
2813 ---------------------
2814
2815 1. There was no test for integer overflow of quantifier values. A construction
2816 such as {1111111111111111} would give undefined results. What is worse, if
2817 a minimum quantifier for a parenthesized subpattern overflowed and became
2818 negative, the calculation of the memory size went wrong. This could have
2819 led to memory overwriting.
2820
2821 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2822
2823 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2824 operating environments where this matters.
2825
2826 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2827 PCRE options from within the C++ wrapper.
2828
2829 5. Named capturing subpatterns were not being correctly counted when a pattern
2830 was compiled. This caused two problems: (a) If there were more than 100
2831 such subpatterns, the calculation of the memory needed for the whole
2832 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2833 back references of the form \12, where the number was greater than 9, were
2834 not recognized as back references, even though there were sufficient
2835 previous subpatterns.
2836
2837 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2838 versions of gcc, e.g. 2.95.4.
2839
2840
2841 Version 6.1 21-Jun-05
2842 ---------------------
2843
2844 1. There was one reference to the variable "posix" in pcretest.c that was not
2845 surrounded by "#if !defined NOPOSIX".
2846
2847 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2848 the cross-check on the old pcre_info() function, for the benefit of the
2849 cut-down version of PCRE that is currently imported into Exim.
2850
2851 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2852 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2853 patterns that start (?i) but I don't think that matters. The use of (?i) is
2854 just an example; this all applies to the other options as well.
2855
2856 4. Since libtool seems to echo the compile commands it is issuing, the output
2857 from "make" can be reduced a bit by putting "@" in front of each libtool
2858 compile command.
2859
2860 5. Patch from the folks at Google for configure.in to be a bit more thorough
2861 in checking for a suitable C++ installation before trying to compile the
2862 C++ stuff. This should fix a reported problem when a compiler was present,
2863 but no suitable headers.
2864
2865 6. The man pages all had just "PCRE" as their title. I have changed them to
2866 be the relevant file name. I have also arranged that these names are
2867 retained in the file doc/pcre.txt, which is a concatenation in text format
2868 of all the man pages except the little individual ones for each function.
2869
2870 7. The NON-UNIX-USE file had not been updated for the different set of source
2871 files that come with release 6. I also added a few comments about the C++
2872 wrapper.
2873
2874
2875 Version 6.0 07-Jun-05
2876 ---------------------
2877
2878 1. Some minor internal re-organization to help with my DFA experiments.
2879
2880 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2881 didn't matter for the library itself when fully configured, but did matter
2882 when compiling without UCP support, or within Exim, where the ucp files are
2883 not imported.
2884
2885 3. Refactoring of the library code to split up the various functions into
2886 different source modules. The addition of the new DFA matching code (see
2887 below) to a single monolithic source would have made it really too
2888 unwieldy, quite apart from causing all the code to be include in a
2889 statically linked application, when only some functions are used. This is
2890 relevant even without the DFA addition now that patterns can be compiled in
2891 one application and matched in another.
2892
2893 The downside of splitting up is that there have to be some external
2894 functions and data tables that are used internally in different modules of
2895 the library but which are not part of the API. These have all had their
2896 names changed to start with "_pcre_" so that they are unlikely to clash
2897 with other external names.
2898
2899 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2900 a different (DFA) algorithm. Although it is slower than the original
2901 function, it does have some advantages for certain types of matching
2902 problem.
2903
2904 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2905 including restarting after a partial match.
2906
2907 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2908 defined when compiling for Windows was sent to me. I have put it into the
2909 code, though I have no means of testing or verifying it.
2910
2911 7. Added the pcre_refcount() auxiliary function.
2912
2913 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2914 match before or at the first newline in the subject string. In pcretest,
2915 the /f option on a pattern can be used to set this.
2916
2917 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2918 would behave wrongly. This has been present in PCRE since release 4.0.
2919
2920 10. A number of changes to the pcregrep command:
2921
2922 (a) Refactored how -x works; insert ^(...)$ instead of setting
2923 PCRE_ANCHORED and checking the length, in preparation for adding
2924 something similar for -w.
2925
2926 (b) Added the -w (match as a word) option.
2927
2928 (c) Refactored the way lines are read and buffered so as to have more
2929 than one at a time available.
2930
2931 (d) Implemented a pcregrep test script.
2932
2933 (e) Added the -M (multiline match) option. This allows patterns to match
2934 over several lines of the subject. The buffering ensures that at least
2935 8K, or the rest of the document (whichever is the shorter) is available
2936 for matching (and similarly the previous 8K for lookbehind assertions).
2937
2938 (f) Changed the --help output so that it now says
2939
2940 -w, --word-regex(p)
2941
2942 instead of two lines, one with "regex" and the other with "regexp"
2943 because that confused at least one person since the short forms are the
2944 same. (This required a bit of code, as the output is generated
2945 automatically from a table. It wasn't just a text change.)
2946
2947 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2948 option but starts with a hyphen. Could be a pattern or a path name
2949 starting with a hyphen, for instance.
2950
2951 (h) "-" can be given as a file name to represent stdin.
2952
2953 (i) When file names are being printed, "(standard input)" is used for
2954 the standard input, for compatibility with GNU grep. Previously
2955 "<stdin>" was used.
2956
2957 (j) The option --label=xxx can be used to supply a name to be used for
2958 stdin when file names are being printed. There is no short form.
2959
2960 (k) Re-factored the options decoding logic because we are going to add
2961 two more options that take data. Such options can now be given in four
2962 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2963
2964 (l) Added the -A, -B, and -C options for requesting that lines of context
2965 around matches be printed.
2966
2967 (m) Added the -L option to print the names of files that do not contain
2968 any matching lines, that is, the complement of -l.
2969
2970 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2971 continue to scan other files.
2972
2973 (o) The -s option was incorrectly implemented. For compatibility with other
2974 greps, it now suppresses the error message for a non-existent or non-
2975 accessible file (but not the return code). There is a new option called
2976 -q that suppresses the output of matching lines, which was what -s was
2977 previously doing.
2978
2979 (p) Added --include and --exclude options to specify files for inclusion
2980 and exclusion when recursing.
2981
2982 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2983 Hopefully, it now does.
2984
2985 12. Missing cast in pcre_study().
2986
2987 13. Added an "uninstall" target to the makefile.
2988
2989 14. Replaced "extern" in the function prototypes in Makefile.in with
2990 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2991 world, but is set differently for Windows.
2992
2993 15. Added a second compiling function called pcre_compile2(). The only
2994 difference is that it has an extra argument, which is a pointer to an
2995 integer error code. When there is a compile-time failure, this is set
2996 non-zero, in addition to the error test pointer being set to point to an
2997 error message. The new argument may be NULL if no error number is required
2998 (but then you may as well call pcre_compile(), which is now just a
2999 wrapper). This facility is provided because some applications need a
3000 numeric error indication, but it has also enabled me to tidy up the way
3001 compile-time errors are handled in the POSIX wrapper.
3002
3003 16. Added VPATH=.libs to the makefile; this should help when building with one
3004 prefix path and installing with another. (Or so I'm told by someone who
3005 knows more about this stuff than I do.)
3006
3007 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3008 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3009 match everything, including newlines. This is not POSIX-compatible, but
3010 somebody wanted the feature. From pcretest it can be activated by using
3011 both the P and the s flags.
3012
3013 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3014
3015 19. libpcre.pc was being incorrectly installed as executable.
3016
3017 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3018 it now also looks for '\r' so that it will work unmodified on Windows.
3019
3020 21. Added Google's contributed C++ wrapper to the distribution.
3021
3022 22. Added some untidy missing memory free() calls in pcretest, to keep
3023 Electric Fence happy when testing.
3024
3025
3026
3027 Version 5.0 13-Sep-04
3028 ---------------------
3029
3030 1. Internal change: literal characters are no longer packed up into items
3031 containing multiple characters in a single byte-string. Each character
3032 is now matched using a separate opcode. However, there may be more than one
3033 byte in the character in UTF-8 mode.
3034
3035 2. The pcre_callout_block structure has two new fields: pattern_position and
3036 next_item_length. These contain the offset in the pattern to the next match
3037 item, and its length, respectively.
3038
3039 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3040 insertion of callouts before each pattern item. Added the /C option to
3041 pcretest to make use of this.
3042
3043 4. On the advice of a Windows user, the lines
3044
3045 #if defined(_WIN32) || defined(WIN32)
3046 _setmode( _fileno( stdout ), 0x8000 );
3047 #endif /* defined(_WIN32) || defined(WIN32) */
3048
3049 have been added to the source of pcretest. This apparently does useful
3050 magic in relation to line terminators.
3051
3052 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3053 for the benefit of those environments where the "b" makes a difference.
3054
3055 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3056 to know about it. I have put a hack into configure.in that adds in code
3057 to set GCC=yes if CC=icc. This seems to end up at a point in the
3058 generated configure script that is early enough to affect the setting of
3059 compiler options, which is what is needed, but I have no means of testing
3060 whether it really works. (The user who reported this had patched the
3061 generated configure script, which of course I cannot do.)
3062
3063 LATER: After change 22 below (new libtool files), the configure script
3064 seems to know about icc (and also ecc). Therefore, I have commented out
3065 this hack in configure.in.
3066
3067 7. Added support for pkg-config (2 patches were sent in).
3068
3069 8. Negated POSIX character classes that used a combination of internal tables
3070 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3071 [[:^ascii]]. Typically, they would match almost any characters. The other
3072 POSIX classes were not broken in this way.
3073
3074 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3075 to find the match, as PCRE was deluded into thinking that the match had to
3076 start at the start point or following a newline. The same bug applied to
3077 patterns with negative forward assertions or any backward assertions
3078 preceding ".*" at the start, unless the pattern required a fixed first
3079 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3080
3081 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3082 starting at the last subject character, bytes beyond the end of the subject
3083 string were read.
3084
3085 11. Renamed the variable "class" as "classbits" to make life easier for C++
3086 users. (Previously there was a macro definition, but it apparently wasn't
3087 enough.)
3088
3089 12. Added the new field "tables" to the extra data so that tables can be passed
3090 in at exec time, or the internal tables can be re-selected. This allows
3091 a compiled regex to be saved and re-used at a later time by a different
3092 program that might have everything at different addresses.
3093
3094 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3095 -R library as well as a -L library.
3096
3097 14. The debugging options of pcretest (-d on the command line or D on a
3098 pattern) showed incorrect output for anything following an extended class
3099 that contained multibyte characters and which was followed by a quantifier.
3100
3101 15. Added optional support for general category Unicode character properties
3102 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3103 support. It adds about 90K to the size of the library. The meanings of the
3104 inbuilt class escapes such as \d and \s have NOT been changed.
3105
3106 16. Updated pcredemo.c to include calls to free() to release the memory for the
3107 compiled pattern.
3108
3109 17. The generated file chartables.c was being created in the source directory
3110 instead of in the building directory. This caused the build to fail if the
3111 source directory was different from the building directory, and was
3112 read-only.
3113
3114 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3115 file. No doubt somebody will tell me if they don't make sense... Also added
3116 Dan Mooney's comments about building on OpenVMS.
3117
3118 19. Added support for partial matching via the PCRE_PARTIAL option for
3119 pcre_exec() and the \P data escape in pcretest.
3120
3121 20. Extended pcretest with 3 new pattern features:
3122
3123 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3124 write the compiled pattern to the file whose name is "rest-of-line".
3125 This is a straight binary dump of the data, with the saved pointer to
3126 the character tables forced to be NULL. The study data, if any, is
3127 written too. After writing, pcretest reads a new pattern.
3128
3129 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3130 compiled pattern from the given file. There must not be any
3131 occurrences of "<" in the file name (pretty unlikely); if there are,
3132 pcretest will instead treat the initial "<" as a pattern delimiter.
3133 After reading in the pattern, pcretest goes on to read data lines as
3134 usual.
3135
3136 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3137 and 16-bit fields in a compiled pattern, to simulate a pattern that
3138 was compiled on a host of opposite endianness.
3139
3140 21. The pcre-exec() function can now cope with patterns that were compiled on
3141 hosts of opposite endianness, with this restriction:
3142
3143 As for any compiled expression that is saved and used later, the tables
3144 pointer field cannot be preserved; the extra_data field in the arguments
3145 to pcre_exec() should be used to pass in a tables address if a value
3146 other than the default internal tables were used at compile time.
3147
3148 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3149 now diagnosed as an error. Previously, most of the time, a negative number
3150 would have been treated as zero, but if in addition "ovector" was passed as
3151 NULL, a crash could occur.
3152
3153 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3154 new versions from the libtool 1.5 distribution (the last one is a copy of
3155 a file called libtool.m4). This seems to have fixed the need to patch
3156 "configure" to support Darwin 1.3 (which I used to do). However, I still
3157 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3158 workstation).
3159
3160 24. Changed the PCRE licence to be the more standard "BSD" licence.
3161
3162
3163 Version 4.5 01-Dec-03
3164 ---------------------
3165
3166 1. There has been some re-arrangement of the code for the match() function so
3167 that it can be compiled in a version that does not call itself recursively.
3168 Instead, it keeps those local variables that need separate instances for
3169 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3170 needs to "recurse". Keeping track of where control must go is done by means
3171 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3172 hide most of the details from the main code, and operates only if
3173 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3174 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3175 operating.
3176
3177 To make it easier for callers to provide specially tailored get/free
3178 functions for this usage, two new functions, pcre_stack_malloc, and
3179 pcre_stack_free, are used. They are always called in strict stacking order,
3180 and the size of block requested is always the same.
3181
3182 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3183 PCRE has been compiled to use the stack or the heap for recursion. The
3184 -C option of pcretest uses this to show which version is compiled.
3185
3186 A new data escape \S, is added to pcretest; it causes the amounts of store
3187 obtained and freed by both kinds of malloc/free at match time to be added
3188 to the output.
3189
3190 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3191 what's available on my current Linux desktop machine.
3192
3193 3. When matching a UTF-8 string, the test for a valid string at the start has
3194 been extended. If start_offset is not zero, PCRE now checks that it points
3195 to a byte that is the start of a UTF-8 character. If not, it returns
3196 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3197 this is necessary because there may be backward assertions in the pattern.
3198 When matching the same subject several times, it may save resources to use
3199 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3200
3201 4. The code for checking the validity of UTF-8 strings has been tightened so
3202 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3203 containing "overlong sequences".
3204
3205 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3206 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3207 should have been "&", but it just so happened that all the cases this let
3208 through by mistake were picked up later in the function.
3209
3210 6. I had used a variable called "isblank" - this is a C99 function, causing
3211 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3212
3213 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3214 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3215 so the version is shown; (c) stop "make test" echoing "./RunTest".
3216
3217 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3218
3219 9. The prototype for memmove() for systems that don't have it was using
3220 size_t, but the inclusion of the header that defines size_t was later. I've
3221 moved the #includes for the C headers earlier to avoid this.
3222
3223 10. Added some adjustments to the code to make it easier to compiler on certain
3224 special systems:
3225
3226 (a) Some "const" qualifiers were missing.
3227 (b) Added the macro EXPORT before all exported functions; by default this
3228 is defined to be empty.
3229 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3230 that it reads its output file name as an argument instead of writing
3231 to the standard output and assuming this can be redirected.
3232
3233 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3234 class containing characters with values greater than 255, PCRE compilation
3235 went into a loop.
3236
3237 12. A recursive reference to a subpattern that was within another subpattern
3238 that had a minimum quantifier of zero caused PCRE to crash. For example,
3239 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3240 recursion. If the recursively-called subpattern itself had a zero repeat,
3241 that was OK.
3242
3243 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3244 buffer into which it was copied (for escape processing) was still set at
3245 1024, so long lines caused crashes.
3246
3247 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3248 "internal error: code overflow...". This applied to any character class
3249 that was followed by a possessive quantifier.
3250
3251 15. Modified the Makefile to add libpcre.la as a prerequisite for
3252 libpcreposix.la because I was told this is needed for a parallel build to
3253 work.
3254
3255 16. If a pattern that contained .* following optional items at the start was
3256 studied, the wrong optimizing data was generated, leading to matching
3257 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3258 matching string must start with a or b or c. The correct conclusion for
3259 this pattern is that a match can start with any character.
3260
3261
3262 Version 4.4 13-Aug-03
3263 ---------------------
3264
3265 1. In UTF-8 mode, a character class containing characters with values between
3266 127 and 255 was not handled correctly if the compiled pattern was studied.
3267 In fixing this, I have also improved the studying algorithm for such
3268 classes (slightly).
3269
3270 2. Three internal functions had redundant arguments passed to them. Removal
3271 might give a very teeny performance improvement.
3272
3273 3. Documentation bug: the value of the capture_top field in a callout is *one
3274 more than* the number of the hightest numbered captured substring.
3275
3276 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3277 in incorrectly linking with a previously installed version. They now link
3278 explicitly with libpcre.la.
3279
3280 5. configure.in no longer needs to recognize Cygwin specially.
3281
3282 6. A problem in pcre.in for Windows platforms is fixed.
3283
3284 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3285 pcretest, it used to include the size of the study block as part of its
3286 output. Unfortunately, the structure contains a field that has a different
3287 size on different hardware architectures. This meant that the tests that
3288 showed this size failed. As the block is currently always of a fixed size,
3289 this information isn't actually particularly useful in pcretest output, so
3290 I have just removed it.
3291
3292 8. Three pre-processor statements accidentally did not start in column 1.
3293 Sadly, there are *still* compilers around that complain, even though
3294 standard C has not required this for well over a decade. Sigh.
3295
3296 9. In pcretest, the code for checking callouts passed small integers in the
3297 callout_data field, which is a void * field. However, some picky compilers
3298 complained about the casts involved for this on 64-bit systems. Now
3299 pcretest passes the address of the small integer instead, which should get
3300 rid of the warnings.
3301
3302 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3303 both compile and run time, and gives an error if an invalid UTF-8 sequence
3304 is found. There is a option for disabling this check in cases where the
3305 string is known to be correct and/or the maximum performance is wanted.
3306
3307 11. In response to a bug report, I changed one line in Makefile.in from
3308
3309 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3310 to
3311 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3312
3313 to look similar to other lines, but I have no way of telling whether this
3314 is the right thing to do, as I do not use Windows. No doubt I'll get told
3315 if it's wrong...
3316
3317
3318 Version 4.3 21-May-03
3319 ---------------------
3320
3321 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3322 Makefile.
3323
3324 2. Some refactoring to improve the quality of the code:
3325
3326 (i) The utf8_table... variables are now declared "const".
3327
3328 (ii) The code for \cx, which used the "case flipping" table to upper case
3329 lower case letters, now just substracts 32. This is ASCII-specific,
3330 but the whole concept of \cx is ASCII-specific, so it seems
3331 reasonable.
3332
3333 (iii) PCRE was using its character types table to recognize decimal and
3334 hexadecimal digits in the pattern. This is silly, because it handles
3335 only 0-9, a-f, and A-F, but the character types table is locale-
3336 specific, which means strange things might happen. A private
3337 table is now used for this - though it costs 256 bytes, a table is
3338 much faster than multiple explicit tests. Of course, the standard
3339 character types table is still used for matching digits in subject
3340 strings against \d.
3341
3342 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3343 ending in _t are). So I've renamed it as ESC_tee.
3344
3345 3. The first argument for regexec() in the POSIX wrapper should have been
3346 defined as "const".
3347
3348 4. Changed pcretest to use malloc() for its buffers so that they can be
3349 Electric Fenced for debugging.
3350
3351 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3352 to read one or more bytes before the start of the subject string. Often this
3353 had no effect on PCRE's behaviour, but in some circumstances it could
3354 provoke a segmentation fault.
3355
3356 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3357 to try to read one or more bytes before the start of the subject string.
3358
3359 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3360 UTF-8 support could misbehave in various ways if the subject string
3361 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3362 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3363 back over UTF-8 characters.)
3364
3365
3366 Version 4.2 14-Apr-03
3367 ---------------------
3368
3369 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3370
3371 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3372 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3373 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3374 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3375 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3376 and BUILD_EXEEXT
3377 Note: automatic setting of the BUILD variables is not yet working
3378 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3379 compile-time but not at link-time
3380 [LINK]: use for linking executables only
3381 make different versions for Windows and non-Windows
3382 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3383 libraries
3384 [LINK_FOR_BUILD]: new variable
3385 [OBJEXT]: use throughout
3386 [EXEEXT]: use throughout
3387 <winshared>: new target
3388 <wininstall>: new target
3389 <dftables.o>: use native compiler
3390 <dftables>: use native linker
3391 <install>: handle Windows platform correctly
3392 <clean>: ditto
3393 <check>: ditto
3394 copy DLL to top builddir before testing
3395
3396 As part of these changes, -no-undefined was removed again. This was reported
3397 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3398 in any case.
3399
3400 3. Some tidies to get rid of compiler warnings:
3401
3402 . In the match_data structure, match_limit was an unsigned long int, whereas
3403 match_call_count was an int. I've made them both unsigned long ints.
3404
3405 . In pcretest the fact that a const uschar * doesn't automatically cast to
3406 a void * provoked a warning.
3407
3408 . Turning on some more compiler warnings threw up some "shadow" variables
3409 and a few more missing casts.
3410
3411 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3412 option, a class that contained a single character with a value between 128
3413 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3414
3415 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3416 option, a class that contained several characters, but with at least one
3417 whose value was between 128 and 255 caused PCRE to crash.
3418
3419
3420 Version 4.1 12-Mar-03
3421 ---------------------
3422
3423 1. Compiling with gcc -pedantic found a couple of places where casts were
3424 needed, and a string in dftables.c that was longer than standard compilers are
3425 required to support.
3426
3427 2. Compiling with Sun's compiler found a few more places where the code could
3428 be tidied up in order to avoid warnings.
3429
3430 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3431 first of these names is deprecated in the latest Autoconf in favour of the name
3432 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3433 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3434 analogy I have changed it to CFLAGS_FOR_BUILD.
3435
3436 4. Added -no-undefined to the linking command in the Makefile, because this is
3437 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3438 linking step for the pcreposix library.
3439
3440 5. PCRE was failing to diagnose the case of two named groups with the same
3441 name.
3442
3443 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3444 literal character that is needed in the subject for a match, and scans along to
3445 ensure that it is present before embarking on the full matching process. This
3446 saves time in cases of nested unlimited repeats that are never going to match.
3447 Problem: the scan can take a lot of time if the subject is very long (e.g.
3448 megabytes), thus penalizing straightforward matches. It is now done only if the
3449 amount of subject to be scanned is less than 1000 bytes.
3450
3451 7. A lesser problem with the same optimization is that it was recording the
3452 first character of an anchored pattern as "needed", thus provoking a search
3453 right along the subject, even when the first match of the pattern was going to
3454 fail. The "needed" character is now not set for anchored patterns, unless it
3455 follows something in the pattern that is of non-fixed length. Thus, it still
3456 fulfils its original purpose of finding quick non-matches in cases of nested
3457 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3458
3459
3460 Version 4.0 17-Feb-03
3461 ---------------------
3462
3463 1. If a comment in an extended regex that started immediately after a meta-item
3464 extended to the end of string, PCRE compiled incorrect data. This could lead to
3465 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3466
3467 2. Moved to autoconf 2.53 and libtool 1.4.2.
3468
3469 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3470 the special perltest8 script is no longer needed - all the tests can be run
3471 from a single perltest script.
3472
3473 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3474 by \s. It has now been removed in PCRE. This means it isn't recognized as
3475 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3476 class [:space:] *does* include VT, thereby creating a mess.
3477
3478 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3479 space and tab.
3480
3481 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3482 its new features into the main test script, reducing the number of scripts.
3483
3484 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3485 were backward compatible, and made the (?i) apply to the whole pattern, as if
3486 /i were given. Now it behaves more logically, and applies the option setting
3487 only to what follows. PCRE has been changed to follow suit. However, if it
3488 finds options settings right at the start of the pattern, it extracts them into
3489 the global options, as before. Thus, they show up in the info data.
3490
3491 8. Added support for the \Q...\E escape sequence. Characters in between are
3492 treated as literals. This is slightly different from Perl in that $ and @ are
3493 also handled as literals inside the quotes. In Perl, they will cause variable
3494 interpolation. Note the following examples:
3495
3496 Pattern PCRE matches Perl matches
3497
3498 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3499 \Qabc\$xyz\E abc\$xyz abc\$xyz
3500 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3501
3502 For compatibility with Perl, \Q...\E sequences are recognized inside character
3503 classes as well as outside them.
3504
3505 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3506 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3507 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3508 signed/unsigned warnings.
3509
3510 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3511 option for pcretest, so I've replaced it by a simple function that does just
3512 that job.
3513
3514 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3515 "pcregrep -".
3516
3517 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3518 Java package. This provides some syntactic sugar for simple cases of what my
3519 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3520 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3521 item, you can use this simplified notation. Note that only makes sense with
3522 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3523 greediness, whatever the setting of the PCRE_UNGREEDY option.
3524
3525 13. A change of greediness default within a pattern was not taking effect at
3526 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3527 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3528 was abstracted outside.
3529
3530 14. PCRE now supports the \G assertion. It is true when the current matching
3531 position is at the start point of the match. This differs from \A when the
3532 starting offset is non-zero. Used with the /g option of pcretest (or similar
3533 code), it works in the same way as it does for Perl's /g option. If all
3534 alternatives of a regex begin with \G, the expression is anchored to the start
3535 match position, and the "anchored" flag is set in the compiled expression.
3536
3537 15. Some bugs concerning the handling of certain option changes within patterns
3538 have been fixed. These applied to options other than (?ims). For example,
3539 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3540 been the other way round. Some of this was related to change 7 above.
3541
3542 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3543 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3544 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3545 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3546
3547 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3548 mode. Unlike ".", it always matches newline, whatever the setting of
3549 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3550 assertions. Perl allows it, but it doesn't (in general) work because it can't
3551 calculate the length of the lookbehind. At least, that's the case for Perl
3552 5.8.0 - I've been told they are going to document that it doesn't work in
3553 future.
3554
3555 18. Added an error diagnosis for escapes that PCRE does not support: these are
3556 \L, \l, \N, \P, \p, \U, \u, and \X.
3557
3558 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3559 reading past the end of the pattern in cases such as /[abcd/.
3560
3561 20. PCRE was getting more memory than necessary for patterns with classes that
3562 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3563
3564 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3565 compiling PCRE for use with Virtual Pascal.
3566
3567 22. Small fix to the Makefile to make it work properly if the build is done
3568 outside the source tree.
3569
3570 23. Added a new extension: a condition to go with recursion. If a conditional
3571 subpattern starts with (?(R) the "true" branch is used if recursion has
3572 happened, whereas the "false" branch is used only at the top level.
3573
3574 24. When there was a very long string of literal characters (over 255 bytes
3575 without UTF support, over 250 bytes with UTF support), the computation of how
3576 much memory was required could be incorrect, leading to segfaults or other
3577 strange effects.
3578
3579 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3580 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3581 there was a subsequent back reference to those brackets. This meant that, for
3582 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3583 possible to check for precisely this case. All we can do is abandon the
3584 optimization if .* occurs inside capturing brackets when there are any back
3585 references whatsoever. (See below for a better fix that came later.)
3586
3587 26. The handling of the optimization for finding the first character of a
3588 non-anchored pattern, and for finding a character that is required later in the
3589 match were failing in some cases. This didn't break the matching; it just
3590 failed to optimize when it could. The way this is done has been re-implemented.
3591
3592 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3593
3594 28. Added a new feature that provides some of the functionality that Perl
3595 provides with (?{...}). The facility is termed a "callout". The way it is done
3596 in PCRE is for the caller to provide an optional function, by setting
3597 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3598 global variable. By default it is unset, which disables all calling out. To get
3599 the function called, the regex must include (?C) at appropriate points. This
3600 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3601 This provides a means of identifying different callout points. When PCRE
3602 reaches such a point in the regex, if pcre_callout has been set, the external
3603 function is called. It is provided with data in a structure called
3604 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3605 matching continues; if it returns a non-zero value, the match at the current
3606 point fails. However, backtracking will occur if possible. [This was changed
3607 later and other features added - see item 49 below.]
3608
3609 29. pcretest is upgraded to test the callout functionality. It provides a
3610 callout function that displays information. By default, it shows the start of
3611 the match and the current position in the text. There are some new data escapes
3612 to vary what happens:
3613
3614 \C+ in addition, show current contents of captured substrings
3615 \C- do not supply a callout function
3616 \C!n return 1 when callout number n is reached
3617 \C!n!m return 1 when callout number n is reached for the mth time
3618
3619 30. If pcregrep was called with the -l option and just a single file name, it
3620 output "<stdin>" if a match was found, instead of the file name.
3621
3622 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3623 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3624 pcre_exec(). This saves a malloc/free per call. The default value of
3625 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3626 when configuring.
3627
3628 32. The default maximum size of a compiled pattern is 64K. There have been a
3629 few cases of people hitting this limit. The code now uses macros to handle the
3630 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3631 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3632 configuring. Tests 2 and 5 work only with 2-byte links because they output
3633 debugging information about compiled patterns.
3634
3635 33. Internal code re-arrangements:
3636
3637 (a) Moved the debugging function for printing out a compiled regex into
3638 its own source file (printint.c) and used #include to pull it into
3639 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3640 separate copies.
3641
3642 (b) Defined the list of op-code names for debugging as a macro in
3643 internal.h so that it is next to the definition of the opcodes.
3644
3645 (c) Defined a table of op-code lengths for simpler skipping along compiled
3646 code. This is again a macro in internal.h so that it is next to the
3647 definition of the opcodes.
3648
3649 34. Added support for recursive calls to individual subpatterns, along the
3650 lines of Robin Houston's patch (but implemented somewhat differently).
3651
3652 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3653 allow it to read and process whole directories in Win32. This code was
3654 contributed by Lionel Fourquaux; it has not been tested by me.
3655
3656 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3657 used to name a group. Names consist of alphanumerics and underscores, and must
3658 be unique. Back references use the syntax (?P=name) and recursive calls use
3659 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3660 numbers. The function pcre_fullinfo() can be used after compilation to extract
3661 a name/number map. There are three relevant calls:
3662
3663 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3664 PCRE_INFO_NAMECOUNT yields the number of entries
3665 PCRE_INFO_NAMETABLE yields a pointer to the map.
3666
3667 The map is a vector of fixed-size entries. The size of each entry depends on
3668 the length of the longest name used. The first two bytes of each entry are the
3669 group number, most significant byte first. There follows the corresponding
3670 name, zero terminated. The names are in alphabetical order.
3671
3672 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3673 case instead of 255. Making it the same both with and without UTF-8 support
3674 means that the same test output works with both.
3675
3676 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3677 calling malloc() with a zero argument.
3678
3679 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3680 optimization. I've improved things by keeping a bitmap of backreferences with
3681 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3682 fact referenced, the optimization can be applied. It is unlikely that a
3683 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3684 the match to follow \n) will appear inside brackets with a number greater than
3685 31, but if it does, any back reference > 31 suppresses the optimization.
3686
3687 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3688 of disabling numbered capturing parentheses. Any opening parenthesis that is
3689 not followed by ? behaves as if it were followed by ?: but named parentheses
3690 can still be used for capturing (and they will acquire numbers in the usual
3691 way).
3692
3693 41. Redesigned the return codes from the match() function into yes/no/error so
3694 that errors can be passed back from deep inside the nested calls. A malloc
3695 failure while inside a recursive subpattern call now causes the
3696 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3697
3698 42. It is now possible to set a limit on the number of times the match()
3699 function is called in a call to pcre_exec(). This facility makes it possible to
3700 limit the amount of recursion and backtracking, though not in a directly
3701 obvious way, because the match() function is used in a number of different
3702 circumstances. The count starts from zero for each position in the subject
3703 string (for non-anchored patterns). The default limit is, for compatibility, a
3704 large number, namely 10 000 000. You can change this in two ways:
3705
3706 (a) When configuring PCRE before making, you can use --with-match-limit=n
3707 to set a default value for the compiled library.
3708
3709 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3710 a different value is set. See 45 below.
3711
3712 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3713
3714 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3715 of things that can be changed at compile time. The first argument specifies
3716 what is wanted and the second points to where the information is to be placed.
3717 The current list of available information is:
3718
3719 PCRE_CONFIG_UTF8
3720
3721 The output is an integer that is set to one if UTF-8 support is available;
3722 otherwise it is set to zero.
3723
3724 PCRE_CONFIG_NEWLINE
3725
3726 The output is an integer that it set to the value of the code that is used for
3727 newline. It is either LF (10) or CR (13).
3728
3729 PCRE_CONFIG_LINK_SIZE
3730
3731 The output is an integer that contains the number of bytes used for internal
3732 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3733
3734 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3735
3736 The output is an integer that contains the threshold above which the POSIX
3737 interface uses malloc() for output vectors. See item 31 above.
3738
3739 PCRE_CONFIG_MATCH_LIMIT
3740
3741 The output is an unsigned integer that contains the default limit of the number
3742 of match() calls in a pcre_exec() execution. See 42 above.
3743
3744 44. pcretest has been upgraded by the addition of the -C option. This causes it
3745 to extract all the available output from the new pcre_config() function, and to
3746 output it. The program then exits immediately.
3747
3748 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3749 order to support additional features. One way would have been to define
3750 pcre_exec2() (for example) with extra arguments, but this would not have been
3751 extensible, and would also have required all calls to the original function to
3752 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3753 is used for passing in "extra" data from pcre_study().
3754
3755 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3756 contains the following fields:
3757
3758 flags a bitmap indicating which of the following fields are set
3759 study_data opaque data from pcre_study()
3760 match_limit a way of specifying a limit on match() calls for a specific
3761 call to pcre_exec()
3762 callout_data data for callouts (see 49 below)
3763
3764 The flag bits are also defined in pcre.h, and are
3765
3766 PCRE_EXTRA_STUDY_DATA
3767 PCRE_EXTRA_MATCH_LIMIT
3768 PCRE_EXTRA_CALLOUT_DATA
3769
3770 The pcre_study() function now returns one of these new pcre_extra blocks, with
3771 the actual study data pointed to by the study_data field, and the
3772 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3773 before. That is, this change is entirely upwards-compatible and requires no
3774 change to existing code.
3775
3776 If you want to pass in additional data to pcre_exec(), you can either place it
3777 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3778 block.
3779
3780 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3781 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3782 times with different match limits, until it finds the minimum value needed for
3783 pcre_exec() to complete. The value is then output. This can be instructive; for
3784 most simple matches the number is quite small, but for pathological cases it
3785 gets very large very quickly.
3786
3787 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3788 returns the size of the data block pointed to by the study_data field in a
3789 pcre_extra block, that is, the value that was passed as the argument to
3790 pcre_malloc() when PCRE was getting memory in which to place the information
3791 created by pcre_study(). The fourth argument should point to a size_t variable.
3792 pcretest has been extended so that this information is shown after a successful
3793 pcre_study() call when information about the compiled regex is being displayed.
3794
3795 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3796 because what follows is always an absolute path. (Later: it turns out that this
3797 is more than cosmetic for MinGW, because it doesn't like empty path
3798 components.)
3799
3800 49. Some changes have been made to the callout feature (see 28 above):
3801
3802 (i) A callout function now has three choices for what it returns:
3803
3804 0 => success, carry on matching
3805 > 0 => failure at this point, but backtrack if possible
3806 < 0 => serious error, return this value from pcre_exec()
3807
3808 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3809 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3810 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3811 use by callout functions. It will never be used by PCRE itself.
3812
3813 (ii) The pcre_extra structure (see 45 above) has a void * field called
3814 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3815 pcre_callout_block structure has a field of the same name. The contents of
3816 the field passed in the pcre_extra structure are passed to the callout
3817 function in the corresponding field in the callout block. This makes it
3818 easier to use the same callout-containing regex from multiple threads. For
3819 testing, the pcretest program has a new data escape
3820
3821 \C*n pass the number n (may be negative) as callout_data
3822
3823 If the callout function in pcretest receives a non-zero value as
3824 callout_data, it returns that value.
3825
3826 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3827 there were some redundant $(CFLAGS) in commands that are now specified as
3828 $(LINK), which already includes $(CFLAGS).
3829
3830 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3831 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3832 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3833 one-byte characters throughout. Note that case-insensitive matching applies
3834 only to characters whose values are less than 256. PCRE doesn't support the
3835 notion of cases for higher-valued characters.
3836
3837 (i) A character class whose characters are all within 0-255 is handled as
3838 a bit map, and the map is inverted for negative classes. Previously, a
3839 character > 255 always failed to match such a class; however it should
3840 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3841
3842 (ii) A negated character class with a single character < 255 is coded as
3843 "not this character" (OP_NOT). This wasn't working properly when the test
3844 character was multibyte, either singly or repeated.
3845
3846 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3847 mode, for example: \x{100}{2,3}.
3848
3849 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3850 singly or repeated) now correctly test multibyte characters. However,
3851 PCRE doesn't recognize any characters with values greater than 255 as
3852 digits, spaces, or word characters. Such characters always match \D, \S,
3853 and \W, and never match \d, \s, or \w.
3854
3855 (v) Classes may now contain characters and character ranges with values
3856 greater than 255. For example: [ab\x{100}-\x{400}].
3857
3858 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3859 PCRE in UTF-8 mode.
3860
3861 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3862 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3863 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3864 value.)
3865
3866 53. The single man page has become too large. I have therefore split it up into
3867 a number of separate man pages. These also give rise to individual HTML pages;
3868 these are now put in a separate directory, and there is an index.html page that
3869 lists them all. Some hyperlinking between the pages has been installed.
3870
3871 54. Added convenience functions for handling named capturing parentheses.
3872
3873 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3874 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3875 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3876 are faulted.
3877
3878 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3879 calling configure. These values are used when compiling the dftables.c program
3880 which is run to generate the source of the default character tables. They
3881 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3882 you will need to set these values.
3883
3884 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3885
3886
3887 Version 3.9 02-Jan-02
3888 ---------------------
3889
3890 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3891
3892 2. If --disable-static was given, the building process failed when trying to
3893 build pcretest and pcregrep. (For some reason it was using libtool to compile
3894 them, which is not right, as they aren't part of the library.)
3895
3896
3897 Version 3.8 18-Dec-01
3898 ---------------------
3899
3900 1. The experimental UTF-8 code was completely screwed up. It was packing the
3901 bytes in the wrong order. How dumb can you get?
3902
3903
3904 Version 3.7 29-Oct-01
3905 ---------------------
3906
3907 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3908 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3909 this didn't happen under Solaris 8, where I normally test things.
3910
3911 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3912 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3913 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3914 this fix an hour or so after the initial 3.7 release.)
3915
3916
3917 Version 3.6 23-Oct-01
3918 ---------------------
3919
3920 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3921 offsets passed as NULL with zero offset count.
3922
3923 2. The config.guess and config.sub files had not been updated when I moved to
3924 the latest autoconf.
3925
3926
3927 Version 3.5 15-Aug-01
3928 ---------------------
3929
3930 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3931 had been forgotten.
3932
3933 2. By using declared but undefined structures, we can avoid using "void"
3934 definitions in pcre.h while keeping the internal definitions of the structures
3935 private.
3936
3937 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3938 user point of view, this means that both static and shared libraries are built
3939 by default, but this can be individually controlled. More of the work of
3940 handling this static/shared cases is now inside libtool instead of PCRE's make
3941 file.
3942
3943 4. The pcretest utility is now installed along with pcregrep because it is
3944 useful for users (to test regexs) and by doing this, it automatically gets
3945 relinked by libtool. The documentation has been turned into a man page, so
3946 there are now .1, .txt, and .html versions in /doc.
3947
3948 5. Upgrades to pcregrep:
3949 (i) Added long-form option names like gnu grep.
3950 (ii) Added --help to list all options with an explanatory phrase.
3951 (iii) Added -r, --recursive to recurse into sub-directories.
3952 (iv) Added -f, --file to read patterns from a file.
3953
3954 6. pcre_exec() was referring to its "code" argument before testing that
3955 argument for NULL (and giving an error if it was NULL).
3956
3957 7. Upgraded Makefile.in to allow for compiling in a different directory from
3958 the source directory.
3959
3960 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3961 options bits, the pointer it was passed was to an int instead of to an unsigned
3962 long int. This mattered only on 64-bit systems.
3963
3964 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3965 generated) instead of pcre.in, which it its source. Also made the same change
3966 in several of the .c files.
3967
3968 10. A new release of gcc defines printf() as a macro, which broke pcretest
3969 because it had an ifdef in the middle of a string argument for printf(). Fixed
3970 by using separate calls to printf().
3971
3972 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3973 script, to force use of CR or LF instead of \n in the source. On non-Unix
3974 systems, the value can be set in config.h.
3975
3976 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3977 absolute limit. Changed the text of the error message to make this clear, and
3978 likewise updated the man page.
3979
3980 13. The limit of 99 on the number of capturing subpatterns has been removed.
3981 The new limit is 65535, which I hope will not be a "real" limit.
3982
3983
3984 Version 3.4 22-Aug-00
3985 ---------------------
3986
3987 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3988
3989 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3990
3991
3992 Version 3.3 01-Aug-00
3993 ---------------------
3994
3995 1. If an octal character was given, but the value was greater than \377, it
3996 was not getting masked to the least significant bits, as documented. This could
3997 lead to crashes in some systems.
3998
3999 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4000 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4001
4002 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4003 These just pass their arguments on to (pcre_free)(), but they are provided
4004 because some uses of PCRE bind it to non-C systems that can call its functions,
4005 but cannot call free() or pcre_free() directly.
4006
4007 4. Add "make test" as a synonym for "make check". Corrected some comments in
4008 the Makefile.
4009
4010 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4011 Makefile.
4012
4013 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4014 command called pgrep for grepping around the active processes.
4015
4016 7. Added the beginnings of support for UTF-8 character strings.
4017
4018 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4019 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4020 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4021 out for the ar command.)
4022
4023
4024 Version 3.2 12-May-00
4025 ---------------------
4026
4027 This is purely a bug fixing release.
4028
4029 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4030 of ZA. This was just one example of several cases that could provoke this bug,
4031 which was introduced by change 9 of version 2.00. The code for breaking
4032 infinite loops after an iteration that matches an empty string was't working
4033 correctly.
4034
4035 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4036 when matched against abbab (for example). After matching an empty string, it
4037 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4038 caused it to match further down the string than it should.
4039
4040 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4041 was there because it doesn't seem to be needed, and it causes trouble on some
4042 systems, as it is not a Standard C header. It has been removed.
4043
4044 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4045 were reported on the Macintosh. The changes were from
4046
4047 while ((c = *(++ptr)) != 0 && c != '\n');
4048 to
4049 while ((c = *(++ptr)) != 0 && c != '\n') ;
4050
4051 Totally extraordinary, but if that's what it takes...
4052
4053 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4054 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4055 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4056 assumes the way PCRE uses memmove() (always moving upwards).
4057
4058 6. PCRE is being used in one environment where strchr() is not available. There
4059 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4060 faster code anyway.
4061
4062
4063 Version 3.1 09-Feb-00
4064 ---------------------
4065
4066 The only change in this release is the fixing of some bugs in Makefile.in for
4067 the "install" target:
4068
4069 (1) It was failing to install pcreposix.h.
4070
4071 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4072
4073
4074 Version 3.0 01-Feb-00
4075 ---------------------
4076
4077 1. Add support for the /+ modifier to perltest (to output $` like it does in
4078 pcretest).
4079
4080 2. Add support for the /g modifier to perltest.
4081
4082 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4083 matches null strings.
4084
4085 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4086 pattern. Perl treats empty patterns specially - it reuses the most recent
4087 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4088 effect.
4089
4090 5. The POSIX interface was broken in that it was just handing over the POSIX
4091 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4092 required a bigger vector, with some working space on the end. This means that
4093 the POSIX wrapper now has to get and free some memory, and copy the results.
4094
4095 6. Added some simple autoconf support, placing the test data and the
4096 documentation in separate directories, re-organizing some of the
4097 information files, and making it build pcre-config (a GNU standard). Also added
4098 libtool support for building PCRE as a shared library, which is now the
4099 default.
4100
4101 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4102 09 are not valid octal constants. Single digits will be used for minor values
4103 less than 10.
4104
4105 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4106 existing programs that set these in the POSIX interface can use PCRE without
4107 modification.
4108
4109 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4110 return all that pcre_info() returns, plus additional data. The pcre_info()
4111 function is retained for compatibility, but is considered to be obsolete.
4112
4113 10. Added experimental recursion feature (?R) to handle one common case that
4114 Perl 5.6 will be able to do with (?p{...}).
4115
4116 11. Added support for POSIX character classes like [:alpha:], which Perl is
4117 adopting.
4118
4119
4120 Version 2.08 31-Aug-99
4121 ----------------------
4122
4123 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4124 trying to match at the startoffset position, but instead was moving forward to
4125 the next newline as if a previous match had failed.
4126
4127 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4128 and could get into a loop if a null string was matched other than at the start
4129 of the subject.
4130
4131 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4132 be distinguished at compile time, and for completeness also added PCRE_DATE.
4133
4134 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4135 in GnuWin32 environments.
4136
4137
4138 Version 2.07 29-Jul-99
4139 ----------------------
4140
4141 1. The documentation is now supplied in plain text form and HTML as well as in
4142 the form of man page sources.
4143
4144 2. C++ compilers don't like assigning (void *) values to other pointer types.
4145 In particular this affects malloc(). Although there is no problem in Standard
4146 C, I've put in casts to keep C++ compilers happy.
4147
4148 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4149 should be (const char *).
4150
4151 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4152 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4153 However, I haven't made this a standard facility. The documentation doesn't
4154 mention it, and the Makefile doesn't support it.
4155
4156 5. The Makefile now contains an "install" target, with editable destinations at
4157 the top of the file. The pcretest program is not installed.
4158
4159 6. pgrep -V now gives the PCRE version number and date.
4160
4161 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4162 causing the entire string to be ignored, instead of just the last character.
4163
4164 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4165 non-matching string, it can take a very, very long time, even for strings of
4166 quite modest length, because of the nested recursion. PCRE now does better in
4167 some of these cases. It does this by remembering the last required literal
4168 character in the pattern, and pre-searching the subject to ensure it is present
4169 before running the real match. In other words, it applies a heuristic to detect
4170 some types of certain failure quickly, and in the above example, if presented
4171 with a string that has no trailing " it gives "no match" very quickly.
4172
4173 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4174 other alternatives are tried instead.
4175
4176
4177 Version 2.06 09-Jun-99
4178 ----------------------
4179
4180 1. Change pcretest's output for amount of store used to show just the code
4181 space, because the remainder (the data block) varies in size between 32-bit and
4182 64-bit systems.
4183
4184 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4185 start matching at. This allows lookbehinds to work when searching for multiple
4186 occurrences in a string.
4187
4188 3. Added additional options to pcretest for testing multiple occurrences:
4189
4190 /+ outputs the rest of the string that follows a match
4191 /g loops for multiple occurrences, using the new startoffset argument
4192 /G loops for multiple occurrences by passing an incremented pointer
4193
4194 4. PCRE wasn't doing the "first character" optimization for patterns starting
4195 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4196 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4197 the letter 'x'. On long subject strings, this gives a significant speed-up.
4198
4199
4200 Version 2.05 21-Apr-99
4201 ----------------------
4202
4203 1. Changed the type of magic_number from int to long int so that it works
4204 properly on 16-bit systems.
4205
4206 2. Fixed a bug which caused patterns starting with .* not to work correctly
4207 when the subject string contained newline characters. PCRE was assuming
4208 anchoring for such patterns in all cases, which is not correct because .* will
4209 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4210 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4211 must be retried after every newline in the subject.
4212
4213
4214 Version 2.04 18-Feb-99
4215 ----------------------
4216
4217 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4218 computation of the store needed to hold the pattern was incorrect (too large).
4219 If such patterns were nested a few deep, this could multiply and become a real
4220 problem.
4221
4222 2. Added /M option to pcretest to show the memory requirement of a specific
4223 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4224
4225 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4226 compiled in such a way that the backtracking after subsequent failure was
4227 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4228 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4229
4230
4231 Version 2.03 02-Feb-99
4232 ----------------------
4233
4234 1. Fixed typo and small mistake in man page.
4235
4236 2. Added 4th condition (GPL supersedes if conflict) and created separate
4237 LICENCE file containing the conditions.
4238
4239 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4240 Perl, that is the internal \ allows the delimiter to be included in the
4241 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4242 the final delimiter, add \ to the end of the pattern (to test the error).
4243
4244 4. Added the convenience functions for extracting substrings after a successful
4245 match. Updated pcretest to make it able to test these functions.
4246
4247
4248 Version 2.02 14-Jan-99
4249 ----------------------
4250
4251 1. Initialized the working variables associated with each extraction so that
4252 their saving and restoring doesn't refer to uninitialized store.
4253
4254 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4255 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4256 fix the problem.
4257
4258 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4259 calls, and wasn't printing the correct value for compiling calls. Increased the
4260 default value of LOOPREPEAT, and the number of significant figures in the
4261 times.
4262
4263 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4264
4265 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4266 a building problem on Windows NT with a FAT file system.
4267
4268
4269 Version 2.01 21-Oct-98
4270 ----------------------
4271
4272 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4273 to character tables built by pcre_maketables() in the current locale. If NULL
4274 is passed, the default tables are used.
4275
4276
4277 Version 2.00 24-Sep-98
4278 ----------------------
4279
4280 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4281 it any more.
4282
4283 2. Allow quantification of (?>) groups, and make it work correctly.
4284
4285 3. The first character computation wasn't working for (?>) groups.
4286
4287 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4288 end of the subject) and add 5.005's \z, which really does match only at the
4289 very end of the subject.
4290
4291 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4292
4293 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4294 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4295 localized options. All options to pcre_study() were also removed.
4296
4297 7. Add other new features from 5.005:
4298
4299 $(?<= positive lookbehind
4300 $(?<! negative lookbehind
4301 (?imsx-imsx) added the unsetting capability
4302 such a setting is global if at outer level; local otherwise
4303 (?imsx-imsx:) non-capturing groups with option setting
4304 (?(cond)re|re) conditional pattern matching
4305
4306 A backreference to itself in a repeated group matches the previous
4307 captured string.
4308
4309 8. General tidying up of studying (both automatic and via "study")
4310 consequential on the addition of new assertions.
4311
4312 9. As in 5.005, unlimited repeated groups that could match an empty substring
4313 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4314 runtime if any iteration does actually match an empty substring.
4315
4316 10. Include the RunTest script in the distribution.
4317
4318 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4319 discrepancies, some of which were old and were also with respect to 5.004. They
4320 have now been fixed.
4321
4322
4323 Version 1.09 28-Apr-98
4324 ----------------------
4325
4326 1. A negated single character class followed by a quantifier with a minimum
4327 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4328 program crashes, or just wrong answers. This did not apply to negated classes
4329 containing more than one character, or to minima other than one.
4330
4331
4332 Version 1.08 27-Mar-98
4333 ----------------------
4334
4335 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4336
4337 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4338 latter must appear before anything that relies on it in the pattern.
4339
4340
4341 Version 1.07 16-Feb-98
4342 ----------------------
4343
4344 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4345 repeat of a potentially empty string).
4346
4347
4348 Version 1.06 23-Jan-98
4349 ----------------------
4350
4351 1. Added Markus Oberhumer's little patches for C++.
4352
4353 2. Literal strings longer than 255 characters were broken.
4354
4355
4356 Version 1.05 23-Dec-97
4357 ----------------------
4358
4359 1. Negated character classes containing more than one character were failing if
4360 PCRE_CASELESS was set at run time.
4361
4362
4363 Version 1.04 19-Dec-97
4364 ----------------------
4365
4366 1. Corrected the man page, where some "const" qualifiers had been omitted.
4367
4368 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4369 input syntax.
4370
4371 3. Fixed memory leak which occurred when a regex with back references was
4372 matched with an offsets vector that wasn't big enough. The temporary memory
4373 that is used in this case wasn't being freed if the match failed.
4374
4375 4. Tidied pcretest to ensure it frees memory that it gets.
4376
4377 5. Temporary memory was being obtained in the case where the passed offsets
4378 vector was exactly big enough.
4379
4380 6. Corrected definition of offsetof() from change 5 below.
4381
4382 7. I had screwed up change 6 below and broken the rules for the use of
4383 setjmp(). Now fixed.
4384
4385
4386 Version 1.03 18-Dec-97
4387 ----------------------
4388
4389 1. A erroneous regex with a missing opening parenthesis was correctly
4390 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4391 on some systems.
4392
4393 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4394 it was reported that one broken compiler failed on the former because "code" is
4395 also an independent variable.
4396
4397 3. The erroneous regex a[]b caused an array overrun reference.
4398
4399 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4400 fail on data ending with that character. (It was going on too far, and checking
4401 the next character, typically a binary zero.) This was specific to the
4402 optimized code for single-character negative classes.
4403
4404 5. Added a contributed patch from the TIN world which does the following:
4405
4406 + Add an undef for memmove, in case the the system defines a macro for it.
4407
4408 + Add a definition of offsetof(), in case there isn't one. (I don't know
4409 the reason behind this - offsetof() is part of the ANSI standard - but
4410 it does no harm).
4411
4412 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4413 most of the places where whitespace preceded '#'. I have given up and
4414 allowed the remaining 2 cases to be at the margin.
4415
4416 + Rename some variables in pcre to eliminate shadowing. This seems very
4417 pedantic, but does no harm, of course.
4418
4419 6. Moved the call to setjmp() into its own function, to get rid of warnings
4420 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4421
4422 7. Constructs such as \d{8,} were compiling into the equivalent of
4423 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4424 outcome, but in this particular case used more store than had been allocated,
4425 which caused the bug to be discovered because it threw up an internal error.
4426
4427 8. The debugging code in both pcre and pcretest for outputting the compiled
4428 form of a regex was going wrong in the case of back references followed by
4429 curly-bracketed repeats.
4430
4431
4432 Version 1.02 12-Dec-97
4433 ----------------------
4434
4435 1. Typos in pcre.3 and comments in the source fixed.
4436
4437 2. Applied a contributed patch to get rid of places where it used to remove
4438 'const' from variables, and fixed some signed/unsigned and uninitialized
4439 variable warnings.
4440
4441 3. Added the "runtest" target to Makefile.
4442
4443 4. Set default compiler flag to -O2 rather than just -O.
4444
4445
4446 Version 1.01 19-Nov-97
4447 ----------------------
4448
4449 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4450 like /([ab]*)*/, that is, for classes with more than one character in them.
4451
4452 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4453 as /((?>a*))*/ (a PCRE_EXTRA facility).
4454
4455
4456 Version 1.00 18-Nov-97
4457 ----------------------
4458
4459 1. Added compile-time macros to support systems such as SunOS4 which don't have
4460 memmove() or strerror() but have other things that can be used instead.
4461
4462 2. Arranged that "make clean" removes the executables.
4463
4464
4465 Version 0.99 27-Oct-97
4466 ----------------------
4467
4468 1. Fixed bug in code for optimizing classes with only one character. It was
4469 initializing a 32-byte map regardless, which could cause it to run off the end
4470 of the memory it had got.
4471
4472 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4473
4474
4475 Version 0.98 22-Oct-97
4476 ----------------------
4477
4478 1. Fixed bug in code for handling temporary memory usage when there are more
4479 back references than supplied space in the ovector. This could cause segfaults.
4480
4481
4482 Version 0.97 21-Oct-97
4483 ----------------------
4484
4485 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4486
4487 2. Optimized negated single characters not to use a bit map.
4488
4489 3. Brought error texts together as macro definitions; clarified some of them;
4490 fixed one that was wrong - it said "range out of order" when it meant "invalid
4491 escape sequence".
4492
4493 4. Changed some char * arguments to const char *.
4494
4495 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4496
4497 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4498 pcretest.
4499
4500
4501 Version 0.96 16-Oct-97
4502 ----------------------
4503
4504 1. Added a simple "pgrep" utility to the distribution.
4505
4506 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4507 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4508 where "ddd" means "one or more decimal digits".
4509
4510 3. Fixed serious bug. If a pattern had a back reference, but the call to
4511 pcre_exec() didn't supply a large enough ovector to record the related
4512 identifying subpattern, the match always failed. PCRE now remembers the number
4513 of the largest back reference, and gets some temporary memory in which to save
4514 the offsets during matching if necessary, in order to ensure that
4515 backreferences always work.
4516
4517 4. Increased the compatibility with Perl in a number of ways:
4518
4519 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4520 to request this handling. The option can be set at compile or exec time.
4521
4522 (b) $ matches before a terminating newline by default; an option
4523 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4524 mode). The option can be set at compile or exec time.
4525
4526 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4527 the same as Perl's. If the decimal number it represents is less than 10
4528 or there aren't that many previous left capturing parentheses, an octal
4529 escape is read. Inside a character class, it's always an octal escape,
4530 even if it is a single digit.
4531
4532 (d) An escaped but undefined alphabetic character is taken as a literal,
4533 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4534 escapes.
4535
4536 (e) {0} is now permitted. (The previous item is removed from the compiled
4537 pattern).
4538
4539 5. Changed all the names of code files so that the basic parts are no longer
4540 than 10 characters, and abolished the teeny "globals.c" file.
4541
4542 6. Changed the handling of character classes; they are now done with a 32-byte
4543 bit map always.
4544
4545 7. Added the -d and /D options to pcretest to make it possible to look at the
4546 internals of compilation without having to recompile pcre.
4547
4548
4549 Version 0.95 23-Sep-97
4550 ----------------------
4551
4552 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4553 \x20 at the start of a run of normal characters. These were being treated as
4554 real characters, instead of the source characters being re-checked.
4555
4556
4557 Version 0.94 18-Sep-97
4558 ----------------------
4559
4560 1. The functions are now thread-safe, with the caveat that the global variables
4561 containing pointers to malloc() and free() or alternative functions are the
4562 same for all threads.
4563
4564 2. Get pcre_study() to generate a bitmap of initial characters for non-
4565 anchored patterns when this is possible, and use it if passed to pcre_exec().
4566
4567
4568 Version 0.93 15-Sep-97
4569 ----------------------
4570
4571 1. /(b)|(:+)/ was computing an incorrect first character.
4572
4573 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4574 but not actually doing anything yet.
4575
4576 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4577 as Perl does (e.g. [-az] or [az-]).
4578
4579 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4580 all possible positions.
4581
4582 5. Split up into different modules to avoid including unneeded functions in a
4583 compiled binary. However, compile and exec are still in one module. The "study"
4584 function is split off.
4585
4586 6. The character tables are now in a separate module whose source is generated
4587 by an auxiliary program - but can then be edited by hand if required. There are
4588 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4589 toupper() in the code.
4590
4591 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4592 make them global. Abolish the function for setting them, as the caller can now
4593 set them directly.
4594
4595
4596 Version 0.92 11-Sep-97
4597 ----------------------
4598
4599 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4600 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4601
4602 2. Caseless matching was not working in character classes if the characters in
4603 the pattern were in upper case.
4604
4605 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4606
4607 4. Make PCRE_ANCHORED public and accept as a compile option.
4608
4609 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4610 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4611 pass them.
4612
4613 6. Give an error if bad option bits passed at compile or run time.
4614
4615 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4616 pcretest to cause it to pass that flag.
4617
4618 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4619 options, and the first character, if set.
4620
4621 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4622
4623
4624 Version 0.91 10-Sep-97
4625 ----------------------
4626
4627 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4628 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4629
4630 2. PCRE was looping on encountering an indefinitely repeated back reference to
4631 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4632 Perl does - treats the match as successful.
4633
4634 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12