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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 890 - (show annotations) (download)
Wed Jan 18 16:25:19 2012 UTC (2 years, 6 months ago) by ph10
File size: 225623 byte(s)
Add PCRE_CONFIG_JITTARGET output to pcretest -C.

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 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
34 rounding is not applied in this particular case).
35
36 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
37 if they appear, or are escaped, in patterns.
38
39 10. Get rid of a number of -Wunused-but-set-variable warnings.
40
41 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
42 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
43 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
44 also returns the mark "x". This bug applied to capturing parentheses,
45 non-capturing parentheses, and atomic parentheses. It also applied to some
46 assertions.
47
48 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
49 information out of configure.ac instead of relying on pcre.h.generic, which
50 is not stored in the repository.
51
52 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
53 -lreadline.
54
55 14. ZH added PCRE_CONFIG_JITTARGET; added it's output to pcretest -C.
56
57
58 Version 8.21 12-Dec-2011
59 ------------------------
60
61 1. Updating the JIT compiler.
62
63 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
64 are added as well.
65
66 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
67 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
68 calling _pcre_jit_exec. Some extra comments are added.
69
70 4. (*MARK) settings inside atomic groups that do not contain any capturing
71 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
72 was introduced by change 18 for 8.20.
73
74 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
75 ECMA-262 standard.
76
77 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
78 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
79 This bug was probably introduced by change 9 of 8.13.
80
81 7. While fixing 6 above, I noticed that a number of other items were being
82 incorrectly rejected as "not fixed length". This arose partly because newer
83 opcodes had not been added to the fixed-length checking code. I have (a)
84 corrected the bug and added tests for these items, and (b) arranged for an
85 error to occur if an unknown opcode is encountered while checking for fixed
86 length instead of just assuming "not fixed length". The items that were
87 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
88 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
89 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
90
91 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
92 being incorrectly compiled and would have given unpredicatble results.
93
94 9. A possessively repeated subpattern with minimum repeat count greater than
95 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
96 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
97 the first (A) could occur when it should not.
98
99 10. Add a cast and remove a redundant test from the code.
100
101 11. JIT should use pcre_malloc/pcre_free for allocation.
102
103 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
104 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
105 is anything other than /usr, -L is still shown).
106
107 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
108
109 14. Perl does not support \N without a following name in a [] class; PCRE now
110 also gives an error.
111
112 15. If a forward reference was repeated with an upper limit of around 2000,
113 it caused the error "internal error: overran compiling workspace". The
114 maximum number of forward references (including repeats) was limited by the
115 internal workspace, and dependent on the LINK_SIZE. The code has been
116 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
117 the default depends on LINK_SIZE. There is a new upper limit (for safety)
118 of around 200,000 forward references. While doing this, I also speeded up
119 the filling in of repeated forward references.
120
121 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
122 incorrectly expecting the subject to contain another "a" after the start.
123
124 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
125 in the match, the SKIP should be ignored. This was not happening; instead
126 the SKIP was being treated as NOMATCH. For patterns such as
127 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
128 tested.
129
130 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
131 now much more compatible with Perl, in particular in cases where the result
132 is a non-match for a non-anchored pattern. For example, if
133 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
134 "m", where previously it did not return a name. A side effect of this
135 change is that for partial matches, the last encountered mark name is
136 returned, as for non matches. A number of tests that were previously not
137 Perl-compatible have been moved into the Perl-compatible test files. The
138 refactoring has had the pleasing side effect of removing one argument from
139 the match() function, thus reducing its stack requirements.
140
141 19. If the /S+ option was used in pcretest to study a pattern using JIT,
142 subsequent uses of /S (without +) incorrectly behaved like /S+.
143
144 21. Retrieve executable code size support for the JIT compiler and fixing
145 some warnings.
146
147 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
148 not work when the shorter character appeared right at the end of the
149 subject string.
150
151 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
152 systems.
153
154 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
155 output it when the /M option is used in pcretest.
156
157 25. The CheckMan script was not being included in the distribution. Also, added
158 an explicit "perl" to run Perl scripts from the PrepareRelease script
159 because this is reportedly needed in Windows.
160
161 26. If study data was being save in a file and studying had not found a set of
162 "starts with" bytes for the pattern, the data written to the file (though
163 never used) was taken from uninitialized memory and so caused valgrind to
164 complain.
165
166 27. Updated RunTest.bat as provided by Sheri Pierce.
167
168 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
169
170 29. Computation of memory usage for the table of capturing group names was
171 giving an unnecessarily large value.
172
173
174 Version 8.20 21-Oct-2011
175 ------------------------
176
177 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
178 a POSIX class. After further experiments with Perl, which convinced me that
179 Perl has bugs and confusions, a closing square bracket is no longer allowed
180 in a POSIX name. This bug also affected patterns with classes that started
181 with full stops.
182
183 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
184 captured substring, but while checking the failing first alternative,
185 substring 1 is temporarily captured. If the output vector supplied to
186 pcre_exec() was not big enough for this capture, the yield of the function
187 was still zero ("insufficient space for captured substrings"). This cannot
188 be totally fixed without adding another stack variable, which seems a lot
189 of expense for a edge case. However, I have improved the situation in cases
190 such as /(a)(b)x|abc/ matched against "abc", where the return code
191 indicates that fewer than the maximum number of slots in the ovector have
192 been set.
193
194 3. Related to (2) above: when there are more back references in a pattern than
195 slots in the output vector, pcre_exec() uses temporary memory during
196 matching, and copies in the captures as far as possible afterwards. It was
197 using the entire output vector, but this conflicts with the specification
198 that only 2/3 is used for passing back captured substrings. Now it uses
199 only the first 2/3, for compatibility. This is, of course, another edge
200 case.
201
202 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
203 main code base, and can be used by building with --enable-jit. When this is
204 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
205 runtime --no-jit option is given.
206
207 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
208 ovector, the return from the function was zero, implying that there were
209 other matches that did not fit. The correct "exactly full" value is now
210 returned.
211
212 6. If a subpattern that was called recursively or as a subroutine contained
213 (*PRUNE) or any other control that caused it to give a non-standard return,
214 invalid errors such as "Error -26 (nested recursion at the same subject
215 position)" or even infinite loops could occur.
216
217 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
218 computing the minimum length on reaching *ACCEPT, and so ended up with the
219 wrong value of 1 rather than 0. Further investigation indicates that
220 computing a minimum subject length in the presence of *ACCEPT is difficult
221 (think back references, subroutine calls), and so I have changed the code
222 so that no minimum is registered for a pattern that contains *ACCEPT.
223
224 8. If (*THEN) was present in the first (true) branch of a conditional group,
225 it was not handled as intended. [But see 16 below.]
226
227 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
228 Sheri Pierce.
229
230 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
231 the first byte in a match must be "a".
232
233 11. Change 17 for 8.13 increased the recursion depth for patterns like
234 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
235 pattern contains any instances of (*THEN). If it does not, the old
236 optimizations are restored. It would be nice to do this on a per-group
237 basis, but at the moment that is not feasible.
238
239 12. In some environments, the output of pcretest -C is CRLF terminated. This
240 broke RunTest's code that checks for the link size. A single white space
241 character after the value is now allowed for.
242
243 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
244 For "fr", it uses the Windows-specific input and output files.
245
246 14. If (*THEN) appeared in a group that was called recursively or as a
247 subroutine, it did not work as intended. [But see next item.]
248
249 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
250 pattern fragments (but not containing any | characters). If A and B are
251 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
252 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
253 D. In other words, Perl considers parentheses that do not contain any |
254 characters to be part of a surrounding alternative, whereas PCRE was
255 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
256 differently. PCRE now behaves in the same way as Perl, except in the case
257 of subroutine/recursion calls such as (?1) which have in any case always
258 been different (but PCRE had them first :-).
259
260 16. Related to 15 above: Perl does not treat the | in a conditional group as
261 creating alternatives. Such a group is treated in the same way as an
262 ordinary group without any | characters when processing (*THEN). PCRE has
263 been changed to match Perl's behaviour.
264
265 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
266 RunGrepTest script failed.
267
268 18. Change 22 for version 13 caused atomic groups to use more stack. This is
269 inevitable for groups that contain captures, but it can lead to a lot of
270 stack use in large patterns. The old behaviour has been restored for atomic
271 groups that do not contain any capturing parentheses.
272
273 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
274 suppress the check for a minimum subject length at run time. (If it was
275 given to pcre_exec() or pcre_dfa_exec() it did work.)
276
277 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
278 fail to work when decoding hex characters in data strings in EBCDIC
279 environments.
280
281 21. It appears that in at least one Mac OS environment, the isxdigit() function
282 is implemented as a macro that evaluates to its argument more than once,
283 contravening the C 90 Standard (I haven't checked a later standard). There
284 was an instance in pcretest which caused it to go wrong when processing
285 \x{...} escapes in subject strings. The has been rewritten to avoid using
286 things like p++ in the argument of isxdigit().
287
288
289 Version 8.13 16-Aug-2011
290 ------------------------
291
292 1. The Unicode data tables have been updated to Unicode 6.0.0.
293
294 2. Two minor typos in pcre_internal.h have been fixed.
295
296 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
297 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
298 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
299
300 4. There were a number of related bugs in the code for matching backrefences
301 caselessly in UTF-8 mode when codes for the characters concerned were
302 different numbers of bytes. For example, U+023A and U+2C65 are an upper
303 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
304 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
305 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
306 2-byte code at the end of the subject (it thought there wasn't enough data
307 left).
308
309 5. Comprehensive information about what went wrong is now returned by
310 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
311 as the output vector has at least 2 elements. The offset of the start of
312 the failing character and a reason code are placed in the vector.
313
314 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
315 now returned is for the first byte of the failing character, instead of the
316 last byte inspected. This is an incompatible change, but I hope it is small
317 enough not to be a problem. It makes the returned offset consistent with
318 pcre_exec() and pcre_dfa_exec().
319
320 7. pcretest now gives a text phrase as well as the error number when
321 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
322 failure, the offset and reason code are output.
323
324 8. When \R was used with a maximizing quantifier it failed to skip backwards
325 over a \r\n pair if the subsequent match failed. Instead, it just skipped
326 back over a single character (\n). This seems wrong (because it treated the
327 two characters as a single entity when going forwards), conflicts with the
328 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
329 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
330 has been changed.
331
332 9. Some internal refactoring has changed the processing so that the handling
333 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
334 time (the PCRE_DOTALL option was changed this way some time ago: version
335 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
336 which was always a bit of a fudge. It also means that there is one less
337 argument for the match() function, which reduces its stack requirements
338 slightly. This change also fixes an incompatibility with Perl: the pattern
339 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
340
341 10. More internal refactoring has drastically reduced the number of recursive
342 calls to match() for possessively repeated groups such as (abc)++ when
343 using pcre_exec().
344
345 11. While implementing 10, a number of bugs in the handling of groups were
346 discovered and fixed:
347
348 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
349 (a|)*(?1) gave a compile-time internal error.
350 ((a|)+)+ did not notice that the outer group could match an empty string.
351 (^a|^)+ was not marked as anchored.
352 (.*a|.*)+ was not marked as matching at start or after a newline.
353
354 12. Yet more internal refactoring has removed another argument from the match()
355 function. Special calls to this function are now indicated by setting a
356 value in a variable in the "match data" data block.
357
358 13. Be more explicit in pcre_study() instead of relying on "default" for
359 opcodes that mean there is no starting character; this means that when new
360 ones are added and accidentally left out of pcre_study(), testing should
361 pick them up.
362
363 14. The -s option of pcretest has been documented for ages as being an old
364 synonym of -m (show memory usage). I have changed it to mean "force study
365 for every regex", that is, assume /S for every regex. This is similar to -i
366 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
367 using it. It makes it easier to run collections of tests with and without
368 study enabled, and thereby test pcre_study() more easily. All the standard
369 tests are now run with and without -s (but some patterns can be marked as
370 "never study" - see 20 below).
371
372 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
373 restoration of the capturing data to the outer values was not happening
374 correctly.
375
376 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
377 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
378 pattern had matched an empty string, and so incorrectly returned a no
379 match.
380
381 17. There was optimizing code for the last branch of non-capturing parentheses,
382 and also for the obeyed branch of a conditional subexpression, which used
383 tail recursion to cut down on stack usage. Unfortunately, now that there is
384 the possibility of (*THEN) occurring in these branches, tail recursion is
385 no longer possible because the return has to be checked for (*THEN). These
386 two optimizations have therefore been removed. [But see 8.20/11 above.]
387
388 18. If a pattern containing \R was studied, it was assumed that \R always
389 matched two bytes, thus causing the minimum subject length to be
390 incorrectly computed because \R can also match just one byte.
391
392 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
393 was incorrectly computed.
394
395 20. If /S is present twice on a test pattern in pcretest input, it now
396 *disables* studying, thereby overriding the use of -s on the command line
397 (see 14 above). This is necessary for one or two tests to keep the output
398 identical in both cases.
399
400 21. When (*ACCEPT) was used in an assertion that matched an empty string and
401 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
402
403 22. When an atomic group that contained a capturing parenthesis was
404 successfully matched, but the branch in which it appeared failed, the
405 capturing was not being forgotten if a higher numbered group was later
406 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
407 group 1 to "a", when in fact it should be unset. This applied to multi-
408 branched capturing and non-capturing groups, repeated or not, and also to
409 positive assertions (capturing in negative assertions does not happen
410 in PCRE) and also to nested atomic groups.
411
412 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
413 subject after a captured substring, to make it easier to tell which of a
414 number of identical substrings has been captured.
415
416 24. The way atomic groups are processed by pcre_exec() has been changed so that
417 if they are repeated, backtracking one repetition now resets captured
418 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
419 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
420 "aaa". Previously, it would have been "a". As part of this code
421 refactoring, the way recursive calls are handled has also been changed.
422
423 25. If an assertion condition captured any substrings, they were not passed
424 back unless some other capturing happened later. For example, if
425 (?(?=(a))a) was matched against "a", no capturing was returned.
426
427 26. When studying a pattern that contained subroutine calls or assertions,
428 the code for finding the minimum length of a possible match was handling
429 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
430 group 1 called group 2 while simultaneously a separate group 2 called group
431 1). A stack overflow occurred in this case. I have fixed this by limiting
432 the recursion depth to 10.
433
434 27. Updated RunTest.bat in the distribution to the version supplied by Tom
435 Fortmann. This supports explicit test numbers on the command line, and has
436 argument validation and error reporting.
437
438 28. An instance of \X with an unlimited repeat could fail if at any point the
439 first character it looked at was a mark character.
440
441 29. Some minor code refactoring concerning Unicode properties and scripts
442 should reduce the stack requirement of match() slightly.
443
444 30. Added the '=' option to pcretest to check the setting of unused capturing
445 slots at the end of the pattern, which are documented as being -1, but are
446 not included in the return count.
447
448 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
449 compiled something random. Now it gives a compile-time error (as does
450 Perl).
451
452 32. A *MARK encountered during the processing of a positive assertion is now
453 recorded and passed back (compatible with Perl).
454
455 33. If --only-matching or --colour was set on a pcregrep call whose pattern
456 had alternative anchored branches, the search for a second match in a line
457 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
458 matched the line "0102" twice. The same bug affected patterns that started
459 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
460 twice.
461
462 34. Previously, PCRE did not allow quantification of assertions. However, Perl
463 does, and because of capturing effects, quantifying parenthesized
464 assertions may at times be useful. Quantifiers are now allowed for
465 parenthesized assertions.
466
467 35. A minor code tidy in pcre_compile() when checking options for \R usage.
468
469 36. \g was being checked for fancy things in a character class, when it should
470 just be a literal "g".
471
472 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
473 appearance of a nested POSIX class supersedes an apparent external class.
474 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
475 unescaped square brackets may also appear as part of class names. For
476 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
477 more like Perl. (But see 8.20/1 above.)
478
479 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
480 was because it thought it was \N{name}, which is not supported).
481
482 39. Add minix to OS list not supporting the -S option in pcretest.
483
484 40. PCRE tries to detect cases of infinite recursion at compile time, but it
485 cannot analyze patterns in sufficient detail to catch mutual recursions
486 such as ((?1))((?2)). There is now a runtime test that gives an error if a
487 subgroup is called recursively as a subpattern for a second time at the
488 same position in the subject string. In previous releases this might have
489 been caught by the recursion limit, or it might have run out of stack.
490
491 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
492 happen only once. PCRE was, however incorrectly giving a compile time error
493 "recursive call could loop indefinitely" because it cannot analyze the
494 pattern in sufficient detail. The compile time test no longer happens when
495 PCRE is compiling a conditional subpattern, but actual runaway loops are
496 now caught at runtime (see 40 above).
497
498 42. It seems that Perl allows any characters other than a closing parenthesis
499 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
500 has been changed to be the same.
501
502 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
503 as not to get warnings when autogen.sh is called. Also changed
504 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
505
506 44. To help people who use pcregrep to scan files containing exceedingly long
507 lines, the following changes have been made:
508
509 (a) The default value of the buffer size parameter has been increased from
510 8K to 20K. (The actual buffer used is three times this size.)
511
512 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
513 PCRE is built.
514
515 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
516 to be set at run time.
517
518 (d) Numerical values in pcregrep options can be followed by K or M, for
519 example --buffer-size=50K.
520
521 (e) If a line being scanned overflows pcregrep's buffer, an error is now
522 given and the return code is set to 2.
523
524 45. Add a pointer to the latest mark to the callout data block.
525
526 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
527 partial match of an empty string instead of no match. This was specific to
528 the use of ".".
529
530 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
531 complete match instead of a partial match. This bug was dependent on both
532 the PCRE_UTF8 and PCRE_DOTALL options being set.
533
534 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
535 starting byte set, because \b was not being ignored.
536
537
538 Version 8.12 15-Jan-2011
539 ------------------------
540
541 1. Fixed some typos in the markup of the man pages, and wrote a script that
542 checks for such things as part of the documentation building process.
543
544 2. On a big-endian 64-bit system, pcregrep did not correctly process the
545 --match-limit and --recursion-limit options (added for 8.11). In
546 particular, this made one of the standard tests fail. (The integer value
547 went into the wrong half of a long int.)
548
549 3. If the --colour option was given to pcregrep with -v (invert match), it
550 did strange things, either producing crazy output, or crashing. It should,
551 of course, ignore a request for colour when reporting lines that do not
552 match.
553
554 4. Another pcregrep bug caused similar problems if --colour was specified with
555 -M (multiline) and the pattern match finished with a line ending.
556
557 5. In pcregrep, when a pattern that ended with a literal newline sequence was
558 matched in multiline mode, the following line was shown as part of the
559 match. This seems wrong, so I have changed it.
560
561 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
562 the check for further matches in the same line (so they could be coloured)
563 to overrun the end of the current line. If another match was found, it was
564 incorrectly shown (and then shown again when found in the next line).
565
566 7. If pcregrep was compiled under Windows, there was a reference to the
567 function pcregrep_exit() before it was defined. I am assuming this was
568 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
569 reported by a user. I've moved the definition above the reference.
570
571
572 Version 8.11 10-Dec-2010
573 ------------------------
574
575 1. (*THEN) was not working properly if there were untried alternatives prior
576 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
577 backtracked to try for "b" instead of moving to the next alternative branch
578 at the same level (in this case, to look for "c"). The Perl documentation
579 is clear that when (*THEN) is backtracked onto, it goes to the "next
580 alternative in the innermost enclosing group".
581
582 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
583 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
584 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
585 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
586 (*THEN).
587
588 3. If \s appeared in a character class, it removed the VT character from
589 the class, even if it had been included by some previous item, for example
590 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
591 of \s, but is part of the POSIX "space" class.)
592
593 4. A partial match never returns an empty string (because you can always
594 match an empty string at the end of the subject); however the checking for
595 an empty string was starting at the "start of match" point. This has been
596 changed to the "earliest inspected character" point, because the returned
597 data for a partial match starts at this character. This means that, for
598 example, /(?<=abc)def/ gives a partial match for the subject "abc"
599 (previously it gave "no match").
600
601 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
602 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
603 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
604 has an implication that the given string is incomplete (because a partial
605 match is preferred over a full match). For this reason, these items now
606 give a partial match in this situation. [Aside: previously, the one case
607 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
608 match rather than a full match, which was wrong by the old rules, but is
609 now correct.]
610
611 6. There was a bug in the handling of #-introduced comments, recognized when
612 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
613 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
614 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
615 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
616 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
617 places in pcre_compile().
618
619 7. Related to (6) above, when pcre_compile() was skipping #-introduced
620 comments when looking ahead for named forward references to subpatterns,
621 the only newline sequence it recognized was NL. It now handles newlines
622 according to the set newline convention.
623
624 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
625 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
626 cater for a lack of strerror(). These oversights have been fixed.
627
628 9. Added --match-limit and --recursion-limit to pcregrep.
629
630 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
631
632 11. When the -o option was used, pcregrep was setting a return code of 1, even
633 when matches were found, and --line-buffered was not being honoured.
634
635 12. Added an optional parentheses number to the -o and --only-matching options
636 of pcregrep.
637
638 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
639 can match an empty string. The code to do it in pcretest and pcredemo
640 needed fixing:
641
642 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
643 only one byte after an empty string match just before CRLF (this case
644 just got forgotten; "any" and "anycrlf" were OK).
645
646 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
647 mode when an empty string match preceded an ASCII character followed by
648 a non-ASCII character. (The code for advancing by one character rather
649 than one byte was nonsense.)
650
651 (c) The pcredemo.c sample program did not have any code at all to handle
652 the cases when CRLF is a valid newline sequence.
653
654 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
655 as a starting offset was within the subject string. There is now a new
656 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
657 negative or greater than the length of the string. In order to test this,
658 pcretest is extended to allow the setting of negative starting offsets.
659
660 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
661 starting offset points to the beginning of a UTF-8 character was
662 unnecessarily clumsy. I tidied it up.
663
664 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
665 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
666
667 17. Nobody had reported that the --include_dir option, which was added in
668 release 7.7 should have been called --include-dir (hyphen, not underscore)
669 for compatibility with GNU grep. I have changed it to --include-dir, but
670 left --include_dir as an undocumented synonym, and the same for
671 --exclude-dir, though that is not available in GNU grep, at least as of
672 release 2.5.4.
673
674 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
675 characters from a string of bytes) have been redefined so as not to use
676 loops, in order to improve performance in some environments. At the same
677 time, I abstracted some of the common code into auxiliary macros to save
678 repetition (this should not affect the compiled code).
679
680 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
681 compile-time error is now given if \c is not followed by an ASCII
682 character, that is, a byte less than 128. (In EBCDIC mode, the code is
683 different, and any byte value is allowed.)
684
685 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
686 START_OPTIMIZE option, which is now allowed at compile time - but just
687 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
688 to pcregrep and other applications that have no direct access to PCRE
689 options. The new /Y option in pcretest sets this option when calling
690 pcre_compile().
691
692 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
693 back references. Groups containing recursive back references were forced to
694 be atomic by that change, but in the case of named groups, the amount of
695 memory required was incorrectly computed, leading to "Failed: internal
696 error: code overflow". This has been fixed.
697
698 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
699 pcretest.c, to avoid build problems in some Borland environments.
700
701
702 Version 8.10 25-Jun-2010
703 ------------------------
704
705 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
706 THEN.
707
708 2. (*ACCEPT) was not working when inside an atomic group.
709
710 3. Inside a character class, \B is treated as a literal by default, but
711 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
712 causes the error). The code is unchanged, but I tidied the documentation.
713
714 4. Inside a character class, PCRE always treated \R and \X as literals,
715 whereas Perl faults them if its -w option is set. I have changed PCRE so
716 that it faults them when PCRE_EXTRA is set.
717
718 5. Added support for \N, which always matches any character other than
719 newline. (It is the same as "." when PCRE_DOTALL is not set.)
720
721 6. When compiling pcregrep with newer versions of gcc which may have
722 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
723 declared with attribute warn_unused_result" were given. Just casting the
724 result to (void) does not stop the warnings; a more elaborate fudge is
725 needed. I've used a macro to implement this.
726
727 7. Minor change to pcretest.c to avoid a compiler warning.
728
729 8. Added four artifical Unicode properties to help with an option to make
730 \s etc use properties (see next item). The new properties are: Xan
731 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
732
733 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
734 use Unicode properties. (*UCP) at the start of a pattern can be used to set
735 this option. Modified pcretest to add /W to test this facility. Added
736 REG_UCP to make it available via the POSIX interface.
737
738 10. Added --line-buffered to pcregrep.
739
740 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
741 studied, and the match started with a letter with a code point greater than
742 127 whose first byte was different to the first byte of the other case of
743 the letter, the other case of this starting letter was not recognized
744 (#976).
745
746 12. If a pattern that was studied started with a repeated Unicode property
747 test, for example, \p{Nd}+, there was the theoretical possibility of
748 setting up an incorrect bitmap of starting bytes, but fortunately it could
749 not have actually happened in practice until change 8 above was made (it
750 added property types that matched character-matching opcodes).
751
752 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
753 possible starting bytes for non-anchored patterns.
754
755 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
756 \R, and also a number of cases that involve Unicode properties, both
757 explicit and implicit when PCRE_UCP is set.
758
759 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
760 input, it could crash or give wrong results if characters with values
761 greater than 0xc0 were present in the subject string. (Detail: it assumed
762 UTF-8 input when processing these items.)
763
764 16. Added a lot of (int) casts to avoid compiler warnings in systems where
765 size_t is 64-bit (#991).
766
767 17. Added a check for running out of memory when PCRE is compiled with
768 --disable-stack-for-recursion (#990).
769
770 18. If the last data line in a file for pcretest does not have a newline on
771 the end, a newline was missing in the output.
772
773 19. The default pcre_chartables.c file recognizes only ASCII characters (values
774 less than 128) in its various bitmaps. However, there is a facility for
775 generating tables according to the current locale when PCRE is compiled. It
776 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
777 characters, are recognized by isspace() and therefore were getting set in
778 these tables, and indeed these tables seem to approximate to ISO 8859. This
779 caused a problem in UTF-8 mode when pcre_study() was used to create a list
780 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
781 which of course cannot start UTF-8 characters. I have changed the code so
782 that only real ASCII characters (less than 128) and the correct starting
783 bytes for UTF-8 encodings are set for characters greater than 127 when in
784 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
785 altogether.)
786
787 20. Added the /T option to pcretest so as to be able to run tests with non-
788 standard character tables, thus making it possible to include the tests
789 used for 19 above in the standard set of tests.
790
791 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
792 reference to a subpattern the other side of a comment that contains an
793 opening parenthesis caused either an internal compiling error, or a
794 reference to the wrong subpattern.
795
796
797 Version 8.02 19-Mar-2010
798 ------------------------
799
800 1. The Unicode data tables have been updated to Unicode 5.2.0.
801
802 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
803 configured.
804
805 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
806 original author of that file, following a query about its status.
807
808 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
809 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
810
811 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
812 quantifier applied to a forward-referencing subroutine call, could compile
813 incorrect code or give the error "internal error: previously-checked
814 referenced subpattern not found".
815
816 6. Both MS Visual Studio and Symbian OS have problems with initializing
817 variables to point to external functions. For these systems, therefore,
818 pcre_malloc etc. are now initialized to local functions that call the
819 relevant global functions.
820
821 7. There were two entries missing in the vectors called coptable and poptable
822 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
823 I've fixed the data, and added a kludgy way of testing at compile time that
824 the lengths are correct (equal to the number of opcodes).
825
826 8. Following on from 7, I added a similar kludge to check the length of the
827 eint vector in pcreposix.c.
828
829 9. Error texts for pcre_compile() are held as one long string to avoid too
830 much relocation at load time. To find a text, the string is searched,
831 counting zeros. There was no check for running off the end of the string,
832 which could happen if a new error number was added without updating the
833 string.
834
835 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
836
837 11. \K was not working if it appeared in an atomic group or in a group that
838 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
839 \K is "not well defined" if used in an assertion. PCRE now accepts it if
840 the assertion is positive, but not if it is negative.
841
842 12. Change 11 fortuitously reduced the size of the stack frame used in the
843 "match()" function of pcre_exec.c by one pointer. Forthcoming
844 implementation of support for (*MARK) will need an extra pointer on the
845 stack; I have reserved it now, so that the stack frame size does not
846 decrease.
847
848 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
849 item in branch that calls a recursion is a subroutine call - as in the
850 second branch in the above example - was incorrectly given the compile-
851 time error "recursive call could loop indefinitely" because pcre_compile()
852 was not correctly checking the subroutine for matching a non-empty string.
853
854 14. The checks for overrunning compiling workspace could trigger after an
855 overrun had occurred. This is a "should never occur" error, but it can be
856 triggered by pathological patterns such as hundreds of nested parentheses.
857 The checks now trigger 100 bytes before the end of the workspace.
858
859 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
860
861
862 Version 8.01 19-Jan-2010
863 ------------------------
864
865 1. If a pattern contained a conditional subpattern with only one branch (in
866 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
867 computed the wrong minimum data length (which is of course zero for such
868 subpatterns). This could cause incorrect "no match" results.
869
870 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
871 the pattern is reset in the first branch, pcre_compile() failed with
872 "internal error: code overflow at offset...". This happened only when
873 the reset was to the original external option setting. (An optimization
874 abstracts leading options settings into an external setting, which was the
875 cause of this.)
876
877 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
878 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
879 assertion pattern did not match (meaning that the assertion was true), it
880 was incorrectly treated as false if the SKIP had been reached during the
881 matching. This also applied to assertions used as conditions.
882
883 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
884 assertion subpattern, including such a pattern used as a condition,
885 unpredictable results occurred, instead of the error return
886 PCRE_ERROR_DFA_UITEM.
887
888 5. The C++ GlobalReplace function was not working like Perl for the special
889 situation when an empty string is matched. It now does the fancy magic
890 stuff that is necessary.
891
892 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
893 removed. (These were left over from very, very early versions of PCRE.)
894
895 7. Some cosmetic changes to the code to make life easier when compiling it
896 as part of something else:
897
898 (a) Change DEBUG to PCRE_DEBUG.
899
900 (b) In pcre_compile(), rename the member of the "branch_chain" structure
901 called "current" as "current_branch", to prevent a collision with the
902 Linux macro when compiled as a kernel module.
903
904 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
905 prevent a collision with the Linux macro when compiled as a kernel
906 module.
907
908 8. In pcre_compile() there are some checks for integer overflows that used to
909 cast potentially large values to (double). This has been changed to that
910 when building, a check for int64_t is made, and if it is found, it is used
911 instead, thus avoiding the use of floating point arithmetic. (There is no
912 other use of FP in PCRE.) If int64_t is not found, the fallback is to
913 double.
914
915 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
916 2005 (difference between two addresses compared to an unsigned value).
917
918 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
919 custom one, because of the following reported problem in Windows:
920
921 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
922 under Win32.
923 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
924 therefore missing the function definition.
925 - The compiler thus generates a "C" signature for the test function.
926 - The linker fails to find the "C" function.
927 - PCRE fails to configure if asked to do so against libbz2.
928
929 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
930 messages were output:
931
932 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
933 rerunning libtoolize, to keep the correct libtool macros in-tree.
934 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
935
936 I have done both of these things.
937
938 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
939 most of the time, it *can* run out if it is given a pattern that contains a
940 runaway infinite recursion. I updated the discussion in the pcrestack man
941 page.
942
943 13. Now that we have gone to the x.xx style of version numbers, the minor
944 version may start with zero. Using 08 or 09 is a bad idea because users
945 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
946 interpreted as invalid octal numbers. I've updated the previous comment in
947 configure.ac, and also added a check that gives an error if 08 or 09 are
948 used.
949
950 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
951 causing partial matching to fail when the end of the subject matched \W
952 in a UTF-8 pattern where \W was quantified with a minimum of 3.
953
954 15. There were some discrepancies between the declarations in pcre_internal.h
955 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
956 their definitions. The declarations used "const uschar *" and the
957 definitions used USPTR. Even though USPTR is normally defined as "const
958 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
959 reported that: "This difference in casting confuses some C++ compilers, for
960 example, SunCC recognizes above declarations as different functions and
961 generates broken code for hbpcre." I have changed the declarations to use
962 USPTR.
963
964 16. GNU libtool is named differently on some systems. The autogen.sh script now
965 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
966 (FreeBSD).
967
968 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
969 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
970 comment: "Figure out how to create a longlong from a string: strtoll and
971 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
972 instance, but it only takes 2 args instead of 3!"
973
974 18. A subtle bug concerned with back references has been fixed by a change of
975 specification, with a corresponding code fix. A pattern such as
976 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
977 refers, was giving matches when it shouldn't. For example, xa=xaaa would
978 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
979 same bug. Such groups have to be quantified to be useful, or contained
980 inside another quantified group. (If there's no repetition, the reference
981 can never match.) The problem arises because, having left the group and
982 moved on to the rest of the pattern, a later failure that backtracks into
983 the group uses the captured value from the final iteration of the group
984 rather than the correct earlier one. I have fixed this in PCRE by forcing
985 any group that contains a reference to itself to be an atomic group; that
986 is, there cannot be any backtracking into it once it has completed. This is
987 similar to recursive and subroutine calls.
988
989
990 Version 8.00 19-Oct-09
991 ----------------------
992
993 1. The table for translating pcre_compile() error codes into POSIX error codes
994 was out-of-date, and there was no check on the pcre_compile() error code
995 being within the table. This could lead to an OK return being given in
996 error.
997
998 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
999 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1000 in a Windows environment.
1001
1002 3. The pcregrep --count option prints the count for each file even when it is
1003 zero, as does GNU grep. However, pcregrep was also printing all files when
1004 --files-with-matches was added. Now, when both options are given, it prints
1005 counts only for those files that have at least one match. (GNU grep just
1006 prints the file name in this circumstance, but including the count seems
1007 more useful - otherwise, why use --count?) Also ensured that the
1008 combination -clh just lists non-zero counts, with no names.
1009
1010 4. The long form of the pcregrep -F option was incorrectly implemented as
1011 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1012 but it seems right to fix it, and I didn't think it was worth preserving
1013 the old behaviour.
1014
1015 5. The command line items --regex=pattern and --regexp=pattern were not
1016 recognized by pcregrep, which required --regex pattern or --regexp pattern
1017 (with a space rather than an '='). The man page documented the '=' forms,
1018 which are compatible with GNU grep; these now work.
1019
1020 6. No libpcreposix.pc file was created for pkg-config; there was just
1021 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1022
1023 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1024 when UCP support is not needed, by modifying the Python script that
1025 generates it from Unicode data files. This should not matter if the module
1026 is correctly used as a library, but I received one complaint about 50K of
1027 unwanted data. My guess is that the person linked everything into his
1028 program rather than using a library. Anyway, it does no harm.
1029
1030 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1031 was a minimum greater than 1 for a wide character in a possessive
1032 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1033 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1034 character. Chaos in the form of incorrect output or a compiling loop could
1035 result.
1036
1037 9. The restrictions on what a pattern can contain when partial matching is
1038 requested for pcre_exec() have been removed. All patterns can now be
1039 partially matched by this function. In addition, if there are at least two
1040 slots in the offset vector, the offset of the earliest inspected character
1041 for the match and the offset of the end of the subject are set in them when
1042 PCRE_ERROR_PARTIAL is returned.
1043
1044 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1045 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1046 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1047 and may be more useful for multi-segment matching.
1048
1049 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1050 used to be given if ever the end of the subject was reached; now it is
1051 given only if matching could not proceed because another character was
1052 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1053 string "Z", which now yields "no match" instead of "partial match". In the
1054 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1055 final character ended with (*FAIL).
1056
1057 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1058 if the pattern had a "must contain" character that was already found in the
1059 earlier partial match, unless partial matching was again requested. For
1060 example, with the pattern /dog.(body)?/, the "must contain" character is
1061 "g". If the first part-match was for the string "dog", restarting with
1062 "sbody" failed. This bug has been fixed.
1063
1064 13. The string returned by pcre_dfa_exec() after a partial match has been
1065 changed so that it starts at the first inspected character rather than the
1066 first character of the match. This makes a difference only if the pattern
1067 starts with a lookbehind assertion or \b or \B (\K is not supported by
1068 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1069 matching functions compatible, and I think it's the right thing to do.
1070
1071 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1072 so that the demonstration program is easily available in environments where
1073 PCRE has not been installed from source.
1074
1075 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1076 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1077 library.
1078
1079 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1080 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1081 is not the first non-POSIX option to be added. Clearly some people find
1082 these options useful.
1083
1084 17. If a caller to the POSIX matching function regexec() passes a non-zero
1085 value for nmatch with a NULL value for pmatch, the value of
1086 nmatch is forced to zero.
1087
1088 18. RunGrepTest did not have a test for the availability of the -u option of
1089 the diff command, as RunTest does. It now checks in the same way as
1090 RunTest, and also checks for the -b option.
1091
1092 19. If an odd number of negated classes containing just a single character
1093 interposed, within parentheses, between a forward reference to a named
1094 subpattern and the definition of the subpattern, compilation crashed with
1095 an internal error, complaining that it could not find the referenced
1096 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1097 [The bug was that it was starting one character too far in when skipping
1098 over the character class, thus treating the ] as data rather than
1099 terminating the class. This meant it could skip too much.]
1100
1101 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1102 /g option in pcretest when the pattern contains \K, which makes it possible
1103 to have an empty string match not at the start, even when the pattern is
1104 anchored. Updated pcretest and pcredemo to use this option.
1105
1106 21. If the maximum number of capturing subpatterns in a recursion was greater
1107 than the maximum at the outer level, the higher number was returned, but
1108 with unset values at the outer level. The correct (outer level) value is
1109 now given.
1110
1111 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1112 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1113 make it do so. The string so far is captured, making this feature
1114 compatible with Perl.
1115
1116 23. The tests have been re-organized, adding tests 11 and 12, to make it
1117 possible to check the Perl 5.10 features against Perl 5.10.
1118
1119 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1120 pattern matches a fixed length string. PCRE did not allow this; now it
1121 does. Neither allows recursion.
1122
1123 25. I finally figured out how to implement a request to provide the minimum
1124 length of subject string that was needed in order to match a given pattern.
1125 (It was back references and recursion that I had previously got hung up
1126 on.) This code has now been added to pcre_study(); it finds a lower bound
1127 to the length of subject needed. It is not necessarily the greatest lower
1128 bound, but using it to avoid searching strings that are too short does give
1129 some useful speed-ups. The value is available to calling programs via
1130 pcre_fullinfo().
1131
1132 26. While implementing 25, I discovered to my embarrassment that pcretest had
1133 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1134 study optimizations had never been tested with that matching function.
1135 Oops. What is worse, even when it was passed study data, there was a bug in
1136 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1137 were also very few tests of studied patterns with pcre_dfa_exec().
1138
1139 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1140 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1141 on the other side of the coin, they are no longer allowed to have different
1142 names, because these cannot be distinguished in PCRE, and this has caused
1143 confusion. (This is a difference from Perl.)
1144
1145 28. When duplicate subpattern names are present (necessarily with different
1146 numbers, as required by 27 above), and a test is made by name in a
1147 conditional pattern, either for a subpattern having been matched, or for
1148 recursion in such a pattern, all the associated numbered subpatterns are
1149 tested, and the overall condition is true if the condition is true for any
1150 one of them. This is the way Perl works, and is also more like the way
1151 testing by number works.
1152
1153
1154 Version 7.9 11-Apr-09
1155 ---------------------
1156
1157 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1158 (pcretest), all targets were linked against these libraries. This included
1159 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1160 libraries. This caused unwanted dependencies to be created. This problem
1161 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1162 pcretest is linked with readline.
1163
1164 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1165 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1166 moved outside it again, because FALSE and TRUE are already defined in AIX,
1167 but BOOL is not.
1168
1169 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1170 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1171
1172 4. The pcregrep documentation said spaces were inserted as well as colons (or
1173 hyphens) following file names and line numbers when outputting matching
1174 lines. This is not true; no spaces are inserted. I have also clarified the
1175 wording for the --colour (or --color) option.
1176
1177 5. In pcregrep, when --colour was used with -o, the list of matching strings
1178 was not coloured; this is different to GNU grep, so I have changed it to be
1179 the same.
1180
1181 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1182 each matching line was coloured. Now it goes on to look for further matches
1183 of any of the test patterns, which is the same behaviour as GNU grep.
1184
1185 7. A pattern that could match an empty string could cause pcregrep to loop; it
1186 doesn't make sense to accept an empty string match in pcregrep, so I have
1187 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1188 seems to be how GNU grep behaves.
1189
1190 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1191 start or after a newline", because the conditional assertion was not being
1192 correctly handled. The rule now is that both the assertion and what follows
1193 in the first alternative must satisfy the test.
1194
1195 9. If auto-callout was enabled in a pattern with a conditional group whose
1196 condition was an assertion, PCRE could crash during matching, both with
1197 pcre_exec() and pcre_dfa_exec().
1198
1199 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1200 used for matching.
1201
1202 11. Unicode property support in character classes was not working for
1203 characters (bytes) greater than 127 when not in UTF-8 mode.
1204
1205 12. Added the -M command line option to pcretest.
1206
1207 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1208
1209 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1210
1211 16. Added comments and documentation about mis-use of no_arg in the C++
1212 wrapper.
1213
1214 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1215 from Martin Jerabek that uses macro names for all relevant character and
1216 string constants.
1217
1218 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1219 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1220 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1221 these, but not everybody uses configure.
1222
1223 19. A conditional group that had only one branch was not being correctly
1224 recognized as an item that could match an empty string. This meant that an
1225 enclosing group might also not be so recognized, causing infinite looping
1226 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1227 with the subject "ab", where knowledge that the repeated group can match
1228 nothing is needed in order to break the loop.
1229
1230 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1231 exec(), but without supplying a callout function, matching went wrong.
1232
1233 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1234 leak if the size of the offset vector was greater than 30. When the vector
1235 is smaller, the saved offsets during recursion go onto a local stack
1236 vector, but for larger vectors malloc() is used. It was failing to free
1237 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1238 error, in fact).
1239
1240 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1241 heapframe that is used only when UTF-8 support is enabled. This caused no
1242 problem, but was untidy.
1243
1244 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1245 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1246 included within another project.
1247
1248 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1249 slightly modified by me:
1250
1251 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1252 not building pcregrep.
1253
1254 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1255 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1256
1257 25. Forward references, both numeric and by name, in patterns that made use of
1258 duplicate group numbers, could behave incorrectly or give incorrect errors,
1259 because when scanning forward to find the reference group, PCRE was not
1260 taking into account the duplicate group numbers. A pattern such as
1261 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1262
1263 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1264 the feature of a custom pointer more persuasive (as requested by a user).
1265
1266 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1267 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1268 pre-defined.
1269
1270 28. Added support for (*UTF8) at the start of a pattern.
1271
1272 29. Arrange for flags added by the "release type" setting in CMake to be shown
1273 in the configuration summary.
1274
1275
1276 Version 7.8 05-Sep-08
1277 ---------------------
1278
1279 1. Replaced UCP searching code with optimized version as implemented for Ad
1280 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1281 stage table and inline lookup instead of a function, giving speed ups of 2
1282 to 5 times on some simple patterns that I tested. Permission was given to
1283 distribute the MultiStage2.py script that generates the tables (it's not in
1284 the tarball, but is in the Subversion repository).
1285
1286 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1287 scripts.
1288
1289 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1290 a group with a zero qualifier. The result of the study could be incorrect,
1291 or the function might crash, depending on the pattern.
1292
1293 4. Caseless matching was not working for non-ASCII characters in back
1294 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1295 It now works when Unicode Property Support is available.
1296
1297 5. In pcretest, an escape such as \x{de} in the data was always generating
1298 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1299 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1300 truncation.
1301
1302 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1303
1304 7. Added two (int) casts to pcregrep when printing the difference of two
1305 pointers, in case they are 64-bit values.
1306
1307 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1308 test 2 if it fails.
1309
1310 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1311 and a #define of that name to empty if it is not externally set. This is to
1312 allow users of MSVC to set it if necessary.
1313
1314 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1315 the convenience functions in the pcre_get.c source file.
1316
1317 11. An option change at the start of a pattern that had top-level alternatives
1318 could cause overwriting and/or a crash. This command provoked a crash in
1319 some environments:
1320
1321 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1322
1323 This potential security problem was recorded as CVE-2008-2371.
1324
1325 12. For a pattern where the match had to start at the beginning or immediately
1326 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1327 pcre_dfa_exec() could read past the end of the passed subject if there was
1328 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1329 pcretest so that it places the subject at the end of its malloc-ed buffer.
1330
1331 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1332 exec() might read past the end of the data buffer in UTF-8 mode.
1333
1334 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1335 the data contained the byte 0x85 as part of a UTF-8 character within its
1336 first line. This applied both to normal and DFA matching.
1337
1338 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1339 /^[^d]*?$/8 failed to match "abc".
1340
1341 16. Added a missing copyright notice to pcrecpp_internal.h.
1342
1343 17. Make it more clear in the documentation that values returned from
1344 pcre_exec() in ovector are byte offsets, not character counts.
1345
1346 18. Tidied a few places to stop certain compilers from issuing warnings.
1347
1348 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1349 supplied by Stefan Weber. I made a further small update for 7.8 because
1350 there is a change of source arrangements: the pcre_searchfuncs.c module is
1351 replaced by pcre_ucd.c.
1352
1353
1354 Version 7.7 07-May-08
1355 ---------------------
1356
1357 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1358 a string to a long long, pretend we don't even have a long long." This is
1359 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1360
1361 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1362 pre-7.6 versions, which defined a global no_arg variable instead of putting
1363 it in the RE class. (See also #8 below.)
1364
1365 3. Remove a line of dead code, identified by coverity and reported by Nuno
1366 Lopes.
1367
1368 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1369
1370 (1) The include/exclude patterns were being applied to the whole pathnames
1371 of files, instead of just to the final components.
1372
1373 (2) If there was more than one level of directory, the subdirectories were
1374 skipped unless they satisfied the include/exclude conditions. This is
1375 inconsistent with GNU grep (and could even be seen as contrary to the
1376 pcregrep specification - which I improved to make it absolutely clear).
1377 The action now is always to scan all levels of directory, and just
1378 apply the include/exclude patterns to regular files.
1379
1380 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1381 --exclude_dir in the tests to avoid scanning .svn directories.
1382
1383 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1384 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1385 doesn't support NULs in patterns.
1386
1387 7. Added some missing "const"s to declarations of static tables in
1388 pcre_compile.c and pcre_dfa_exec.c.
1389
1390 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1391 caused by fix #2 above. (Subsequently also a second patch to fix the
1392 first patch. And a third patch - this was a messy problem.)
1393
1394 9. Applied Craig's patch to remove the use of push_back().
1395
1396 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1397 matching function regexec().
1398
1399 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1400 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1401 references. PCRE supports relative numbers with this syntax (I don't think
1402 Oniguruma does).
1403
1404 12. Previously, a group with a zero repeat such as (...){0} was completely
1405 omitted from the compiled regex. However, this means that if the group
1406 was called as a subroutine from elsewhere in the pattern, things went wrong
1407 (an internal error was given). Such groups are now left in the compiled
1408 pattern, with a new opcode that causes them to be skipped at execution
1409 time.
1410
1411 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1412 to the way PCRE behaves:
1413
1414 (a) A lone ] character is dis-allowed (Perl treats it as data).
1415
1416 (b) A back reference to an unmatched subpattern matches an empty string
1417 (Perl fails the current match path).
1418
1419 (c) A data ] in a character class must be notated as \] because if the
1420 first data character in a class is ], it defines an empty class. (In
1421 Perl it is not possible to have an empty class.) The empty class []
1422 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1423 The negative empty class [^] matches any one character, independently
1424 of the DOTALL setting.
1425
1426 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1427 non-existent subpattern following a character class starting with ']' and
1428 containing () gave an internal compiling error instead of "reference to
1429 non-existent subpattern". Fortunately, when the pattern did exist, the
1430 compiled code was correct. (When scanning forwards to check for the
1431 existencd of the subpattern, it was treating the data ']' as terminating
1432 the class, so got the count wrong. When actually compiling, the reference
1433 was subsequently set up correctly.)
1434
1435 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1436 it was being rejected as not supported by pcre_dfa_exec(), even though
1437 other assertions are supported. I have made pcre_dfa_exec() support
1438 (*FAIL).
1439
1440 16. The implementation of 13c above involved the invention of a new opcode,
1441 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1442 cannot be changed at match time, I realized I could make a small
1443 improvement to matching performance by compiling OP_ALLANY instead of
1444 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1445 on the OP_ANY path.
1446
1447 17. Compiling pcretest on Windows with readline support failed without the
1448 following two fixes: (1) Make the unistd.h include conditional on
1449 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1450
1451 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1452 ncurses library to be included for pcretest when ReadLine support is
1453 requested, but also to allow for it to be overridden. This patch came from
1454 Daniel Bergström.
1455
1456 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1457 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1458 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1459 spotting this.
1460
1461
1462 Version 7.6 28-Jan-08
1463 ---------------------
1464
1465 1. A character class containing a very large number of characters with
1466 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1467 overflow.
1468
1469 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1470 HAVE_LONG_LONG is not defined.
1471
1472 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1473 bring them up to date and include new features. This patch includes:
1474
1475 - Fixed PH's badly added libz and libbz2 support.
1476 - Fixed a problem with static linking.
1477 - Added pcredemo. [But later removed - see 7 below.]
1478 - Fixed dftables problem and added an option.
1479 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1480 HAVE_LONG_LONG.
1481 - Added readline support for pcretest.
1482 - Added an listing of the option settings after cmake has run.
1483
1484 4. A user submitted a patch to Makefile that makes it easy to create
1485 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1486 Makefile.am that cause it to include this special target, without
1487 affecting anything else. Note that the same mingw target plus all
1488 the other distribution libraries and programs are now supported
1489 when configuring with CMake (see 6 below) instead of with
1490 Configure/Make.
1491
1492 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1493 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1494 exported in the Windows port". It has not yet been confirmed that the patch
1495 solves the problem, but it does no harm.
1496
1497 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1498 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1499 with CMake, and also correct the comment about stack recursion.
1500
1501 7. Remove the automatic building of pcredemo from the ./configure system and
1502 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1503 of a program that users should build themselves after PCRE is installed, so
1504 building it automatically is not really right. What is more, it gave
1505 trouble in some build environments.
1506
1507 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1508
1509
1510 Version 7.5 10-Jan-08
1511 ---------------------
1512
1513 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1514 values in parens when parsing an RE using the C++ wrapper."
1515
1516 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1517 Characters greater than 255 were excluded from the class instead of being
1518 included.
1519
1520 3. The same bug as (2) above applied to negated POSIX classes such as
1521 [:^space:].
1522
1523 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1524 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1525 I have changed it.
1526
1527 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1528 first named subpattern) and a construct such as (?&a) would reference the
1529 first named subpattern whose name started with "a" (in other words, the
1530 length check was missing). Both these problems are fixed. "Subpattern name
1531 expected" is now given for (?&) (a zero-length name), and this patch also
1532 makes it give the same error for \k'' (previously it complained that that
1533 was a reference to a non-existent subpattern).
1534
1535 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1536 this is right because (?- can be followed by option settings as well as by
1537 digits. I have, however, made the messages clearer.
1538
1539 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1540 than the number used in the conditional) now cause a compile-time error.
1541 This is actually not compatible with Perl, which accepts such patterns, but
1542 treats the conditional as always being FALSE (as PCRE used to), but it
1543 seems to me that giving a diagnostic is better.
1544
1545 8. Change "alphameric" to the more common word "alphanumeric" in comments
1546 and messages.
1547
1548 9. Fix two occurrences of "backslash" in comments that should have been
1549 "backspace".
1550
1551 10. Remove two redundant lines of code that can never be obeyed (their function
1552 was moved elsewhere).
1553
1554 11. The program that makes PCRE's Unicode character property table had a bug
1555 which caused it to generate incorrect table entries for sequences of
1556 characters that have the same character type, but are in different scripts.
1557 It amalgamated them into a single range, with the script of the first of
1558 them. In other words, some characters were in the wrong script. There were
1559 thirteen such cases, affecting characters in the following ranges:
1560
1561 U+002b0 - U+002c1
1562 U+0060c - U+0060d
1563 U+0061e - U+00612
1564 U+0064b - U+0065e
1565 U+0074d - U+0076d
1566 U+01800 - U+01805
1567 U+01d00 - U+01d77
1568 U+01d9b - U+01dbf
1569 U+0200b - U+0200f
1570 U+030fc - U+030fe
1571 U+03260 - U+0327f
1572 U+0fb46 - U+0fbb1
1573 U+10450 - U+1049d
1574
1575 12. The -o option (show only the matching part of a line) for pcregrep was not
1576 compatible with GNU grep in that, if there was more than one match in a
1577 line, it showed only the first of them. It now behaves in the same way as
1578 GNU grep.
1579
1580 13. If the -o and -v options were combined for pcregrep, it printed a blank
1581 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1582 does the same. The return code can be used to tell if there were any
1583 non-matching lines.
1584
1585 14. Added --file-offsets and --line-offsets to pcregrep.
1586
1587 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1588 infinitely looping recursion. The bug was that positive lookaheads were not
1589 being skipped when checking for a possible empty match (negative lookaheads
1590 and both kinds of lookbehind were skipped).
1591
1592 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1593 inclusion of <windows.h> to before rather than after the definition of
1594 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1595
1596 17. Specifying a possessive quantifier with a specific limit for a Unicode
1597 character property caused pcre_compile() to compile bad code, which led at
1598 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1599 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1600 caused the error; without that there was no problem.
1601
1602 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1603
1604 19. Added --enable-pcretest-libreadline.
1605
1606 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1607 RE::GlobalReplace(). As a result, the number of replacements returned was
1608 double what it should be. I removed one of the increments, but Craig sent a
1609 later patch that removed the other one (the right fix) and added unit tests
1610 that check the return values (which was not done before).
1611
1612 21. Several CMake things:
1613
1614 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1615 the names libpcre and libpcreposix, not just pcre and pcreposix.
1616
1617 (2) The above change means that pcretest and pcregrep are now correctly
1618 linked with the newly-built libraries, not previously installed ones.
1619
1620 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1621
1622 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1623 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1624 UTF-8 newline character). The key issue is that the pattern starts .*;
1625 this means that the match must be either at the beginning, or after a
1626 newline. The bug was in the code for advancing after a failed match and
1627 checking that the new position followed a newline. It was not taking
1628 account of UTF-8 characters correctly.
1629
1630 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1631 character classes. PCRE was not treating the sequence [:...:] as a
1632 character class unless the ... were all letters. Perl, however, seems to
1633 allow any characters between [: and :], though of course it rejects as
1634 unknown any "names" that contain non-letters, because all the known class
1635 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1636 for example, whereas PCRE did not - it did not recognize a POSIX character
1637 class. This seemed a bit dangerous, so the code has been changed to be
1638 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1639 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1640 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1641 Perl does, and where it didn't before.
1642
1643 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1644 Windows environments %n is disabled by default.
1645
1646
1647 Version 7.4 21-Sep-07
1648 ---------------------
1649
1650 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1651 means that a class such as [\s] counted as "explicit reference to CR or
1652 LF". That isn't really right - the whole point of the change was to try to
1653 help when there was an actual mention of one of the two characters. So now
1654 the change happens only if \r or \n (or a literal CR or LF) character is
1655 encountered.
1656
1657 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1658 of both had grown to the point where there were only 3 bits left.
1659 Fortunately, there was spare space in the data structure, and so I have
1660 moved the internal flags into a new 16-bit field to free up more option
1661 bits.
1662
1663 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1664 but did not set the internal JCHANGED flag - either of these is enough to
1665 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1666 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1667 start sets both bits.
1668
1669 4. Added options (at build time, compile time, exec time) to change \R from
1670 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1671
1672 5. doc/pcresyntax.html was missing from the distribution.
1673
1674 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1675 compatibility, even though it is no longer used.
1676
1677 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1678 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1679 windows.h file is present (where different names are used). [This was
1680 reversed later after testing - see 16 below.]
1681
1682 8. Changed all #include <config.h> to #include "config.h". There were also
1683 some further <pcre.h> cases that I changed to "pcre.h".
1684
1685 9. When pcregrep was used with the --colour option, it missed the line ending
1686 sequence off the lines that it output.
1687
1688 10. It was pointed out to me that arrays of string pointers cause lots of
1689 relocations when a shared library is dynamically loaded. A technique of
1690 using a single long string with a table of offsets can drastically reduce
1691 these. I have refactored PCRE in four places to do this. The result is
1692 dramatic:
1693
1694 Originally: 290
1695 After changing UCP table: 187
1696 After changing error message table: 43
1697 After changing table of "verbs" 36
1698 After changing table of Posix names 22
1699
1700 Thanks to the folks working on Gregex for glib for this insight.
1701
1702 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1703 unicode-properties was also set.
1704
1705 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1706
1707 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1708 checked only for CRLF.
1709
1710 14. Added casts to pcretest.c to avoid compiler warnings.
1711
1712 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1713
1714 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1715 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1716 entirely. This removes changes made in 7 above.
1717
1718 17. The CMake files have been updated, and there is now more information about
1719 building with CMake in the NON-UNIX-USE document.
1720
1721
1722 Version 7.3 28-Aug-07
1723 ---------------------
1724
1725 1. In the rejigging of the build system that eventually resulted in 7.1, the
1726 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1727 brackets there is not right, since it causes compilers to look for an
1728 installed pcre.h, not the version that is in the source that is being
1729 compiled (which of course may be different). I have changed it back to:
1730
1731 #include "pcre.h"
1732
1733 I have a vague recollection that the change was concerned with compiling in
1734 different directories, but in the new build system, that is taken care of
1735 by the VPATH setting the Makefile.
1736
1737 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1738 when the subject happened to end in the byte 0x85 (e.g. if the last
1739 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1740 characters but of course it shouldn't be taken as a newline when it is part
1741 of another character. The bug was that, for an unlimited repeat of . in
1742 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1743 characters when looking for a newline.
1744
1745 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1746
1747 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1748 in debug output.
1749
1750 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1751 long printing in the pcrecpp unittest when running under MinGW.
1752
1753 6. ESC_K was left out of the EBCDIC table.
1754
1755 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1756 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1757 limit also applies to "virtual nesting" when a pattern is recursive, and in
1758 this case 1000 isn't so big. I have been able to remove this limit at the
1759 expense of backing off one optimization in certain circumstances. Normally,
1760 when pcre_exec() would call its internal match() function recursively and
1761 immediately return the result unconditionally, it uses a "tail recursion"
1762 feature to save stack. However, when a subpattern that can match an empty
1763 string has an unlimited repetition quantifier, it no longer makes this
1764 optimization. That gives it a stack frame in which to save the data for
1765 checking that an empty string has been matched. Previously this was taken
1766 from the 1000-entry workspace that had been reserved. So now there is no
1767 explicit limit, but more stack is used.
1768
1769 8. Applied Daniel's patches to solve problems with the import/export magic
1770 syntax that is required for Windows, and which was going wrong for the
1771 pcreposix and pcrecpp parts of the library. These were overlooked when this
1772 problem was solved for the main library.
1773
1774 9. There were some crude static tests to avoid integer overflow when computing
1775 the size of patterns that contain repeated groups with explicit upper
1776 limits. As the maximum quantifier is 65535, the maximum group length was
1777 set at 30,000 so that the product of these two numbers did not overflow a
1778 32-bit integer. However, it turns out that people want to use groups that
1779 are longer than 30,000 bytes (though not repeat them that many times).
1780 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1781 made it possible to implement the integer overflow checks in a much more
1782 dynamic way, which I have now done. The artificial limitation on group
1783 length has been removed - we now have only the limit on the total length of
1784 the compiled pattern, which depends on the LINK_SIZE setting.
1785
1786 10. Fixed a bug in the documentation for get/copy named substring when
1787 duplicate names are permitted. If none of the named substrings are set, the
1788 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1789 empty string.
1790
1791 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1792 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1793 because the ] is interpreted as the first data character and the
1794 terminating ] is not found. PCRE has been made compatible with Perl in this
1795 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1796 cause memory overwriting.
1797
1798 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1799 string has been matched (to stop an infinite loop). It was not recognizing
1800 a conditional subpattern that could match an empty string if that
1801 subpattern was within another subpattern. For example, it looped when
1802 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1803 condition was not nested. This bug has been fixed.
1804
1805 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1806 past the start of the subject in the presence of bytes with the top bit
1807 set, for example "\x8aBCD".
1808
1809 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1810 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1811
1812 14. Optimized (?!) to (*FAIL).
1813
1814 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1815 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1816 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1817 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1818 does: it's just the validity check that is more restrictive.
1819
1820 16. Inserted checks for integer overflows during escape sequence (backslash)
1821 processing, and also fixed erroneous offset values for syntax errors during
1822 backslash processing.
1823
1824 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1825 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1826
1827 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1828 caused an overrun.
1829
1830 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1831 something other than just ASCII characters) inside a group that had an
1832 unlimited repeat caused a loop at compile time (while checking to see
1833 whether the group could match an empty string).
1834
1835 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1836 [\P{Any}] did so. (Error in the code for printing property names.)
1837
1838 21. An orphan \E inside a character class could cause a crash.
1839
1840 22. A repeated capturing bracket such as (A)? could cause a wild memory
1841 reference during compilation.
1842
1843 23. There are several functions in pcre_compile() that scan along a compiled
1844 expression for various reasons (e.g. to see if it's fixed length for look
1845 behind). There were bugs in these functions when a repeated \p or \P was
1846 present in the pattern. These operators have additional parameters compared
1847 with \d, etc, and these were not being taken into account when moving along
1848 the compiled data. Specifically:
1849
1850 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1851 length.
1852
1853 (b) An item such as \pL+ within a repeated group could cause crashes or
1854 loops.
1855
1856 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1857 "reference to non-existent subpattern" error.
1858
1859 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1860
1861 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1862 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1863
1864 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1865
1866 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1867 character were causing crashes (broken optimization).
1868
1869 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1870 \p or \P) caused a compile-time loop.
1871
1872 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1873 break. For example, the unstudied pattern [\r\n]A does not match the string
1874 "\r\nA" because change 7.0/46 below moves the current point on by two
1875 characters after failing to match at the start. However, the pattern \nA
1876 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1877 the same is true. There doesn't seem any very clean way out of this, but
1878 what I have chosen to do makes the common cases work: PCRE now takes note
1879 of whether there can be an explicit match for \r or \n anywhere in the
1880 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1881 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1882 pattern has explicit CR or LF references.
1883
1884 29. Added (*CR) etc for changing newline setting at start of pattern.
1885
1886
1887 Version 7.2 19-Jun-07
1888 ---------------------
1889
1890 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1891 which is apparently normally available under Windows.
1892
1893 2. Re-jig the pcregrep tests with different newline settings in an attempt
1894 to make them independent of the local environment's newline setting.
1895
1896 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1897
1898 4. Some of the "internals" tests were previously cut out when the link size
1899 was not 2, because the output contained actual offsets. The recent new
1900 "Z" feature of pcretest means that these can be cut out, making the tests
1901 usable with all link sizes.
1902
1903 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1904 stack recursion. This gives a massive performance boost under BSD, but just
1905 a small improvement under Linux. However, it saves one field in the frame
1906 in all cases.
1907
1908 6. Added more features from the forthcoming Perl 5.10:
1909
1910 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1911 recursion call. It refers to the nth most recently opened parentheses.
1912
1913 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1914 to be opened parentheses.
1915
1916 (c) Conditions that refer to capturing parentheses can be specified
1917 relatively, for example, (?(-2)... or (?(+3)...
1918
1919 (d) \K resets the start of the current match so that everything before
1920 is not part of it.
1921
1922 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1923
1924 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1925 reference syntax.
1926
1927 (g) (?| introduces a group in which the numbering of parentheses in each
1928 alternative starts with the same number.
1929
1930 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1931
1932 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1933 PCRE_INFO_JCHANGED.
1934
1935 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1936 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1937 for detecting groups that can match an empty string.
1938
1939 9. A pattern with a very large number of alternatives (more than several
1940 hundred) was running out of internal workspace during the pre-compile
1941 phase, where pcre_compile() figures out how much memory will be needed. A
1942 bit of new cunning has reduced the workspace needed for groups with
1943 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1944 workspace instead of running out of the 4096 that are available.
1945
1946 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1947
1948 11. Applied patch from Google to remove an optimization that didn't quite work.
1949 The report of the bug said:
1950
1951 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1952 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1953 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1954
1955 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1956 it matched the wrong number of bytes.
1957
1958
1959 Version 7.1 24-Apr-07
1960 ---------------------
1961
1962 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1963 that is more "standard", making use of automake and other Autotools. There
1964 is some re-arrangement of the files and adjustment of comments consequent
1965 on this.
1966
1967 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1968 for recursive directory scanning broke on some systems because the files
1969 are not scanned in any specific order and on different systems the order
1970 was different. A call to "sort" has been inserted into RunGrepTest for the
1971 approprate test as a short-term fix. In the longer term there may be an
1972 alternative.
1973
1974 3. I had an email from Eric Raymond about problems translating some of PCRE's
1975 man pages to HTML (despite the fact that I distribute HTML pages, some
1976 people do their own conversions for various reasons). The problems
1977 concerned the use of low-level troff macros .br and .in. I have therefore
1978 removed all such uses from the man pages (some were redundant, some could
1979 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1980 HTML has been updated to handle .nf/.fi and to complain if it encounters
1981 .br or .in.
1982
1983 4. Updated comments in configure.ac that get placed in config.h.in and also
1984 arranged for config.h to be included in the distribution, with the name
1985 config.h.generic, for the benefit of those who have to compile without
1986 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1987
1988 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1989 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1990 makevp.bat for the current PCRE, using the additional files
1991 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1992
1993 6. A Windows user reported a minor discrepancy with test 2, which turned out
1994 to be caused by a trailing space on an input line that had got lost in his
1995 copy. The trailing space was an accident, so I've just removed it.
1996
1997 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1998 that is needed.
1999
2000 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2001 as "const" (a) because they are and (b) because it helps the PHP
2002 maintainers who have recently made a script to detect big data structures
2003 in the php code that should be moved to the .rodata section. I remembered
2004 to update Builducptable as well, so it won't revert if ucptable.h is ever
2005 re-created.
2006
2007 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2008 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2009 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2010 support is not required. This saves 1.5-2K of code, which is important in
2011 some applications.
2012
2013 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2014 so as not to refer to the tables, even though these functions will never be
2015 called when UTF-8 support is disabled. Otherwise there are problems with a
2016 shared library.
2017
2018 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2019
2020 (a) It was defining its arguments as char * instead of void *.
2021
2022 (b) It was assuming that all moves were upwards in memory; this was true
2023 a long time ago when I wrote it, but is no longer the case.
2024
2025 The emulated memove() is provided for those environments that have neither
2026 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2027 is clearly not the case, as these two bugs were recently reported.
2028
2029 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2030 and Detrail to create the HTML documentation, the .txt form of the man
2031 pages, and it removes trailing spaces from listed files. It also creates
2032 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2033 case, it wraps all the #defines with #ifndefs. This script should be run
2034 before "make dist".
2035
2036 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2037 with Unicode property support.
2038
2039 (a) For a maximizing quantifier, if the two different cases of the
2040 character were of different lengths in their UTF-8 codings (there are
2041 some cases like this - I found 11), and the matching function had to
2042 back up over a mixture of the two cases, it incorrectly assumed they
2043 were both the same length.
2044
2045 (b) When PCRE was configured to use the heap rather than the stack for
2046 recursion during matching, it was not correctly preserving the data for
2047 the other case of a UTF-8 character when checking ahead for a match
2048 while processing a minimizing repeat. If the check also involved
2049 matching a wide character, but failed, corruption could cause an
2050 erroneous result when trying to check for a repeat of the original
2051 character.
2052
2053 13. Some tidying changes to the testing mechanism:
2054
2055 (a) The RunTest script now detects the internal link size and whether there
2056 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2057 values substituted by "configure". (The RunGrepTest script already did
2058 this for UTF-8.) The configure.ac script no longer substitutes the
2059 relevant variables.
2060
2061 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2062 with length and offset values. This means that the output is different
2063 for different internal link sizes. Test 2 is skipped for link sizes
2064 other than 2 because of this, bypassing the problem. Unfortunately,
2065 there was also a test in test 3 (the locale tests) that used /B and
2066 failed for link sizes other than 2. Rather than cut the whole test out,
2067 I have added a new /Z option to pcretest that replaces the length and
2068 offset values with spaces. This is now used to make test 3 independent
2069 of link size. (Test 2 will be tidied up later.)
2070
2071 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2072 segmentation fault instead of returning the appropriate error message.
2073
2074 15. In multiline mode when the newline sequence was set to "any", the pattern
2075 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2076 This doesn't seem right; it now treats the CRLF combination as the line
2077 ending, and so does not match in that case. It's only a pattern such as ^$
2078 that would hit this one: something like ^ABC$ would have failed after \r
2079 and then tried again after \r\n.
2080
2081 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2082 in an attempt to make files that differ only in their line terminators
2083 compare equal. This works on Linux.
2084
2085 17. Under certain error circumstances pcregrep might try to free random memory
2086 as it exited. This is now fixed, thanks to valgrind.
2087
2088 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2089 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2090 was because its rules for how to advance for /g after matching an empty
2091 string at the end of a line did not allow for this case. They now check for
2092 it specially.
2093
2094 20. pcretest is supposed to handle patterns and data of any length, by
2095 extending its buffers when necessary. It was getting this wrong when the
2096 buffer for a data line had to be extended.
2097
2098 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2099 CRLF as a newline sequence.
2100
2101 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2102 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2103 I have nevertheless tidied it up.
2104
2105 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2106
2107 24. Added a man page for pcre-config.
2108
2109
2110 Version 7.0 19-Dec-06
2111 ---------------------
2112
2113 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2114 moving to gcc 4.1.1.
2115
2116 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2117 sys/time.h, which is documented as needed for this function. It doesn't
2118 seem to matter on Linux, but it showed up on some releases of OS X.
2119
2120 3. It seems that there are systems where bytes whose values are greater than
2121 127 match isprint() in the "C" locale. The "C" locale should be the
2122 default when a C program starts up. In most systems, only ASCII printing
2123 characters match isprint(). This difference caused the output from pcretest
2124 to vary, making some of the tests fail. I have changed pcretest so that:
2125
2126 (a) When it is outputting text in the compiled version of a pattern, bytes
2127 other than 32-126 are always shown as hex escapes.
2128
2129 (b) When it is outputting text that is a matched part of a subject string,
2130 it does the same, unless a different locale has been set for the match
2131 (using the /L modifier). In this case, it uses isprint() to decide.
2132
2133 4. Fixed a major bug that caused incorrect computation of the amount of memory
2134 required for a compiled pattern when options that changed within the
2135 pattern affected the logic of the preliminary scan that determines the
2136 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2137 that the computed length was too small. The symptoms of this bug were
2138 either the PCRE error "internal error: code overflow" from pcre_compile(),
2139 or a glibc crash with a message such as "pcretest: free(): invalid next
2140 size (fast)". Examples of patterns that provoked this bug (shown in
2141 pcretest format) are:
2142
2143 /(?-x: )/x
2144 /(?x)(?-x: \s*#\s*)/
2145 /((?i)[\x{c0}])/8
2146 /(?i:[\x{c0}])/8
2147
2148 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2149 is now done differently.
2150
2151 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2152 wrapper classes; (b) implement a new function in the C++ scanner that is
2153 more efficient than the old way of doing things because it avoids levels of
2154 recursion in the regex matching; (c) add a paragraph to the documentation
2155 for the FullMatch() function.
2156
2157 6. The escape sequence \n was being treated as whatever was defined as
2158 "newline". Not only was this contrary to the documentation, which states
2159 that \n is character 10 (hex 0A), but it also went horribly wrong when
2160 "newline" was defined as CRLF. This has been fixed.
2161
2162 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2163 was being set to -1 for the "end of line" case (supposedly a value that no
2164 character can have). Though this value is never used (the check for end of
2165 line is "zero bytes in current character"), it caused compiler complaints.
2166 I've changed it to 0xffffffff.
2167
2168 8. In pcre_version.c, the version string was being built by a sequence of
2169 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2170 string (as it is for production releases) called a macro with an empty
2171 argument. The C standard says the result of this is undefined. The gcc
2172 compiler treats it as an empty string (which was what was wanted) but it is
2173 reported that Visual C gives an error. The source has been hacked around to
2174 avoid this problem.
2175
2176 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2177 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2178 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2179 of them did).
2180
2181 10. Originally, pcretest opened its input and output without "b"; then I was
2182 told that "b" was needed in some environments, so it was added for release
2183 5.0 to both the input and output. (It makes no difference on Unix-like
2184 systems.) Later I was told that it is wrong for the input on Windows. I've
2185 now abstracted the modes into two macros, to make it easier to fiddle with
2186 them, and removed "b" from the input mode under Windows.
2187
2188 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2189
2190 12. Added -help and --help to pcretest as an official way of being reminded
2191 of the options.
2192
2193 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2194 and pcrecpp.cc because they annoy compilers at high warning levels.
2195
2196 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2197
2198 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2199 scripts are not C programs :-) and which was not noticed because it works
2200 on Linux.
2201
2202 16. pcretest is supposed to handle any length of pattern and data line (as one
2203 line or as a continued sequence of lines) by extending its input buffer if
2204 necessary. This feature was broken for very long pattern lines, leading to
2205 a string of junk being passed to pcre_compile() if the pattern was longer
2206 than about 50K.
2207
2208 17. I have done a major re-factoring of the way pcre_compile() computes the
2209 amount of memory needed for a compiled pattern. Previously, there was code
2210 that made a preliminary scan of the pattern in order to do this. That was
2211 OK when PCRE was new, but as the facilities have expanded, it has become
2212 harder and harder to keep it in step with the real compile phase, and there
2213 have been a number of bugs (see for example, 4 above). I have now found a
2214 cunning way of running the real compile function in a "fake" mode that
2215 enables it to compute how much memory it would need, while actually only
2216 ever using a few hundred bytes of working memory and without too many
2217 tests of the mode. This should make future maintenance and development
2218 easier. A side effect of this work is that the limit of 200 on the nesting
2219 depth of parentheses has been removed (though this was never a serious
2220 limitation, I suspect). However, there is a downside: pcre_compile() now
2221 runs more slowly than before (30% or more, depending on the pattern). I
2222 hope this isn't a big issue. There is no effect on runtime performance.
2223
2224 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2225 newline (only possible for the last line of a file) and it was a
2226 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2227
2228 19. Added additional timing features to pcretest. (1) The -tm option now times
2229 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2230 separate command line item, by a number that specifies the number of
2231 repeats to use when timing. The default is 50000; this gives better
2232 precision, but takes uncomfortably long for very large patterns.
2233
2234 20. Extended pcre_study() to be more clever in cases where a branch of a
2235 subpattern has no definite first character. For example, (a*|b*)[cd] would
2236 previously give no result from pcre_study(). Now it recognizes that the
2237 first character must be a, b, c, or d.
2238
2239 21. There was an incorrect error "recursive call could loop indefinitely" if
2240 a subpattern (or the entire pattern) that was being tested for matching an
2241 empty string contained only one non-empty item after a nested subpattern.
2242 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2243 incorrectly, because the \d was being skipped in the check.
2244
2245 22. The pcretest program now has a new pattern option /B and a command line
2246 option -b, which is equivalent to adding /B to every pattern. This causes
2247 it to show the compiled bytecode, without the additional information that
2248 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2249 is the same as /B/I).
2250
2251 23. A new optimization is now able automatically to treat some sequences such
2252 as a*b as a*+b. More specifically, if something simple (such as a character
2253 or a simple class like \d) has an unlimited quantifier, and is followed by
2254 something that cannot possibly match the quantified thing, the quantifier
2255 is automatically "possessified".
2256
2257 24. A recursive reference to a subpattern whose number was greater than 39
2258 went wrong under certain circumstances in UTF-8 mode. This bug could also
2259 have affected the operation of pcre_study().
2260
2261 25. Realized that a little bit of performance could be had by replacing
2262 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2263
2264 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2265
2266 27. Possessive quantifiers such as a++ were previously implemented by turning
2267 them into atomic groups such as ($>a+). Now they have their own opcodes,
2268 which improves performance. This includes the automatically created ones
2269 from 23 above.
2270
2271 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2272 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2273 the first matched character to be a colon. This applied both to named and
2274 numbered groups.
2275
2276 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2277
2278 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2279 building PCRE on Windows easier, so I have included it in the distribution.
2280
2281 31. There is now a check in pcretest against a ridiculously large number being
2282 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2283 loop, the loop is abandoned.
2284
2285 32. Forward references to subpatterns in conditions such as (?(2)...) where
2286 subpattern 2 is defined later cause pcre_compile() to search forwards in
2287 the pattern for the relevant set of parentheses. This search went wrong
2288 when there were unescaped parentheses in a character class, parentheses
2289 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2290
2291 33. "Subroutine" calls and backreferences were previously restricted to
2292 referencing subpatterns earlier in the regex. This restriction has now
2293 been removed.
2294
2295 34. Added a number of extra features that are going to be in Perl 5.10. On the
2296 whole, these are just syntactic alternatives for features that PCRE had
2297 previously implemented using the Python syntax or my own invention. The
2298 other formats are all retained for compatibility.
2299
2300 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2301 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2302 also .NET compatible.
2303
2304 (b) A recursion or subroutine call to a named group can now be defined as
2305 (?&name) as well as (?P>name).
2306
2307 (c) A backreference to a named group can now be defined as \k<name> or
2308 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2309 5.10, are also .NET compatible.
2310
2311 (d) A conditional reference to a named group can now use the syntax
2312 (?(<name>) or (?('name') as well as (?(name).
2313
2314 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2315 groups (named and numbered) that are never evaluated inline, but can be
2316 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2317 is always false. There may be only one alternative in such a group.
2318
2319 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2320 as the simple (?(R). The condition is true only if the most recent
2321 recursion is that of the given number or name. It does not search out
2322 through the entire recursion stack.
2323
2324 (g) The escape \gN or \g{N} has been added, where N is a positive or
2325 negative number, specifying an absolute or relative reference.
2326
2327 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2328 some "unreachable code" warnings.
2329
2330 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2331 things, this adds five new scripts.
2332
2333 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2334 There were also incompatibilities regarding the handling of \Q..\E inside
2335 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2336 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2337
2338 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2339 matches an empty string, and forcibly breaks the loop. There were bugs in
2340 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2341 against aaaa the result was just "a" rather than "aaaa", for example. Two
2342 separate and independent bugs (that affected different cases) have been
2343 fixed.
2344
2345 39. Refactored the code to abolish the use of different opcodes for small
2346 capturing bracket numbers. This is a tidy that I avoided doing when I
2347 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2348 The new approach is not only tidier, it makes it possible to reduce the
2349 memory needed to fix the previous bug (38).
2350
2351 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2352 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2353 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2354 mode.
2355
2356 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2357 report.
2358
2359 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2360 copy construction and assignment in the C++ wrapper.
2361
2362 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2363 couple of bugs that could have given wrong results in the "--newline=crlf"
2364 case.
2365
2366 44. Added a number of casts and did some reorganization of signed/unsigned int
2367 variables following suggestions from Dair Grant. Also renamed the variable
2368 "this" as "item" because it is a C++ keyword.
2369
2370 45. Arranged for dftables to add
2371
2372 #include "pcre_internal.h"
2373
2374 to pcre_chartables.c because without it, gcc 4.x may remove the array
2375 definition from the final binary if PCRE is built into a static library and
2376 dead code stripping is activated.
2377
2378 46. For an unanchored pattern, if a match attempt fails at the start of a
2379 newline sequence, and the newline setting is CRLF or ANY, and the next two
2380 characters are CRLF, advance by two characters instead of one.
2381
2382
2383 Version 6.7 04-Jul-06
2384 ---------------------
2385
2386 1. In order to handle tests when input lines are enormously long, pcretest has
2387 been re-factored so that it automatically extends its buffers when
2388 necessary. The code is crude, but this _is_ just a test program. The
2389 default size has been increased from 32K to 50K.
2390
2391 2. The code in pcre_study() was using the value of the re argument before
2392 testing it for NULL. (Of course, in any sensible call of the function, it
2393 won't be NULL.)
2394
2395 3. The memmove() emulation function in pcre_internal.h, which is used on
2396 systems that lack both memmove() and bcopy() - that is, hardly ever -
2397 was missing a "static" storage class specifier.
2398
2399 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2400 containing an extended class (one that cannot be represented by a bitmap
2401 because it contains high-valued characters or Unicode property items, e.g.
2402 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2403 pattern, but PCRE should not loop if you do not (it no longer does).
2404 [Detail: two cases were found: (a) a repeated subpattern containing an
2405 extended class; (b) a recursive reference to a subpattern that followed a
2406 previous extended class. It wasn't skipping over the extended class
2407 correctly when UTF-8 mode was not set.]
2408
2409 5. A negated single-character class was not being recognized as fixed-length
2410 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2411 compile error "lookbehind assertion is not fixed length".
2412
2413 6. The RunPerlTest auxiliary script was showing an unexpected difference
2414 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2415 write a Perl script that can interpret lines of an input file either as
2416 byte characters or as UTF-8, which is what "perltest" was being required to
2417 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2418 can't do is switch easily at run time between having the "use utf8;" pragma
2419 or not. In the end, I fudged it by using the RunPerlTest script to insert
2420 "use utf8;" explicitly for the UTF-8 tests.
2421
2422 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2423 the end of the subject string, contrary to the documentation and to what
2424 Perl does. This was true of both matching functions. Now it matches only at
2425 the start of the subject and immediately after *internal* newlines.
2426
2427 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2428 a pointer to an int instead of a pointer to an unsigned long int. This
2429 caused problems on 64-bit systems.
2430
2431 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2432 instance of the 'standard' template library not being so standard".
2433
2434 10. There was no check on the number of named subpatterns nor the maximum
2435 length of a subpattern name. The product of these values is used to compute
2436 the size of the memory block for a compiled pattern. By supplying a very
2437 long subpattern name and a large number of named subpatterns, the size
2438 computation could be caused to overflow. This is now prevented by limiting
2439 the length of names to 32 characters, and the number of named subpatterns
2440 to 10,000.
2441
2442 11. Subpatterns that are repeated with specific counts have to be replicated in
2443 the compiled pattern. The size of memory for this was computed from the
2444 length of the subpattern and the repeat count. The latter is limited to
2445 65535, but there was no limit on the former, meaning that integer overflow
2446 could in principle occur. The compiled length of a repeated subpattern is
2447 now limited to 30,000 bytes in order to prevent this.
2448
2449 12. Added the optional facility to have named substrings with the same name.
2450
2451 13. Added the ability to use a named substring as a condition, using the
2452 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2453 are numbers (not recommended). Forward references are permitted.
2454
2455 14. Added forward references in named backreferences (if you see what I mean).
2456
2457 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2458 pattern could run off the end of the subject. For example, the pattern
2459 "(?s)(.{1,5})"8 did this with the subject "ab".
2460
2461 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2462 PCRE_CASELESS was set when matching characters that were quantified with ?
2463 or *.
2464
2465 17. A character class other than a single negated character that had a minimum
2466 but no maximum quantifier - for example [ab]{6,} - was not handled
2467 correctly by pce_dfa_exec(). It would match only one character.
2468
2469 18. A valid (though odd) pattern that looked like a POSIX character
2470 class but used an invalid character after [ (for example [[,abc,]]) caused
2471 pcre_compile() to give the error "Failed: internal error: code overflow" or
2472 in some cases to crash with a glibc free() error. This could even happen if
2473 the pattern terminated after [[ but there just happened to be a sequence of
2474 letters, a binary zero, and a closing ] in the memory that followed.
2475
2476 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2477 over the years. Originally (before any Unicode support), just the bottom 8
2478 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2479 output from "man perlunicode" includes this:
2480
2481 The regular expression compiler produces polymorphic opcodes. That
2482 is, the pattern adapts to the data and automatically switches to
2483 the Unicode character scheme when presented with Unicode data--or
2484 instead uses a traditional byte scheme when presented with byte
2485 data.
2486
2487 Sadly, a wide octal escape does not cause a switch, and in a string with
2488 no other multibyte characters, these octal escapes are treated as before.
2489 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2490 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2491 Unicode string.
2492
2493 I have not perpetrated such confusion in PCRE. Up till now, it took just
2494 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2495 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2496 translate to the appropriate multibyte character.
2497
2498 29. Applied some refactoring to reduce the number of warnings from Microsoft
2499 and Borland compilers. This has included removing the fudge introduced
2500 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2501 a warning about an unused variable.
2502
2503 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2504 characters since release 4.0, because Perl (from release 5.004) does not.
2505 [Or at least, is documented not to: some releases seem to be in conflict
2506 with the documentation.] However, when a pattern was studied with
2507 pcre_study() and all its branches started with \s, PCRE still included VT
2508 as a possible starting character. Of course, this did no harm; it just
2509 caused an unnecessary match attempt.
2510
2511 22. Removed a now-redundant internal flag bit that recorded the fact that case
2512 dependency changed within the pattern. This was once needed for "required
2513 byte" processing, but is no longer used. This recovers a now-scarce options
2514 bit. Also moved the least significant internal flag bit to the most-
2515 significant bit of the word, which was not previously used (hangover from
2516 the days when it was an int rather than a uint) to free up another bit for
2517 the future.
2518
2519 23. Added support for CRLF line endings as well as CR and LF. As well as the
2520 default being selectable at build time, it can now be changed at runtime
2521 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2522 specify that it is scanning data with non-default line endings.
2523
2524 24. Changed the definition of CXXLINK to make it agree with the definition of
2525 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2526
2527 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2528 recursions. This makes a big different to stack usage for some patterns.
2529
2530 26. If a subpattern containing a named recursion or subroutine reference such
2531 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2532 the space required for the compiled pattern went wrong and gave too small a
2533 value. Depending on the environment, this could lead to "Failed: internal
2534 error: code overflow at offset 49" or "glibc detected double free or
2535 corruption" errors.
2536
2537 27. Applied patches from Google (a) to support the new newline modes and (b) to
2538 advance over multibyte UTF-8 characters in GlobalReplace.
2539
2540 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2541 difference for some implementation of PCRE in some Windows version.
2542
2543 29. Added some extra testing facilities to pcretest:
2544
2545 \q<number> in a data line sets the "match limit" value
2546 \Q<number> in a data line sets the "match recursion limt" value
2547 -S <number> sets the stack size, where <number> is in megabytes
2548
2549 The -S option isn't available for Windows.
2550
2551
2552 Version 6.6 06-Feb-06
2553 ---------------------
2554
2555 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2556 in pcreposix.h. I have copied the definition from pcre.h.
2557
2558 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2559 because pcre.h is no longer a built file.
2560
2561 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2562 not normally included in the compiled code.
2563
2564
2565 Version 6.5 01-Feb-06
2566 ---------------------
2567
2568 1. When using the partial match feature with pcre_dfa_exec(), it was not
2569 anchoring the second and subsequent partial matches at the new starting
2570 point. This could lead to incorrect results. For example, with the pattern
2571 /1234/, partially matching against "123" and then "a4" gave a match.
2572
2573 2. Changes to pcregrep:
2574
2575 (a) All non-match returns from pcre_exec() were being treated as failures
2576 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2577 error message is output. Some extra information is given for the
2578 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2579 probably the only errors that are likely to be caused by users (by
2580 specifying a regex that has nested indefinite repeats, for instance).
2581 If there are more than 20 of these errors, pcregrep is abandoned.
2582
2583 (b) A binary zero was treated as data while matching, but terminated the
2584 output line if it was written out. This has been fixed: binary zeroes
2585 are now no different to any other data bytes.
2586
2587 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2588 used to set a locale for matching. The --locale=xxxx long option has
2589 been added (no short equivalent) to specify a locale explicitly on the
2590 pcregrep command, overriding the environment variables.
2591
2592 (d) When -B was used with -n, some line numbers in the output were one less
2593 than they should have been.
2594
2595 (e) Added the -o (--only-matching) option.
2596
2597 (f) If -A or -C was used with -c (count only), some lines of context were
2598 accidentally printed for the final match.
2599
2600 (g) Added the -H (--with-filename) option.
2601
2602 (h) The combination of options -rh failed to suppress file names for files
2603 that were found from directory arguments.
2604
2605 (i) Added the -D (--devices) and -d (--directories) options.
2606
2607 (j) Added the -F (--fixed-strings) option.
2608
2609 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2610
2611 (l) Added the --colo(u)r option.
2612
2613 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2614 is not present by default.
2615
2616 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2617 items such as (?R) or (?1), when the recursion could match a number of
2618 alternatives. If it matched one of the alternatives, but subsequently,
2619 outside the recursion, there was a failure, the code tried to back up into
2620 the recursion. However, because of the way PCRE is implemented, this is not
2621 possible, and the result was an incorrect result from the match.
2622
2623 In order to prevent this happening, the specification of recursion has
2624 been changed so that all such subpatterns are automatically treated as
2625 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2626
2627 4. I had overlooked the fact that, in some locales, there are characters for
2628 which isalpha() is true but neither isupper() nor islower() are true. In
2629 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2630 and ordfeminine) are like this. This affected the treatment of \w and \W
2631 when they appeared in character classes, but not when they appeared outside
2632 a character class. The bit map for "word" characters is now created
2633 separately from the results of isalnum() instead of just taking it from the
2634 upper, lower, and digit maps. (Plus the underscore character, of course.)
2635
2636 5. The above bug also affected the handling of POSIX character classes such as
2637 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2638 permanent tables. Instead, the bit maps for such a class were previously
2639 created as the appropriate unions of the upper, lower, and digit bitmaps.
2640 Now they are created by subtraction from the [[:word:]] class, which has
2641 its own bitmap.
2642
2643 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2644 It is created by subtracting the vertical space characters (\x09, \x0a,
2645 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2646 subtraction was done in the overall bitmap for a character class, meaning
2647 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2648 be recognized. This bug has been fixed.
2649
2650 7. Patches from the folks at Google:
2651
2652 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2653 real life, but is still worth protecting against".
2654
2655 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2656 regular expressions".
2657
2658 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2659 have it.
2660
2661 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2662 "configure" and the latter not, in order to fix a problem somebody had
2663 with compiling the Arg class on HP-UX.
2664
2665 (e) Improve the error-handling of the C++ wrapper a little bit.
2666
2667 (f) New tests for checking recursion limiting.
2668
2669 8. The pcre_memmove() function, which is used only if the environment does not
2670 have a standard memmove() function (and is therefore rarely compiled),
2671 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2672 returning a result (though PCRE never actually uses the result).
2673
2674 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2675 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2676 returned instead of calling malloc() with an overflowing number that would
2677 most likely cause subsequent chaos.
2678
2679 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2680
2681 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2682 with this option is matched, the nmatch and pmatch options of regexec() are
2683 ignored.
2684
2685 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2686 provided in case anyone wants to the the POSIX interface with UTF-8
2687 strings.
2688
2689 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2690 C++ linking (needed for some HP-UX environments).
2691
2692 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2693 (unused parameter) and in the pcre_printint() function (omitted "default"
2694 switch label when the default is to do nothing).
2695
2696 15. Added some code to make it possible, when PCRE is compiled as a C++
2697 library, to replace subject pointers for pcre_exec() with a smart pointer
2698 class, thus making it possible to process discontinuous strings.
2699
2700 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2701 much the same function. They were added by different people who were trying
2702 to make PCRE easy to compile on non-Unix systems. It has been suggested
2703 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2704 for compiling on Windows systems. I have therefore replaced it with
2705 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2706 defaults to "extern" for C or "extern C" for C++, which works fine on
2707 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2708 SCOPE with something explicit in config.h. In addition:
2709
2710 (a) pcreposix.h still had just "extern" instead of either of these macros;
2711 I have replaced it with PCRE_DATA_SCOPE.
2712
2713 (b) Functions such as _pcre_xclass(), which are internal to the library,
2714 but external in the C sense, all had PCRE_EXPORT in their definitions.
2715 This is apparently wrong for the Windows case, so I have removed it.
2716 (It makes no difference on Unix-like systems.)
2717
2718 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2719 of recursive calls to match(). This is different to MATCH_LIMIT because
2720 that limits the total number of calls to match(), not all of which increase
2721 the depth of recursion. Limiting the recursion depth limits the amount of
2722 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2723 when PCRE is compiled, and changed at run time. A patch from Google adds
2724 this functionality to the C++ interface.
2725
2726 18. Changes to the handling of Unicode character properties:
2727
2728 (a) Updated the table to Unicode 4.1.0.
2729
2730 (b) Recognize characters that are not in the table as "Cn" (undefined).
2731
2732 (c) I revised the way the table is implemented to a much improved format
2733 which includes recognition of ranges. It now supports the ranges that
2734 are defined in UnicodeData.txt, and it also amalgamates other
2735 characters into ranges. This has reduced the number of entries in the
2736 table from around 16,000 to around 3,000, thus reducing its size
2737 considerably. I realized I did not need to use a tree structure after
2738 all - a binary chop search is just as efficient. Having reduced the
2739 number of entries, I extended their size from 6 bytes to 8 bytes to
2740 allow for more data.
2741
2742 (d) Added support for Unicode script names via properties such as \p{Han}.
2743
2744 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2745 matching that character.
2746
2747 20. When matching a repeated Unicode property with a minimum greater than zero,
2748 (for example \pL{2,}), PCRE could look past the end of the subject if it
2749 reached it while seeking the minimum number of characters. This could
2750 happen only if some of the characters were more than one byte long, because
2751 there is a check for at least the minimum number of bytes.
2752
2753 21. Refactored the implementation of \p and \P so as to be more general, to
2754 allow for more different types of property in future. This has changed the
2755 compiled form incompatibly. Anybody with saved compiled patterns that use
2756 \p or \P will have to recompile them.
2757
2758 22. Added "Any" and "L&" to the supported property types.
2759
2760 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2761 but give a compile time error if the value is greater than 0xff.
2762
2763 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2764 accidentally not being installed or uninstalled.
2765
2766 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2767 made were to insert the current release number. This seemed silly, because
2768 it made things harder for people building PCRE on systems that don't run
2769 "configure". I have turned pcre.h into a distributed file, no longer built
2770 by "configure", with the version identification directly included. There is
2771 no longer a pcre.h.in file.
2772
2773 However, this change necessitated a change to the pcre-config script as
2774 well. It is built from pcre-config.in, and one of the substitutions was the
2775 release number. I have updated configure.ac so that ./configure now finds
2776 the release number by grepping pcre.h.
2777
2778 26. Added the ability to run the tests under valgrind.
2779
2780
2781 Version 6.4 05-Sep-05
2782 ---------------------
2783
2784 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2785 "--" to be printed when multiple files were scanned, even when none of the
2786 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2787 consider it to be a bug, and have restored the previous behaviour.
2788
2789 2. A couple of code tidies to get rid of compiler warnings.
2790
2791 3. The pcretest program used to cheat by referring to symbols in the library
2792 whose names begin with _pcre_. These are internal symbols that are not
2793 really supposed to be visible externally, and in some environments it is
2794 possible to suppress them. The cheating is now confined to including
2795 certain files from the library's source, which is a bit cleaner.
2796
2797 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2798 file's purpose clearer.
2799
2800 5. Reorganized pcre_ucp_findchar().
2801
2802
2803 Version 6.3 15-Aug-05
2804 ---------------------
2805
2806 1. The file libpcre.pc.in did not have general read permission in the tarball.
2807
2808 2. There were some problems when building without C++ support:
2809
2810 (a) If C++ support was not built, "make install" and "make test" still
2811 tried to test it.
2812
2813 (b) There were problems when the value of CXX was explicitly set. Some
2814 changes have been made to try to fix these, and ...
2815
2816 (c) --disable-cpp can now be used to explicitly disable C++ support.
2817
2818 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2819 backslash in a target when C++ was disabled. This confuses some
2820 versions of "make", apparently. Using an intermediate variable solves
2821 this. (Same for CPP_LOBJ.)
2822
2823 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2824 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2825 necessary on certain architectures.
2826
2827 4. Added a setting of -export-symbols-regex to the link command to remove
2828 those symbols that are exported in the C sense, but actually are local
2829 within the library, and not documented. Their names all begin with
2830 "_pcre_". This is not a perfect job, because (a) we have to except some
2831 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2832 available (and never for static libraries). I have made a note to try to
2833 find a way round (a) in the future.
2834
2835
2836 Version 6.2 01-Aug-05
2837 ---------------------
2838
2839 1. There was no test for integer overflow of quantifier values. A construction
2840 such as {1111111111111111} would give undefined results. What is worse, if
2841 a minimum quantifier for a parenthesized subpattern overflowed and became
2842 negative, the calculation of the memory size went wrong. This could have
2843 led to memory overwriting.
2844
2845 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2846
2847 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2848 operating environments where this matters.
2849
2850 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2851 PCRE options from within the C++ wrapper.
2852
2853 5. Named capturing subpatterns were not being correctly counted when a pattern
2854 was compiled. This caused two problems: (a) If there were more than 100
2855 such subpatterns, the calculation of the memory needed for the whole
2856 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2857 back references of the form \12, where the number was greater than 9, were
2858 not recognized as back references, even though there were sufficient
2859 previous subpatterns.
2860
2861 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2862 versions of gcc, e.g. 2.95.4.
2863
2864
2865 Version 6.1 21-Jun-05
2866 ---------------------
2867
2868 1. There was one reference to the variable "posix" in pcretest.c that was not
2869 surrounded by "#if !defined NOPOSIX".
2870
2871 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2872 the cross-check on the old pcre_info() function, for the benefit of the
2873 cut-down version of PCRE that is currently imported into Exim.
2874
2875 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2876 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2877 patterns that start (?i) but I don't think that matters. The use of (?i) is
2878 just an example; this all applies to the other options as well.
2879
2880 4. Since libtool seems to echo the compile commands it is issuing, the output
2881 from "make" can be reduced a bit by putting "@" in front of each libtool
2882 compile command.
2883
2884 5. Patch from the folks at Google for configure.in to be a bit more thorough
2885 in checking for a suitable C++ installation before trying to compile the
2886 C++ stuff. This should fix a reported problem when a compiler was present,
2887 but no suitable headers.
2888
2889 6. The man pages all had just "PCRE" as their title. I have changed them to
2890 be the relevant file name. I have also arranged that these names are
2891 retained in the file doc/pcre.txt, which is a concatenation in text format
2892 of all the man pages except the little individual ones for each function.
2893
2894 7. The NON-UNIX-USE file had not been updated for the different set of source
2895 files that come with release 6. I also added a few comments about the C++
2896 wrapper.
2897
2898
2899 Version 6.0 07-Jun-05
2900 ---------------------
2901
2902 1. Some minor internal re-organization to help with my DFA experiments.
2903
2904 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2905 didn't matter for the library itself when fully configured, but did matter
2906 when compiling without UCP support, or within Exim, where the ucp files are
2907 not imported.
2908
2909 3. Refactoring of the library code to split up the various functions into
2910 different source modules. The addition of the new DFA matching code (see
2911 below) to a single monolithic source would have made it really too
2912 unwieldy, quite apart from causing all the code to be include in a
2913 statically linked application, when only some functions are used. This is
2914 relevant even without the DFA addition now that patterns can be compiled in
2915 one application and matched in another.
2916
2917 The downside of splitting up is that there have to be some external
2918 functions and data tables that are used internally in different modules of
2919 the library but which are not part of the API. These have all had their
2920 names changed to start with "_pcre_" so that they are unlikely to clash
2921 with other external names.
2922
2923 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2924 a different (DFA) algorithm. Although it is slower than the original
2925 function, it does have some advantages for certain types of matching
2926 problem.
2927
2928 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2929 including restarting after a partial match.
2930
2931 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2932 defined when compiling for Windows was sent to me. I have put it into the
2933 code, though I have no means of testing or verifying it.
2934
2935 7. Added the pcre_refcount() auxiliary function.
2936
2937 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2938 match before or at the first newline in the subject string. In pcretest,
2939 the /f option on a pattern can be used to set this.
2940
2941 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2942 would behave wrongly. This has been present in PCRE since release 4.0.
2943
2944 10. A number of changes to the pcregrep command:
2945
2946 (a) Refactored how -x works; insert ^(...)$ instead of setting
2947 PCRE_ANCHORED and checking the length, in preparation for adding
2948 something similar for -w.
2949
2950 (b) Added the -w (match as a word) option.
2951
2952 (c) Refactored the way lines are read and buffered so as to have more
2953 than one at a time available.
2954
2955 (d) Implemented a pcregrep test script.
2956
2957 (e) Added the -M (multiline match) option. This allows patterns to match
2958 over several lines of the subject. The buffering ensures that at least
2959 8K, or the rest of the document (whichever is the shorter) is available
2960 for matching (and similarly the previous 8K for lookbehind assertions).
2961
2962 (f) Changed the --help output so that it now says
2963
2964 -w, --word-regex(p)
2965
2966 instead of two lines, one with "regex" and the other with "regexp"
2967 because that confused at least one person since the short forms are the
2968 same. (This required a bit of code, as the output is generated
2969 automatically from a table. It wasn't just a text change.)
2970
2971 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2972 option but starts with a hyphen. Could be a pattern or a path name
2973 starting with a hyphen, for instance.
2974
2975 (h) "-" can be given as a file name to represent stdin.
2976
2977 (i) When file names are being printed, "(standard input)" is used for
2978 the standard input, for compatibility with GNU grep. Previously
2979 "<stdin>" was used.
2980
2981 (j) The option --label=xxx can be used to supply a name to be used for
2982 stdin when file names are being printed. There is no short form.
2983
2984 (k) Re-factored the options decoding logic because we are going to add
2985 two more options that take data. Such options can now be given in four
2986 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2987
2988 (l) Added the -A, -B, and -C options for requesting that lines of context
2989 around matches be printed.
2990
2991 (m) Added the -L option to print the names of files that do not contain
2992 any matching lines, that is, the complement of -l.
2993
2994 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2995 continue to scan other files.
2996
2997 (o) The -s option was incorrectly implemented. For compatibility with other
2998 greps, it now suppresses the error message for a non-existent or non-
2999 accessible file (but not the return code). There is a new option called
3000 -q that suppresses the output of matching lines, which was what -s was
3001 previously doing.
3002
3003 (p) Added --include and --exclude options to specify files for inclusion
3004 and exclusion when recursing.
3005
3006 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3007 Hopefully, it now does.
3008
3009 12. Missing cast in pcre_study().
3010
3011 13. Added an "uninstall" target to the makefile.
3012
3013 14. Replaced "extern" in the function prototypes in Makefile.in with
3014 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3015 world, but is set differently for Windows.
3016
3017 15. Added a second compiling function called pcre_compile2(). The only
3018 difference is that it has an extra argument, which is a pointer to an
3019 integer error code. When there is a compile-time failure, this is set
3020 non-zero, in addition to the error test pointer being set to point to an
3021 error message. The new argument may be NULL if no error number is required
3022 (but then you may as well call pcre_compile(), which is now just a
3023 wrapper). This facility is provided because some applications need a
3024 numeric error indication, but it has also enabled me to tidy up the way
3025 compile-time errors are handled in the POSIX wrapper.
3026
3027 16. Added VPATH=.libs to the makefile; this should help when building with one
3028 prefix path and installing with another. (Or so I'm told by someone who
3029 knows more about this stuff than I do.)
3030
3031 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3032 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3033 match everything, including newlines. This is not POSIX-compatible, but
3034 somebody wanted the feature. From pcretest it can be activated by using
3035 both the P and the s flags.
3036
3037 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3038
3039 19. libpcre.pc was being incorrectly installed as executable.
3040
3041 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3042 it now also looks for '\r' so that it will work unmodified on Windows.
3043
3044 21. Added Google's contributed C++ wrapper to the distribution.
3045
3046 22. Added some untidy missing memory free() calls in pcretest, to keep
3047 Electric Fence happy when testing.
3048
3049
3050
3051 Version 5.0 13-Sep-04
3052 ---------------------
3053
3054 1. Internal change: literal characters are no longer packed up into items
3055 containing multiple characters in a single byte-string. Each character
3056 is now matched using a separate opcode. However, there may be more than one
3057 byte in the character in UTF-8 mode.
3058
3059 2. The pcre_callout_block structure has two new fields: pattern_position and
3060 next_item_length. These contain the offset in the pattern to the next match
3061 item, and its length, respectively.
3062
3063 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3064 insertion of callouts before each pattern item. Added the /C option to
3065 pcretest to make use of this.
3066
3067 4. On the advice of a Windows user, the lines
3068
3069 #if defined(_WIN32) || defined(WIN32)
3070 _setmode( _fileno( stdout ), 0x8000 );
3071 #endif /* defined(_WIN32) || defined(WIN32) */
3072
3073 have been added to the source of pcretest. This apparently does useful
3074 magic in relation to line terminators.
3075
3076 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3077 for the benefit of those environments where the "b" makes a difference.
3078
3079 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3080 to know about it. I have put a hack into configure.in that adds in code
3081 to set GCC=yes if CC=icc. This seems to end up at a point in the
3082 generated configure script that is early enough to affect the setting of
3083 compiler options, which is what is needed, but I have no means of testing
3084 whether it really works. (The user who reported this had patched the
3085 generated configure script, which of course I cannot do.)
3086
3087 LATER: After change 22 below (new libtool files), the configure script
3088 seems to know about icc (and also ecc). Therefore, I have commented out
3089 this hack in configure.in.
3090
3091 7. Added support for pkg-config (2 patches were sent in).
3092
3093 8. Negated POSIX character classes that used a combination of internal tables
3094 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3095 [[:^ascii]]. Typically, they would match almost any characters. The other
3096 POSIX classes were not broken in this way.
3097
3098 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3099 to find the match, as PCRE was deluded into thinking that the match had to
3100 start at the start point or following a newline. The same bug applied to
3101 patterns with negative forward assertions or any backward assertions
3102 preceding ".*" at the start, unless the pattern required a fixed first
3103 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3104
3105 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3106 starting at the last subject character, bytes beyond the end of the subject
3107 string were read.
3108
3109 11. Renamed the variable "class" as "classbits" to make life easier for C++
3110 users. (Previously there was a macro definition, but it apparently wasn't
3111 enough.)
3112
3113 12. Added the new field "tables" to the extra data so that tables can be passed
3114 in at exec time, or the internal tables can be re-selected. This allows
3115 a compiled regex to be saved and re-used at a later time by a different
3116 program that might have everything at different addresses.
3117
3118 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3119 -R library as well as a -L library.
3120
3121 14. The debugging options of pcretest (-d on the command line or D on a
3122 pattern) showed incorrect output for anything following an extended class
3123 that contained multibyte characters and which was followed by a quantifier.
3124
3125 15. Added optional support for general category Unicode character properties
3126 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3127 support. It adds about 90K to the size of the library. The meanings of the
3128 inbuilt class escapes such as \d and \s have NOT been changed.
3129
3130 16. Updated pcredemo.c to include calls to free() to release the memory for the
3131 compiled pattern.
3132
3133 17. The generated file chartables.c was being created in the source directory
3134 instead of in the building directory. This caused the build to fail if the
3135 source directory was different from the building directory, and was
3136 read-only.
3137
3138 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3139 file. No doubt somebody will tell me if they don't make sense... Also added
3140 Dan Mooney's comments about building on OpenVMS.
3141
3142 19. Added support for partial matching via the PCRE_PARTIAL option for
3143 pcre_exec() and the \P data escape in pcretest.
3144
3145 20. Extended pcretest with 3 new pattern features:
3146
3147 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3148 write the compiled pattern to the file whose name is "rest-of-line".
3149 This is a straight binary dump of the data, with the saved pointer to
3150 the character tables forced to be NULL. The study data, if any, is
3151 written too. After writing, pcretest reads a new pattern.
3152
3153 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3154 compiled pattern from the given file. There must not be any
3155 occurrences of "<" in the file name (pretty unlikely); if there are,
3156 pcretest will instead treat the initial "<" as a pattern delimiter.
3157 After reading in the pattern, pcretest goes on to read data lines as
3158 usual.
3159
3160 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3161 and 16-bit fields in a compiled pattern, to simulate a pattern that
3162 was compiled on a host of opposite endianness.
3163
3164 21. The pcre-exec() function can now cope with patterns that were compiled on
3165 hosts of opposite endianness, with this restriction:
3166
3167 As for any compiled expression that is saved and used later, the tables
3168 pointer field cannot be preserved; the extra_data field in the arguments
3169 to pcre_exec() should be used to pass in a tables address if a value
3170 other than the default internal tables were used at compile time.
3171
3172 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3173 now diagnosed as an error. Previously, most of the time, a negative number
3174 would have been treated as zero, but if in addition "ovector" was passed as
3175 NULL, a crash could occur.
3176
3177 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3178 new versions from the libtool 1.5 distribution (the last one is a copy of
3179 a file called libtool.m4). This seems to have fixed the need to patch
3180 "configure" to support Darwin 1.3 (which I used to do). However, I still
3181 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3182 workstation).
3183
3184 24. Changed the PCRE licence to be the more standard "BSD" licence.
3185
3186
3187 Version 4.5 01-Dec-03
3188 ---------------------
3189
3190 1. There has been some re-arrangement of the code for the match() function so
3191 that it can be compiled in a version that does not call itself recursively.
3192 Instead, it keeps those local variables that need separate instances for
3193 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3194 needs to "recurse". Keeping track of where control must go is done by means
3195 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3196 hide most of the details from the main code, and operates only if
3197 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3198 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3199 operating.
3200
3201 To make it easier for callers to provide specially tailored get/free
3202 functions for this usage, two new functions, pcre_stack_malloc, and
3203 pcre_stack_free, are used. They are always called in strict stacking order,
3204 and the size of block requested is always the same.
3205
3206 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3207 PCRE has been compiled to use the stack or the heap for recursion. The
3208 -C option of pcretest uses this to show which version is compiled.
3209
3210 A new data escape \S, is added to pcretest; it causes the amounts of store
3211 obtained and freed by both kinds of malloc/free at match time to be added
3212 to the output.
3213
3214 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3215 what's available on my current Linux desktop machine.
3216
3217 3. When matching a UTF-8 string, the test for a valid string at the start has
3218 been extended. If start_offset is not zero, PCRE now checks that it points
3219 to a byte that is the start of a UTF-8 character. If not, it returns
3220 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3221 this is necessary because there may be backward assertions in the pattern.
3222 When matching the same subject several times, it may save resources to use
3223 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3224
3225 4. The code for checking the validity of UTF-8 strings has been tightened so
3226 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3227 containing "overlong sequences".
3228
3229 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3230 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3231 should have been "&", but it just so happened that all the cases this let
3232 through by mistake were picked up later in the function.
3233
3234 6. I had used a variable called "isblank" - this is a C99 function, causing
3235 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3236
3237 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3238 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3239 so the version is shown; (c) stop "make test" echoing "./RunTest".
3240
3241 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3242
3243 9. The prototype for memmove() for systems that don't have it was using
3244 size_t, but the inclusion of the header that defines size_t was later. I've
3245 moved the #includes for the C headers earlier to avoid this.
3246
3247 10. Added some adjustments to the code to make it easier to compiler on certain
3248 special systems:
3249
3250 (a) Some "const" qualifiers were missing.
3251 (b) Added the macro EXPORT before all exported functions; by default this
3252 is defined to be empty.
3253 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3254 that it reads its output file name as an argument instead of writing
3255 to the standard output and assuming this can be redirected.
3256
3257 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3258 class containing characters with values greater than 255, PCRE compilation
3259 went into a loop.
3260
3261 12. A recursive reference to a subpattern that was within another subpattern
3262 that had a minimum quantifier of zero caused PCRE to crash. For example,
3263 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3264 recursion. If the recursively-called subpattern itself had a zero repeat,
3265 that was OK.
3266
3267 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3268 buffer into which it was copied (for escape processing) was still set at
3269 1024, so long lines caused crashes.
3270
3271 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3272 "internal error: code overflow...". This applied to any character class
3273 that was followed by a possessive quantifier.
3274
3275 15. Modified the Makefile to add libpcre.la as a prerequisite for
3276 libpcreposix.la because I was told this is needed for a parallel build to
3277 work.
3278
3279 16. If a pattern that contained .* following optional items at the start was
3280 studied, the wrong optimizing data was generated, leading to matching
3281 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3282 matching string must start with a or b or c. The correct conclusion for
3283 this pattern is that a match can start with any character.
3284
3285
3286 Version 4.4 13-Aug-03
3287 ---------------------
3288
3289 1. In UTF-8 mode, a character class containing characters with values between
3290 127 and 255 was not handled correctly if the compiled pattern was studied.
3291 In fixing this, I have also improved the studying algorithm for such
3292 classes (slightly).
3293
3294 2. Three internal functions had redundant arguments passed to them. Removal
3295 might give a very teeny performance improvement.
3296
3297 3. Documentation bug: the value of the capture_top field in a callout is *one
3298 more than* the number of the hightest numbered captured substring.
3299
3300 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3301 in incorrectly linking with a previously installed version. They now link
3302 explicitly with libpcre.la.
3303
3304 5. configure.in no longer needs to recognize Cygwin specially.
3305
3306 6. A problem in pcre.in for Windows platforms is fixed.
3307
3308 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3309 pcretest, it used to include the size of the study block as part of its
3310 output. Unfortunately, the structure contains a field that has a different
3311 size on different hardware architectures. This meant that the tests that
3312 showed this size failed. As the block is currently always of a fixed size,
3313 this information isn't actually particularly useful in pcretest output, so
3314 I have just removed it.
3315
3316 8. Three pre-processor statements accidentally did not start in column 1.
3317 Sadly, there are *still* compilers around that complain, even though
3318 standard C has not required this for well over a decade. Sigh.
3319
3320 9. In pcretest, the code for checking callouts passed small integers in the
3321 callout_data field, which is a void * field. However, some picky compilers
3322 complained about the casts involved for this on 64-bit systems. Now
3323 pcretest passes the address of the small integer instead, which should get
3324 rid of the warnings.
3325
3326 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3327 both compile and run time, and gives an error if an invalid UTF-8 sequence
3328 is found. There is a option for disabling this check in cases where the
3329 string is known to be correct and/or the maximum performance is wanted.
3330
3331 11. In response to a bug report, I changed one line in Makefile.in from
3332
3333 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3334 to
3335 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3336
3337 to look similar to other lines, but I have no way of telling whether this
3338 is the right thing to do, as I do not use Windows. No doubt I'll get told
3339 if it's wrong...
3340
3341
3342 Version 4.3 21-May-03
3343 ---------------------
3344
3345 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3346 Makefile.
3347
3348 2. Some refactoring to improve the quality of the code:
3349
3350 (i) The utf8_table... variables are now declared "const".
3351
3352 (ii) The code for \cx, which used the "case flipping" table to upper case
3353 lower case letters, now just substracts 32. This is ASCII-specific,
3354 but the whole concept of \cx is ASCII-specific, so it seems
3355 reasonable.
3356
3357 (iii) PCRE was using its character types table to recognize decimal and
3358 hexadecimal digits in the pattern. This is silly, because it handles
3359 only 0-9, a-f, and A-F, but the character types table is locale-
3360 specific, which means strange things might happen. A private
3361 table is now used for this - though it costs 256 bytes, a table is
3362 much faster than multiple explicit tests. Of course, the standard
3363 character types table is still used for matching digits in subject
3364 strings against \d.
3365
3366 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3367 ending in _t are). So I've renamed it as ESC_tee.
3368
3369 3. The first argument for regexec() in the POSIX wrapper should have been
3370 defined as "const".
3371
3372 4. Changed pcretest to use malloc() for its buffers so that they can be
3373 Electric Fenced for debugging.
3374
3375 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3376 to read one or more bytes before the start of the subject string. Often this
3377 had no effect on PCRE's behaviour, but in some circumstances it could
3378 provoke a segmentation fault.
3379
3380 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3381 to try to read one or more bytes before the start of the subject string.
3382
3383 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3384 UTF-8 support could misbehave in various ways if the subject string
3385 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3386 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3387 back over UTF-8 characters.)
3388
3389
3390 Version 4.2 14-Apr-03
3391 ---------------------
3392
3393 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3394
3395 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3396 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3397 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3398 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3399 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3400 and BUILD_EXEEXT
3401 Note: automatic setting of the BUILD variables is not yet working
3402 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3403 compile-time but not at link-time
3404 [LINK]: use for linking executables only
3405 make different versions for Windows and non-Windows
3406 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3407 libraries
3408 [LINK_FOR_BUILD]: new variable
3409 [OBJEXT]: use throughout
3410 [EXEEXT]: use throughout
3411 <winshared>: new target
3412 <wininstall>: new target
3413 <dftables.o>: use native compiler
3414 <dftables>: use native linker
3415 <install>: handle Windows platform correctly
3416 <clean>: ditto
3417 <check>: ditto
3418 copy DLL to top builddir before testing
3419
3420 As part of these changes, -no-undefined was removed again. This was reported
3421 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3422 in any case.
3423
3424 3. Some tidies to get rid of compiler warnings:
3425
3426 . In the match_data structure, match_limit was an unsigned long int, whereas
3427 match_call_count was an int. I've made them both unsigned long ints.
3428
3429 . In pcretest the fact that a const uschar * doesn't automatically cast to
3430 a void * provoked a warning.
3431
3432 . Turning on some more compiler warnings threw up some "shadow" variables
3433 and a few more missing casts.
3434
3435 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3436 option, a class that contained a single character with a value between 128
3437 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3438
3439 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3440 option, a class that contained several characters, but with at least one
3441 whose value was between 128 and 255 caused PCRE to crash.
3442
3443
3444 Version 4.1 12-Mar-03
3445 ---------------------
3446
3447 1. Compiling with gcc -pedantic found a couple of places where casts were
3448 needed, and a string in dftables.c that was longer than standard compilers are
3449 required to support.
3450
3451 2. Compiling with Sun's compiler found a few more places where the code could
3452 be tidied up in order to avoid warnings.
3453
3454 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3455 first of these names is deprecated in the latest Autoconf in favour of the name
3456 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3457 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3458 analogy I have changed it to CFLAGS_FOR_BUILD.
3459
3460 4. Added -no-undefined to the linking command in the Makefile, because this is
3461 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3462 linking step for the pcreposix library.
3463
3464 5. PCRE was failing to diagnose the case of two named groups with the same
3465 name.
3466
3467 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3468 literal character that is needed in the subject for a match, and scans along to
3469 ensure that it is present before embarking on the full matching process. This
3470 saves time in cases of nested unlimited repeats that are never going to match.
3471 Problem: the scan can take a lot of time if the subject is very long (e.g.
3472 megabytes), thus penalizing straightforward matches. It is now done only if the
3473 amount of subject to be scanned is less than 1000 bytes.
3474
3475 7. A lesser problem with the same optimization is that it was recording the
3476 first character of an anchored pattern as "needed", thus provoking a search
3477 right along the subject, even when the first match of the pattern was going to
3478 fail. The "needed" character is now not set for anchored patterns, unless it
3479 follows something in the pattern that is of non-fixed length. Thus, it still
3480 fulfils its original purpose of finding quick non-matches in cases of nested
3481 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3482
3483
3484 Version 4.0 17-Feb-03
3485 ---------------------
3486
3487 1. If a comment in an extended regex that started immediately after a meta-item
3488 extended to the end of string, PCRE compiled incorrect data. This could lead to
3489 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3490
3491 2. Moved to autoconf 2.53 and libtool 1.4.2.
3492
3493 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3494 the special perltest8 script is no longer needed - all the tests can be run
3495 from a single perltest script.
3496
3497 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3498 by \s. It has now been removed in PCRE. This means it isn't recognized as
3499 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3500 class [:space:] *does* include VT, thereby creating a mess.
3501
3502 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3503 space and tab.
3504
3505 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3506 its new features into the main test script, reducing the number of scripts.
3507
3508 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3509 were backward compatible, and made the (?i) apply to the whole pattern, as if
3510 /i were given. Now it behaves more logically, and applies the option setting
3511 only to what follows. PCRE has been changed to follow suit. However, if it
3512 finds options settings right at the start of the pattern, it extracts them into
3513 the global options, as before. Thus, they show up in the info data.
3514
3515 8. Added support for the \Q...\E escape sequence. Characters in between are
3516 treated as literals. This is slightly different from Perl in that $ and @ are
3517 also handled as literals inside the quotes. In Perl, they will cause variable
3518 interpolation. Note the following examples:
3519
3520 Pattern PCRE matches Perl matches
3521
3522 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3523 \Qabc\$xyz\E abc\$xyz abc\$xyz
3524 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3525
3526 For compatibility with Perl, \Q...\E sequences are recognized inside character
3527 classes as well as outside them.
3528
3529 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3530 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3531 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3532 signed/unsigned warnings.
3533
3534 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3535 option for pcretest, so I've replaced it by a simple function that does just
3536 that job.
3537
3538 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3539 "pcregrep -".
3540
3541 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3542 Java package. This provides some syntactic sugar for simple cases of what my
3543 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3544 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3545 item, you can use this simplified notation. Note that only makes sense with
3546 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3547 greediness, whatever the setting of the PCRE_UNGREEDY option.
3548
3549 13. A change of greediness default within a pattern was not taking effect at
3550 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3551 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3552 was abstracted outside.
3553
3554 14. PCRE now supports the \G assertion. It is true when the current matching
3555 position is at the start point of the match. This differs from \A when the
3556 starting offset is non-zero. Used with the /g option of pcretest (or similar
3557 code), it works in the same way as it does for Perl's /g option. If all
3558 alternatives of a regex begin with \G, the expression is anchored to the start
3559 match position, and the "anchored" flag is set in the compiled expression.
3560
3561 15. Some bugs concerning the handling of certain option changes within patterns
3562 have been fixed. These applied to options other than (?ims). For example,
3563 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3564 been the other way round. Some of this was related to change 7 above.
3565
3566 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3567 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3568 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3569 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3570
3571 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3572 mode. Unlike ".", it always matches newline, whatever the setting of
3573 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3574 assertions. Perl allows it, but it doesn't (in general) work because it can't
3575 calculate the length of the lookbehind. At least, that's the case for Perl
3576 5.8.0 - I've been told they are going to document that it doesn't work in
3577 future.
3578
3579 18. Added an error diagnosis for escapes that PCRE does not support: these are
3580 \L, \l, \N, \P, \p, \U, \u, and \X.
3581
3582 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3583 reading past the end of the pattern in cases such as /[abcd/.
3584
3585 20. PCRE was getting more memory than necessary for patterns with classes that
3586 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3587
3588 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3589 compiling PCRE for use with Virtual Pascal.
3590
3591 22. Small fix to the Makefile to make it work properly if the build is done
3592 outside the source tree.
3593
3594 23. Added a new extension: a condition to go with recursion. If a conditional
3595 subpattern starts with (?(R) the "true" branch is used if recursion has
3596 happened, whereas the "false" branch is used only at the top level.
3597
3598 24. When there was a very long string of literal characters (over 255 bytes
3599 without UTF support, over 250 bytes with UTF support), the computation of how
3600 much memory was required could be incorrect, leading to segfaults or other
3601 strange effects.
3602
3603 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3604 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3605 there was a subsequent back reference to those brackets. This meant that, for
3606 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3607 possible to check for precisely this case. All we can do is abandon the
3608 optimization if .* occurs inside capturing brackets when there are any back
3609 references whatsoever. (See below for a better fix that came later.)
3610
3611 26. The handling of the optimization for finding the first character of a
3612 non-anchored pattern, and for finding a character that is required later in the
3613 match were failing in some cases. This didn't break the matching; it just
3614 failed to optimize when it could. The way this is done has been re-implemented.
3615
3616 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3617
3618 28. Added a new feature that provides some of the functionality that Perl
3619 provides with (?{...}). The facility is termed a "callout". The way it is done
3620 in PCRE is for the caller to provide an optional function, by setting
3621 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3622 global variable. By default it is unset, which disables all calling out. To get
3623 the function called, the regex must include (?C) at appropriate points. This
3624 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3625 This provides a means of identifying different callout points. When PCRE
3626 reaches such a point in the regex, if pcre_callout has been set, the external
3627 function is called. It is provided with data in a structure called
3628 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3629 matching continues; if it returns a non-zero value, the match at the current
3630 point fails. However, backtracking will occur if possible. [This was changed
3631 later and other features added - see item 49 below.]
3632
3633 29. pcretest is upgraded to test the callout functionality. It provides a
3634 callout function that displays information. By default, it shows the start of
3635 the match and the current position in the text. There are some new data escapes
3636 to vary what happens:
3637
3638 \C+ in addition, show current contents of captured substrings
3639 \C- do not supply a callout function
3640 \C!n return 1 when callout number n is reached
3641 \C!n!m return 1 when callout number n is reached for the mth time
3642
3643 30. If pcregrep was called with the -l option and just a single file name, it
3644 output "<stdin>" if a match was found, instead of the file name.
3645
3646 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3647 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3648 pcre_exec(). This saves a malloc/free per call. The default value of
3649 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3650 when configuring.
3651
3652 32. The default maximum size of a compiled pattern is 64K. There have been a
3653 few cases of people hitting this limit. The code now uses macros to handle the
3654 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3655 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3656 configuring. Tests 2 and 5 work only with 2-byte links because they output
3657 debugging information about compiled patterns.
3658
3659 33. Internal code re-arrangements:
3660
3661 (a) Moved the debugging function for printing out a compiled regex into
3662 its own source file (printint.c) and used #include to pull it into
3663 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3664 separate copies.
3665
3666 (b) Defined the list of op-code names for debugging as a macro in
3667 internal.h so that it is next to the definition of the opcodes.
3668
3669 (c) Defined a table of op-code lengths for simpler skipping along compiled
3670 code. This is again a macro in internal.h so that it is next to the
3671 definition of the opcodes.
3672
3673 34. Added support for recursive calls to individual subpatterns, along the
3674 lines of Robin Houston's patch (but implemented somewhat differently).
3675
3676 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3677 allow it to read and process whole directories in Win32. This code was
3678 contributed by Lionel Fourquaux; it has not been tested by me.
3679
3680 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3681 used to name a group. Names consist of alphanumerics and underscores, and must
3682 be unique. Back references use the syntax (?P=name) and recursive calls use
3683 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3684 numbers. The function pcre_fullinfo() can be used after compilation to extract
3685 a name/number map. There are three relevant calls:
3686
3687 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3688 PCRE_INFO_NAMECOUNT yields the number of entries
3689 PCRE_INFO_NAMETABLE yields a pointer to the map.
3690
3691 The map is a vector of fixed-size entries. The size of each entry depends on
3692 the length of the longest name used. The first two bytes of each entry are the
3693 group number, most significant byte first. There follows the corresponding
3694 name, zero terminated. The names are in alphabetical order.
3695
3696 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3697 case instead of 255. Making it the same both with and without UTF-8 support
3698 means that the same test output works with both.
3699
3700 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3701 calling malloc() with a zero argument.
3702
3703 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3704 optimization. I've improved things by keeping a bitmap of backreferences with
3705 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3706 fact referenced, the optimization can be applied. It is unlikely that a
3707 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3708 the match to follow \n) will appear inside brackets with a number greater than
3709 31, but if it does, any back reference > 31 suppresses the optimization.
3710
3711 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3712 of disabling numbered capturing parentheses. Any opening parenthesis that is
3713 not followed by ? behaves as if it were followed by ?: but named parentheses
3714 can still be used for capturing (and they will acquire numbers in the usual
3715 way).
3716
3717 41. Redesigned the return codes from the match() function into yes/no/error so
3718 that errors can be passed back from deep inside the nested calls. A malloc
3719 failure while inside a recursive subpattern call now causes the
3720 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3721
3722 42. It is now possible to set a limit on the number of times the match()
3723 function is called in a call to pcre_exec(). This facility makes it possible to
3724 limit the amount of recursion and backtracking, though not in a directly
3725 obvious way, because the match() function is used in a number of different
3726 circumstances. The count starts from zero for each position in the subject
3727 string (for non-anchored patterns). The default limit is, for compatibility, a
3728 large number, namely 10 000 000. You can change this in two ways:
3729
3730 (a) When configuring PCRE before making, you can use --with-match-limit=n
3731 to set a default value for the compiled library.
3732
3733 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3734 a different value is set. See 45 below.
3735
3736 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3737
3738 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3739 of things that can be changed at compile time. The first argument specifies
3740 what is wanted and the second points to where the information is to be placed.
3741 The current list of available information is:
3742
3743 PCRE_CONFIG_UTF8
3744
3745 The output is an integer that is set to one if UTF-8 support is available;
3746 otherwise it is set to zero.
3747
3748 PCRE_CONFIG_NEWLINE
3749
3750 The output is an integer that it set to the value of the code that is used for
3751 newline. It is either LF (10) or CR (13).
3752
3753 PCRE_CONFIG_LINK_SIZE
3754
3755 The output is an integer that contains the number of bytes used for internal
3756 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3757
3758 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3759
3760 The output is an integer that contains the threshold above which the POSIX
3761 interface uses malloc() for output vectors. See item 31 above.
3762
3763 PCRE_CONFIG_MATCH_LIMIT
3764
3765 The output is an unsigned integer that contains the default limit of the number
3766 of match() calls in a pcre_exec() execution. See 42 above.
3767
3768 44. pcretest has been upgraded by the addition of the -C option. This causes it
3769 to extract all the available output from the new pcre_config() function, and to
3770 output it. The program then exits immediately.
3771
3772 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3773 order to support additional features. One way would have been to define
3774 pcre_exec2() (for example) with extra arguments, but this would not have been
3775 extensible, and would also have required all calls to the original function to
3776 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3777 is used for passing in "extra" data from pcre_study().
3778
3779 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3780 contains the following fields:
3781
3782 flags a bitmap indicating which of the following fields are set
3783 study_data opaque data from pcre_study()
3784 match_limit a way of specifying a limit on match() calls for a specific
3785 call to pcre_exec()
3786 callout_data data for callouts (see 49 below)
3787
3788 The flag bits are also defined in pcre.h, and are
3789
3790 PCRE_EXTRA_STUDY_DATA
3791 PCRE_EXTRA_MATCH_LIMIT
3792 PCRE_EXTRA_CALLOUT_DATA
3793
3794 The pcre_study() function now returns one of these new pcre_extra blocks, with
3795 the actual study data pointed to by the study_data field, and the
3796 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3797 before. That is, this change is entirely upwards-compatible and requires no
3798 change to existing code.
3799
3800 If you want to pass in additional data to pcre_exec(), you can either place it
3801 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3802 block.
3803
3804 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3805 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3806 times with different match limits, until it finds the minimum value needed for
3807 pcre_exec() to complete. The value is then output. This can be instructive; for
3808 most simple matches the number is quite small, but for pathological cases it
3809 gets very large very quickly.
3810
3811 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3812 returns the size of the data block pointed to by the study_data field in a
3813 pcre_extra block, that is, the value that was passed as the argument to
3814 pcre_malloc() when PCRE was getting memory in which to place the information
3815 created by pcre_study(). The fourth argument should point to a size_t variable.
3816 pcretest has been extended so that this information is shown after a successful
3817 pcre_study() call when information about the compiled regex is being displayed.
3818
3819 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3820 because what follows is always an absolute path. (Later: it turns out that this
3821 is more than cosmetic for MinGW, because it doesn't like empty path
3822 components.)
3823
3824 49. Some changes have been made to the callout feature (see 28 above):
3825
3826 (i) A callout function now has three choices for what it returns:
3827
3828 0 => success, carry on matching
3829 > 0 => failure at this point, but backtrack if possible
3830 < 0 => serious error, return this value from pcre_exec()
3831
3832 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3833 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3834 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3835 use by callout functions. It will never be used by PCRE itself.
3836
3837 (ii) The pcre_extra structure (see 45 above) has a void * field called
3838 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3839 pcre_callout_block structure has a field of the same name. The contents of
3840 the field passed in the pcre_extra structure are passed to the callout
3841 function in the corresponding field in the callout block. This makes it
3842 easier to use the same callout-containing regex from multiple threads. For
3843 testing, the pcretest program has a new data escape
3844
3845 \C*n pass the number n (may be negative) as callout_data
3846
3847 If the callout function in pcretest receives a non-zero value as
3848 callout_data, it returns that value.
3849
3850 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3851 there were some redundant $(CFLAGS) in commands that are now specified as
3852 $(LINK), which already includes $(CFLAGS).
3853
3854 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3855 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3856 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3857 one-byte characters throughout. Note that case-insensitive matching applies
3858 only to characters whose values are less than 256. PCRE doesn't support the
3859 notion of cases for higher-valued characters.
3860
3861 (i) A character class whose characters are all within 0-255 is handled as
3862 a bit map, and the map is inverted for negative classes. Previously, a
3863 character > 255 always failed to match such a class; however it should
3864 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3865
3866 (ii) A negated character class with a single character < 255 is coded as
3867 "not this character" (OP_NOT). This wasn't working properly when the test
3868 character was multibyte, either singly or repeated.
3869
3870 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3871 mode, for example: \x{100}{2,3}.
3872
3873 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3874 singly or repeated) now correctly test multibyte characters. However,
3875 PCRE doesn't recognize any characters with values greater than 255 as
3876 digits, spaces, or word characters. Such characters always match \D, \S,
3877 and \W, and never match \d, \s, or \w.
3878
3879 (v) Classes may now contain characters and character ranges with values
3880 greater than 255. For example: [ab\x{100}-\x{400}].
3881
3882 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3883 PCRE in UTF-8 mode.
3884
3885 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3886 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3887 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3888 value.)
3889
3890 53. The single man page has become too large. I have therefore split it up into
3891 a number of separate man pages. These also give rise to individual HTML pages;
3892 these are now put in a separate directory, and there is an index.html page that
3893 lists them all. Some hyperlinking between the pages has been installed.
3894
3895 54. Added convenience functions for handling named capturing parentheses.
3896
3897 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3898 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3899 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3900 are faulted.
3901
3902 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3903 calling configure. These values are used when compiling the dftables.c program
3904 which is run to generate the source of the default character tables. They
3905 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3906 you will need to set these values.
3907
3908 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3909
3910
3911 Version 3.9 02-Jan-02
3912 ---------------------
3913
3914 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3915
3916 2. If --disable-static was given, the building process failed when trying to
3917 build pcretest and pcregrep. (For some reason it was using libtool to compile
3918 them, which is not right, as they aren't part of the library.)
3919
3920
3921 Version 3.8 18-Dec-01
3922 ---------------------
3923
3924 1. The experimental UTF-8 code was completely screwed up. It was packing the
3925 bytes in the wrong order. How dumb can you get?
3926
3927
3928 Version 3.7 29-Oct-01
3929 ---------------------
3930
3931 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3932 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3933 this didn't happen under Solaris 8, where I normally test things.
3934
3935 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3936 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3937 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3938 this fix an hour or so after the initial 3.7 release.)
3939
3940
3941 Version 3.6 23-Oct-01
3942 ---------------------
3943
3944 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3945 offsets passed as NULL with zero offset count.
3946
3947 2. The config.guess and config.sub files had not been updated when I moved to
3948 the latest autoconf.
3949
3950
3951 Version 3.5 15-Aug-01
3952 ---------------------
3953
3954 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3955 had been forgotten.
3956
3957 2. By using declared but undefined structures, we can avoid using "void"
3958 definitions in pcre.h while keeping the internal definitions of the structures
3959 private.
3960
3961 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3962 user point of view, this means that both static and shared libraries are built
3963 by default, but this can be individually controlled. More of the work of
3964 handling this static/shared cases is now inside libtool instead of PCRE's make
3965 file.
3966
3967 4. The pcretest utility is now installed along with pcregrep because it is
3968 useful for users (to test regexs) and by doing this, it automatically gets
3969 relinked by libtool. The documentation has been turned into a man page, so
3970 there are now .1, .txt, and .html versions in /doc.
3971
3972 5. Upgrades to pcregrep:
3973 (i) Added long-form option names like gnu grep.
3974 (ii) Added --help to list all options with an explanatory phrase.
3975 (iii) Added -r, --recursive to recurse into sub-directories.
3976 (iv) Added -f, --file to read patterns from a file.
3977
3978 6. pcre_exec() was referring to its "code" argument before testing that
3979 argument for NULL (and giving an error if it was NULL).
3980
3981 7. Upgraded Makefile.in to allow for compiling in a different directory from
3982 the source directory.
3983
3984 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3985 options bits, the pointer it was passed was to an int instead of to an unsigned
3986 long int. This mattered only on 64-bit systems.
3987
3988 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3989 generated) instead of pcre.in, which it its source. Also made the same change
3990 in several of the .c files.
3991
3992 10. A new release of gcc defines printf() as a macro, which broke pcretest
3993 because it had an ifdef in the middle of a string argument for printf(). Fixed
3994 by using separate calls to printf().
3995
3996 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3997 script, to force use of CR or LF instead of \n in the source. On non-Unix
3998 systems, the value can be set in config.h.
3999
4000 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4001 absolute limit. Changed the text of the error message to make this clear, and
4002 likewise updated the man page.
4003
4004 13. The limit of 99 on the number of capturing subpatterns has been removed.
4005 The new limit is 65535, which I hope will not be a "real" limit.
4006
4007
4008 Version 3.4 22-Aug-00
4009 ---------------------
4010
4011 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4012
4013 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4014
4015
4016 Version 3.3 01-Aug-00
4017 ---------------------
4018
4019 1. If an octal character was given, but the value was greater than \377, it
4020 was not getting masked to the least significant bits, as documented. This could
4021 lead to crashes in some systems.
4022
4023 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4024 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4025
4026 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4027 These just pass their arguments on to (pcre_free)(), but they are provided
4028 because some uses of PCRE bind it to non-C systems that can call its functions,
4029 but cannot call free() or pcre_free() directly.
4030
4031 4. Add "make test" as a synonym for "make check". Corrected some comments in
4032 the Makefile.
4033
4034 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4035 Makefile.
4036
4037 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4038 command called pgrep for grepping around the active processes.
4039
4040 7. Added the beginnings of support for UTF-8 character strings.
4041
4042 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4043 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4044 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4045 out for the ar command.)
4046
4047
4048 Version 3.2 12-May-00
4049 ---------------------
4050
4051 This is purely a bug fixing release.
4052
4053 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4054 of ZA. This was just one example of several cases that could provoke this bug,
4055 which was introduced by change 9 of version 2.00. The code for breaking
4056 infinite loops after an iteration that matches an empty string was't working
4057 correctly.
4058
4059 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4060 when matched against abbab (for example). After matching an empty string, it
4061 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4062 caused it to match further down the string than it should.
4063
4064 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4065 was there because it doesn't seem to be needed, and it causes trouble on some
4066 systems, as it is not a Standard C header. It has been removed.
4067
4068 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4069 were reported on the Macintosh. The changes were from
4070
4071 while ((c = *(++ptr)) != 0 && c != '\n');
4072 to
4073 while ((c = *(++ptr)) != 0 && c != '\n') ;
4074
4075 Totally extraordinary, but if that's what it takes...
4076
4077 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4078 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4079 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4080 assumes the way PCRE uses memmove() (always moving upwards).
4081
4082 6. PCRE is being used in one environment where strchr() is not available. There
4083 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4084 faster code anyway.
4085
4086
4087 Version 3.1 09-Feb-00
4088 ---------------------
4089
4090 The only change in this release is the fixing of some bugs in Makefile.in for
4091 the "install" target:
4092
4093 (1) It was failing to install pcreposix.h.
4094
4095 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4096
4097
4098 Version 3.0 01-Feb-00
4099 ---------------------
4100
4101 1. Add support for the /+ modifier to perltest (to output $` like it does in
4102 pcretest).
4103
4104 2. Add support for the /g modifier to perltest.
4105
4106 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4107 matches null strings.
4108
4109 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4110 pattern. Perl treats empty patterns specially - it reuses the most recent
4111 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4112 effect.
4113
4114 5. The POSIX interface was broken in that it was just handing over the POSIX
4115 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4116 required a bigger vector, with some working space on the end. This means that
4117 the POSIX wrapper now has to get and free some memory, and copy the results.
4118
4119 6. Added some simple autoconf support, placing the test data and the
4120 documentation in separate directories, re-organizing some of the
4121 information files, and making it build pcre-config (a GNU standard). Also added
4122 libtool support for building PCRE as a shared library, which is now the
4123 default.
4124
4125 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4126 09 are not valid octal constants. Single digits will be used for minor values
4127 less than 10.
4128
4129 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4130 existing programs that set these in the POSIX interface can use PCRE without
4131 modification.
4132
4133 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4134 return all that pcre_info() returns, plus additional data. The pcre_info()
4135 function is retained for compatibility, but is considered to be obsolete.
4136
4137 10. Added experimental recursion feature (?R) to handle one common case that
4138 Perl 5.6 will be able to do with (?p{...}).
4139
4140 11. Added support for POSIX character classes like [:alpha:], which Perl is
4141 adopting.
4142
4143
4144 Version 2.08 31-Aug-99
4145 ----------------------
4146
4147 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4148 trying to match at the startoffset position, but instead was moving forward to
4149 the next newline as if a previous match had failed.
4150
4151 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4152 and could get into a loop if a null string was matched other than at the start
4153 of the subject.
4154
4155 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4156 be distinguished at compile time, and for completeness also added PCRE_DATE.
4157
4158 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4159 in GnuWin32 environments.
4160
4161
4162 Version 2.07 29-Jul-99
4163 ----------------------
4164
4165 1. The documentation is now supplied in plain text form and HTML as well as in
4166 the form of man page sources.
4167
4168 2. C++ compilers don't like assigning (void *) values to other pointer types.
4169 In particular this affects malloc(). Although there is no problem in Standard
4170 C, I've put in casts to keep C++ compilers happy.
4171
4172 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4173 should be (const char *).
4174
4175 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4176 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4177 However, I haven't made this a standard facility. The documentation doesn't
4178 mention it, and the Makefile doesn't support it.
4179
4180 5. The Makefile now contains an "install" target, with editable destinations at
4181 the top of the file. The pcretest program is not installed.
4182
4183 6. pgrep -V now gives the PCRE version number and date.
4184
4185 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4186 causing the entire string to be ignored, instead of just the last character.
4187
4188 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4189 non-matching string, it can take a very, very long time, even for strings of
4190 quite modest length, because of the nested recursion. PCRE now does better in
4191 some of these cases. It does this by remembering the last required literal
4192 character in the pattern, and pre-searching the subject to ensure it is present
4193 before running the real match. In other words, it applies a heuristic to detect
4194 some types of certain failure quickly, and in the above example, if presented
4195 with a string that has no trailing " it gives "no match" very quickly.
4196
4197 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4198 other alternatives are tried instead.
4199
4200
4201 Version 2.06 09-Jun-99
4202 ----------------------
4203
4204 1. Change pcretest's output for amount of store used to show just the code
4205 space, because the remainder (the data block) varies in size between 32-bit and
4206 64-bit systems.
4207
4208 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4209 start matching at. This allows lookbehinds to work when searching for multiple
4210 occurrences in a string.
4211
4212 3. Added additional options to pcretest for testing multiple occurrences:
4213
4214 /+ outputs the rest of the string that follows a match
4215 /g loops for multiple occurrences, using the new startoffset argument
4216 /G loops for multiple occurrences by passing an incremented pointer
4217
4218 4. PCRE wasn't doing the "first character" optimization for patterns starting
4219 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4220 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4221 the letter 'x'. On long subject strings, this gives a significant speed-up.
4222
4223
4224 Version 2.05 21-Apr-99
4225 ----------------------
4226
4227 1. Changed the type of magic_number from int to long int so that it works
4228 properly on 16-bit systems.
4229
4230 2. Fixed a bug which caused patterns starting with .* not to work correctly
4231 when the subject string contained newline characters. PCRE was assuming
4232 anchoring for such patterns in all cases, which is not correct because .* will
4233 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4234 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4235 must be retried after every newline in the subject.
4236
4237
4238 Version 2.04 18-Feb-99
4239 ----------------------
4240
4241 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4242 computation of the store needed to hold the pattern was incorrect (too large).
4243 If such patterns were nested a few deep, this could multiply and become a real
4244 problem.
4245
4246 2. Added /M option to pcretest to show the memory requirement of a specific
4247 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4248
4249 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4250 compiled in such a way that the backtracking after subsequent failure was
4251 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4252 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4253
4254
4255 Version 2.03 02-Feb-99
4256 ----------------------
4257
4258 1. Fixed typo and small mistake in man page.
4259
4260 2. Added 4th condition (GPL supersedes if conflict) and created separate
4261 LICENCE file containing the conditions.
4262
4263 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4264 Perl, that is the internal \ allows the delimiter to be included in the
4265 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4266 the final delimiter, add \ to the end of the pattern (to test the error).
4267
4268 4. Added the convenience functions for extracting substrings after a successful
4269 match. Updated pcretest to make it able to test these functions.
4270
4271
4272 Version 2.02 14-Jan-99
4273 ----------------------
4274
4275 1. Initialized the working variables associated with each extraction so that
4276 their saving and restoring doesn't refer to uninitialized store.
4277
4278 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4279 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4280 fix the problem.
4281
4282 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4283 calls, and wasn't printing the correct value for compiling calls. Increased the
4284 default value of LOOPREPEAT, and the number of significant figures in the
4285 times.
4286
4287 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4288
4289 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4290 a building problem on Windows NT with a FAT file system.
4291
4292
4293 Version 2.01 21-Oct-98
4294 ----------------------
4295
4296 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4297 to character tables built by pcre_maketables() in the current locale. If NULL
4298 is passed, the default tables are used.
4299
4300
4301 Version 2.00 24-Sep-98
4302 ----------------------
4303
4304 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4305 it any more.
4306
4307 2. Allow quantification of (?>) groups, and make it work correctly.
4308
4309 3. The first character computation wasn't working for (?>) groups.
4310
4311 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4312 end of the subject) and add 5.005's \z, which really does match only at the
4313 very end of the subject.
4314
4315 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4316
4317 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4318 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4319 localized options. All options to pcre_study() were also removed.
4320
4321 7. Add other new features from 5.005:
4322
4323 $(?<= positive lookbehind
4324 $(?<! negative lookbehind
4325 (?imsx-imsx) added the unsetting capability
4326 such a setting is global if at outer level; local otherwise
4327 (?imsx-imsx:) non-capturing groups with option setting
4328 (?(cond)re|re) conditional pattern matching
4329
4330 A backreference to itself in a repeated group matches the previous
4331 captured string.
4332
4333 8. General tidying up of studying (both automatic and via "study")
4334 consequential on the addition of new assertions.
4335
4336 9. As in 5.005, unlimited repeated groups that could match an empty substring
4337 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4338 runtime if any iteration does actually match an empty substring.
4339
4340 10. Include the RunTest script in the distribution.
4341
4342 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4343 discrepancies, some of which were old and were also with respect to 5.004. They
4344 have now been fixed.
4345
4346
4347 Version 1.09 28-Apr-98
4348 ----------------------
4349
4350 1. A negated single character class followed by a quantifier with a minimum
4351 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4352 program crashes, or just wrong answers. This did not apply to negated classes
4353 containing more than one character, or to minima other than one.
4354
4355
4356 Version 1.08 27-Mar-98
4357 ----------------------
4358
4359 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4360
4361 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4362 latter must appear before anything that relies on it in the pattern.
4363
4364
4365 Version 1.07 16-Feb-98
4366 ----------------------
4367
4368 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4369 repeat of a potentially empty string).
4370
4371
4372 Version 1.06 23-Jan-98
4373 ----------------------
4374
4375 1. Added Markus Oberhumer's little patches for C++.
4376
4377 2. Literal strings longer than 255 characters were broken.
4378
4379
4380 Version 1.05 23-Dec-97
4381 ----------------------
4382
4383 1. Negated character classes containing more than one character were failing if
4384 PCRE_CASELESS was set at run time.
4385
4386
4387 Version 1.04 19-Dec-97
4388 ----------------------
4389
4390 1. Corrected the man page, where some "const" qualifiers had been omitted.
4391
4392 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4393 input syntax.
4394
4395 3. Fixed memory leak which occurred when a regex with back references was
4396 matched with an offsets vector that wasn't big enough. The temporary memory
4397 that is used in this case wasn't being freed if the match failed.
4398
4399 4. Tidied pcretest to ensure it frees memory that it gets.
4400
4401 5. Temporary memory was being obtained in the case where the passed offsets
4402 vector was exactly big enough.
4403
4404 6. Corrected definition of offsetof() from change 5 below.
4405
4406 7. I had screwed up change 6 below and broken the rules for the use of
4407 setjmp(). Now fixed.
4408
4409
4410 Version 1.03 18-Dec-97
4411 ----------------------
4412
4413 1. A erroneous regex with a missing opening parenthesis was correctly
4414 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4415 on some systems.
4416
4417 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4418 it was reported that one broken compiler failed on the former because "code" is
4419 also an independent variable.
4420
4421 3. The erroneous regex a[]b caused an array overrun reference.
4422
4423 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4424 fail on data ending with that character. (It was going on too far, and checking
4425 the next character, typically a binary zero.) This was specific to the
4426 optimized code for single-character negative classes.
4427
4428 5. Added a contributed patch from the TIN world which does the following:
4429
4430 + Add an undef for memmove, in case the the system defines a macro for it.
4431
4432 + Add a definition of offsetof(), in case there isn't one. (I don't know
4433 the reason behind this - offsetof() is part of the ANSI standard - but
4434 it does no harm).
4435
4436 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4437 most of the places where whitespace preceded '#'. I have given up and
4438 allowed the remaining 2 cases to be at the margin.
4439
4440 + Rename some variables in pcre to eliminate shadowing. This seems very
4441 pedantic, but does no harm, of course.
4442
4443 6. Moved the call to setjmp() into its own function, to get rid of warnings
4444 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4445
4446 7. Constructs such as \d{8,} were compiling into the equivalent of
4447 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4448 outcome, but in this particular case used more store than had been allocated,
4449 which caused the bug to be discovered because it threw up an internal error.
4450
4451 8. The debugging code in both pcre and pcretest for outputting the compiled
4452 form of a regex was going wrong in the case of back references followed by
4453 curly-bracketed repeats.
4454
4455
4456 Version 1.02 12-Dec-97
4457 ----------------------
4458
4459 1. Typos in pcre.3 and comments in the source fixed.
4460
4461 2. Applied a contributed patch to get rid of places where it used to remove
4462 'const' from variables, and fixed some signed/unsigned and uninitialized
4463 variable warnings.
4464
4465 3. Added the "runtest" target to Makefile.
4466
4467 4. Set default compiler flag to -O2 rather than just -O.
4468
4469
4470 Version 1.01 19-Nov-97
4471 ----------------------
4472
4473 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4474 like /([ab]*)*/, that is, for classes with more than one character in them.
4475
4476 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4477 as /((?>a*))*/ (a PCRE_EXTRA facility).
4478
4479
4480 Version 1.00 18-Nov-97
4481 ----------------------
4482
4483 1. Added compile-time macros to support systems such as SunOS4 which don't have
4484 memmove() or strerror() but have other things that can be used instead.
4485
4486 2. Arranged that "make clean" removes the executables.
4487
4488
4489 Version 0.99 27-Oct-97
4490 ----------------------
4491
4492 1. Fixed bug in code for optimizing classes with only one character. It was
4493 initializing a 32-byte map regardless, which could cause it to run off the end
4494 of the memory it had got.
4495
4496 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4497
4498
4499 Version 0.98 22-Oct-97
4500 ----------------------
4501
4502 1. Fixed bug in code for handling temporary memory usage when there are more
4503 back references than supplied space in the ovector. This could cause segfaults.
4504
4505
4506 Version 0.97 21-Oct-97
4507 ----------------------
4508
4509 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4510
4511 2. Optimized negated single characters not to use a bit map.
4512
4513 3. Brought error texts together as macro definitions; clarified some of them;
4514 fixed one that was wrong - it said "range out of order" when it meant "invalid
4515 escape sequence".
4516
4517 4. Changed some char * arguments to const char *.
4518
4519 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4520
4521 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4522 pcretest.
4523
4524
4525 Version 0.96 16-Oct-97
4526 ----------------------
4527
4528 1. Added a simple "pgrep" utility to the distribution.
4529
4530 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4531 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4532 where "ddd" means "one or more decimal digits".
4533
4534 3. Fixed serious bug. If a pattern had a back reference, but the call to
4535 pcre_exec() didn't supply a large enough ovector to record the related
4536 identifying subpattern, the match always failed. PCRE now remembers the number
4537 of the largest back reference, and gets some temporary memory in which to save
4538 the offsets during matching if necessary, in order to ensure that
4539 backreferences always work.
4540
4541 4. Increased the compatibility with Perl in a number of ways:
4542
4543 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4544 to request this handling. The option can be set at compile or exec time.
4545
4546 (b) $ matches before a terminating newline by default; an option
4547 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4548 mode). The option can be set at compile or exec time.
4549
4550 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4551 the same as Perl's. If the decimal number it represents is less than 10
4552 or there aren't that many previous left capturing parentheses, an octal
4553 escape is read. Inside a character class, it's always an octal escape,
4554 even if it is a single digit.
4555
4556 (d) An escaped but undefined alphabetic character is taken as a literal,
4557 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4558 escapes.
4559
4560 (e) {0} is now permitted. (The previous item is removed from the compiled
4561 pattern).
4562
4563 5. Changed all the names of code files so that the basic parts are no longer
4564 than 10 characters, and abolished the teeny "globals.c" file.
4565
4566 6. Changed the handling of character classes; they are now done with a 32-byte
4567 bit map always.
4568
4569 7. Added the -d and /D options to pcretest to make it possible to look at the
4570 internals of compilation without having to recompile pcre.
4571
4572
4573 Version 0.95 23-Sep-97
4574 ----------------------
4575
4576 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4577 \x20 at the start of a run of normal characters. These were being treated as
4578 real characters, instead of the source characters being re-checked.
4579
4580
4581 Version 0.94 18-Sep-97
4582 ----------------------
4583
4584 1. The functions are now thread-safe, with the caveat that the global variables
4585 containing pointers to malloc() and free() or alternative functions are the
4586 same for all threads.
4587
4588 2. Get pcre_study() to generate a bitmap of initial characters for non-
4589 anchored patterns when this is possible, and use it if passed to pcre_exec().
4590
4591
4592 Version 0.93 15-Sep-97
4593 ----------------------
4594
4595 1. /(b)|(:+)/ was computing an incorrect first character.
4596
4597 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4598 but not actually doing anything yet.
4599
4600 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4601 as Perl does (e.g. [-az] or [az-]).
4602
4603 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4604 all possible positions.
4605
4606 5. Split up into different modules to avoid including unneeded functions in a
4607 compiled binary. However, compile and exec are still in one module. The "study"
4608 function is split off.
4609
4610 6. The character tables are now in a separate module whose source is generated
4611 by an auxiliary program - but can then be edited by hand if required. There are
4612 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4613 toupper() in the code.
4614
4615 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4616 make them global. Abolish the function for setting them, as the caller can now
4617 set them directly.
4618
4619
4620 Version 0.92 11-Sep-97
4621 ----------------------
4622
4623 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4624 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4625
4626 2. Caseless matching was not working in character classes if the characters in
4627 the pattern were in upper case.
4628
4629 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4630
4631 4. Make PCRE_ANCHORED public and accept as a compile option.
4632
4633 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4634 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4635 pass them.
4636
4637 6. Give an error if bad option bits passed at compile or run time.
4638
4639 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4640 pcretest to cause it to pass that flag.
4641
4642 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4643 options, and the first character, if set.
4644
4645 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4646
4647
4648 Version 0.91 10-Sep-97
4649 ----------------------
4650
4651 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4652 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4653
4654 2. PCRE was looping on encountering an indefinitely repeated back reference to
4655 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4656 Perl does - treats the match as successful.
4657
4658 ****

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