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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 272 - (show annotations) (download)
Sat Nov 17 16:53:53 2007 UTC (6 years, 11 months ago) by ph10
File size: 143672 byte(s)
Generalize wording for (?&) error as it also applies to \k'' and other cases.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.5 12-Nov-07
5 ---------------------
6
7 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
8 values in parens when parsing an RE using the C++ wrapper."
9
10 2. Negative specials like \S did not work in character classes in UTF-8 mode.
11 Characters greater than 255 were excluded from the class instead of being
12 included.
13
14 3. The same bug as (2) above applied to negated POSIX classes such as
15 [:^space:].
16
17 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
18 defined or documented. It seems to have been a typo for PCRE_STATIC, so
19 I have changed it.
20
21 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
22 first named subpattern) and a construct such as (?&a) would reference the
23 first named subpattern whose name started with "a" (in other words, the
24 length check was missing). Both these problems are fixed. "Subpattern name
25 expected" is now given for (?&) (a zero-length name), and this patch also
26 makes it give the same error for \k'' (previously it complained that that
27 was a reference to a non-existent subpattern).
28
29 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
30 this is right because (?- can be followed by option settings as well as by
31 digits. I have, however, made the messages clearer.
32
33 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
34 than the number used in the conditional) now cause a compile-time error.
35 This is actually not compatible with Perl, which accepts such patterns, but
36 treats the conditional as always being FALSE (as PCRE used to), but it
37 seems to me that giving a diagnostic is better.
38
39
40 Version 7.4 21-Sep-07
41 ---------------------
42
43 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
44 means that a class such as [\s] counted as "explicit reference to CR or
45 LF". That isn't really right - the whole point of the change was to try to
46 help when there was an actual mention of one of the two characters. So now
47 the change happens only if \r or \n (or a literal CR or LF) character is
48 encountered.
49
50 2. The 32-bit options word was also used for 6 internal flags, but the numbers
51 of both had grown to the point where there were only 3 bits left.
52 Fortunately, there was spare space in the data structure, and so I have
53 moved the internal flags into a new 16-bit field to free up more option
54 bits.
55
56 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
57 but did not set the internal JCHANGED flag - either of these is enough to
58 control the way the "get" function works - but the PCRE_INFO_JCHANGED
59 facility is supposed to tell if (?J) was ever used, so now (?J) at the
60 start sets both bits.
61
62 4. Added options (at build time, compile time, exec time) to change \R from
63 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
64
65 5. doc/pcresyntax.html was missing from the distribution.
66
67 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
68 compatibility, even though it is no longer used.
69
70 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
71 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
72 windows.h file is present (where different names are used). [This was
73 reversed later after testing - see 16 below.]
74
75 8. Changed all #include <config.h> to #include "config.h". There were also
76 some further <pcre.h> cases that I changed to "pcre.h".
77
78 9. When pcregrep was used with the --colour option, it missed the line ending
79 sequence off the lines that it output.
80
81 10. It was pointed out to me that arrays of string pointers cause lots of
82 relocations when a shared library is dynamically loaded. A technique of
83 using a single long string with a table of offsets can drastically reduce
84 these. I have refactored PCRE in four places to do this. The result is
85 dramatic:
86
87 Originally: 290
88 After changing UCP table: 187
89 After changing error message table: 43
90 After changing table of "verbs" 36
91 After changing table of Posix names 22
92
93 Thanks to the folks working on Gregex for glib for this insight.
94
95 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
96 unicode-properties was also set.
97
98 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
99
100 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
101 checked only for CRLF.
102
103 14. Added casts to pcretest.c to avoid compiler warnings.
104
105 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
106
107 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
108 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
109 entirely. This removes changes made in 7 above.
110
111 17. The CMake files have been updated, and there is now more information about
112 building with CMake in the NON-UNIX-USE document.
113
114
115 Version 7.3 28-Aug-07
116 ---------------------
117
118 1. In the rejigging of the build system that eventually resulted in 7.1, the
119 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
120 brackets there is not right, since it causes compilers to look for an
121 installed pcre.h, not the version that is in the source that is being
122 compiled (which of course may be different). I have changed it back to:
123
124 #include "pcre.h"
125
126 I have a vague recollection that the change was concerned with compiling in
127 different directories, but in the new build system, that is taken care of
128 by the VPATH setting the Makefile.
129
130 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
131 when the subject happened to end in the byte 0x85 (e.g. if the last
132 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
133 characters but of course it shouldn't be taken as a newline when it is part
134 of another character. The bug was that, for an unlimited repeat of . in
135 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
136 characters when looking for a newline.
137
138 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
139
140 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
141 in debug output.
142
143 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
144 long printing in the pcrecpp unittest when running under MinGW.
145
146 6. ESC_K was left out of the EBCDIC table.
147
148 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
149 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
150 limit also applies to "virtual nesting" when a pattern is recursive, and in
151 this case 1000 isn't so big. I have been able to remove this limit at the
152 expense of backing off one optimization in certain circumstances. Normally,
153 when pcre_exec() would call its internal match() function recursively and
154 immediately return the result unconditionally, it uses a "tail recursion"
155 feature to save stack. However, when a subpattern that can match an empty
156 string has an unlimited repetition quantifier, it no longer makes this
157 optimization. That gives it a stack frame in which to save the data for
158 checking that an empty string has been matched. Previously this was taken
159 from the 1000-entry workspace that had been reserved. So now there is no
160 explicit limit, but more stack is used.
161
162 8. Applied Daniel's patches to solve problems with the import/export magic
163 syntax that is required for Windows, and which was going wrong for the
164 pcreposix and pcrecpp parts of the library. These were overlooked when this
165 problem was solved for the main library.
166
167 9. There were some crude static tests to avoid integer overflow when computing
168 the size of patterns that contain repeated groups with explicit upper
169 limits. As the maximum quantifier is 65535, the maximum group length was
170 set at 30,000 so that the product of these two numbers did not overflow a
171 32-bit integer. However, it turns out that people want to use groups that
172 are longer than 30,000 bytes (though not repeat them that many times).
173 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
174 made it possible to implement the integer overflow checks in a much more
175 dynamic way, which I have now done. The artificial limitation on group
176 length has been removed - we now have only the limit on the total length of
177 the compiled pattern, which depends on the LINK_SIZE setting.
178
179 10. Fixed a bug in the documentation for get/copy named substring when
180 duplicate names are permitted. If none of the named substrings are set, the
181 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
182 empty string.
183
184 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
185 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
186 because the ] is interpreted as the first data character and the
187 terminating ] is not found. PCRE has been made compatible with Perl in this
188 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
189 cause memory overwriting.
190
191 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
192 string has been matched (to stop an infinite loop). It was not recognizing
193 a conditional subpattern that could match an empty string if that
194 subpattern was within another subpattern. For example, it looped when
195 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
196 condition was not nested. This bug has been fixed.
197
198 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
199 past the start of the subject in the presence of bytes with the top bit
200 set, for example "\x8aBCD".
201
202 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
203 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
204
205 14. Optimized (?!) to (*FAIL).
206
207 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
208 This restricts code points to be within the range 0 to 0x10FFFF, excluding
209 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
210 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
211 does: it's just the validity check that is more restrictive.
212
213 16. Inserted checks for integer overflows during escape sequence (backslash)
214 processing, and also fixed erroneous offset values for syntax errors during
215 backslash processing.
216
217 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
218 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
219
220 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
221 caused an overrun.
222
223 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
224 something other than just ASCII characters) inside a group that had an
225 unlimited repeat caused a loop at compile time (while checking to see
226 whether the group could match an empty string).
227
228 20. Debugging a pattern containing \p or \P could cause a crash. For example,
229 [\P{Any}] did so. (Error in the code for printing property names.)
230
231 21. An orphan \E inside a character class could cause a crash.
232
233 22. A repeated capturing bracket such as (A)? could cause a wild memory
234 reference during compilation.
235
236 23. There are several functions in pcre_compile() that scan along a compiled
237 expression for various reasons (e.g. to see if it's fixed length for look
238 behind). There were bugs in these functions when a repeated \p or \P was
239 present in the pattern. These operators have additional parameters compared
240 with \d, etc, and these were not being taken into account when moving along
241 the compiled data. Specifically:
242
243 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
244 length.
245
246 (b) An item such as \pL+ within a repeated group could cause crashes or
247 loops.
248
249 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
250 "reference to non-existent subpattern" error.
251
252 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
253
254 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
255 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
256
257 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
258
259 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
260 character were causing crashes (broken optimization).
261
262 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
263 \p or \P) caused a compile-time loop.
264
265 28. More problems have arisen in unanchored patterns when CRLF is a valid line
266 break. For example, the unstudied pattern [\r\n]A does not match the string
267 "\r\nA" because change 7.0/46 below moves the current point on by two
268 characters after failing to match at the start. However, the pattern \nA
269 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
270 the same is true. There doesn't seem any very clean way out of this, but
271 what I have chosen to do makes the common cases work: PCRE now takes note
272 of whether there can be an explicit match for \r or \n anywhere in the
273 pattern, and if so, 7.0/46 no longer applies. As part of this change,
274 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
275 pattern has explicit CR or LF references.
276
277 29. Added (*CR) etc for changing newline setting at start of pattern.
278
279
280 Version 7.2 19-Jun-07
281 ---------------------
282
283 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
284 which is apparently normally available under Windows.
285
286 2. Re-jig the pcregrep tests with different newline settings in an attempt
287 to make them independent of the local environment's newline setting.
288
289 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
290
291 4. Some of the "internals" tests were previously cut out when the link size
292 was not 2, because the output contained actual offsets. The recent new
293 "Z" feature of pcretest means that these can be cut out, making the tests
294 usable with all link sizes.
295
296 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
297 stack recursion. This gives a massive performance boost under BSD, but just
298 a small improvement under Linux. However, it saves one field in the frame
299 in all cases.
300
301 6. Added more features from the forthcoming Perl 5.10:
302
303 (a) (?-n) (where n is a string of digits) is a relative subroutine or
304 recursion call. It refers to the nth most recently opened parentheses.
305
306 (b) (?+n) is also a relative subroutine call; it refers to the nth next
307 to be opened parentheses.
308
309 (c) Conditions that refer to capturing parentheses can be specified
310 relatively, for example, (?(-2)... or (?(+3)...
311
312 (d) \K resets the start of the current match so that everything before
313 is not part of it.
314
315 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
316
317 (f) \g{name} is another synonym - part of Perl 5.10's unification of
318 reference syntax.
319
320 (g) (?| introduces a group in which the numbering of parentheses in each
321 alternative starts with the same number.
322
323 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
324
325 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
326 PCRE_INFO_JCHANGED.
327
328 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
329 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
330 for detecting groups that can match an empty string.
331
332 9. A pattern with a very large number of alternatives (more than several
333 hundred) was running out of internal workspace during the pre-compile
334 phase, where pcre_compile() figures out how much memory will be needed. A
335 bit of new cunning has reduced the workspace needed for groups with
336 alternatives. The 1000-alternative test pattern now uses 12 bytes of
337 workspace instead of running out of the 4096 that are available.
338
339 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
340
341 11. Applied patch from Google to remove an optimization that didn't quite work.
342 The report of the bug said:
343
344 pcrecpp::RE("a*").FullMatch("aaa") matches, while
345 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
346 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
347
348 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
349 it matched the wrong number of bytes.
350
351
352 Version 7.1 24-Apr-07
353 ---------------------
354
355 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
356 that is more "standard", making use of automake and other Autotools. There
357 is some re-arrangement of the files and adjustment of comments consequent
358 on this.
359
360 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
361 for recursive directory scanning broke on some systems because the files
362 are not scanned in any specific order and on different systems the order
363 was different. A call to "sort" has been inserted into RunGrepTest for the
364 approprate test as a short-term fix. In the longer term there may be an
365 alternative.
366
367 3. I had an email from Eric Raymond about problems translating some of PCRE's
368 man pages to HTML (despite the fact that I distribute HTML pages, some
369 people do their own conversions for various reasons). The problems
370 concerned the use of low-level troff macros .br and .in. I have therefore
371 removed all such uses from the man pages (some were redundant, some could
372 be replaced by .nf/.fi pairs). The 132html script that I use to generate
373 HTML has been updated to handle .nf/.fi and to complain if it encounters
374 .br or .in.
375
376 4. Updated comments in configure.ac that get placed in config.h.in and also
377 arranged for config.h to be included in the distribution, with the name
378 config.h.generic, for the benefit of those who have to compile without
379 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
380
381 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
382 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
383 makevp.bat for the current PCRE, using the additional files
384 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
385
386 6. A Windows user reported a minor discrepancy with test 2, which turned out
387 to be caused by a trailing space on an input line that had got lost in his
388 copy. The trailing space was an accident, so I've just removed it.
389
390 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
391 that is needed.
392
393 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
394 as "const" (a) because they are and (b) because it helps the PHP
395 maintainers who have recently made a script to detect big data structures
396 in the php code that should be moved to the .rodata section. I remembered
397 to update Builducptable as well, so it won't revert if ucptable.h is ever
398 re-created.
399
400 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
401 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
402 order to be able to cut out the UTF-8 tables in the latter when UTF-8
403 support is not required. This saves 1.5-2K of code, which is important in
404 some applications.
405
406 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
407 so as not to refer to the tables, even though these functions will never be
408 called when UTF-8 support is disabled. Otherwise there are problems with a
409 shared library.
410
411 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
412
413 (a) It was defining its arguments as char * instead of void *.
414
415 (b) It was assuming that all moves were upwards in memory; this was true
416 a long time ago when I wrote it, but is no longer the case.
417
418 The emulated memove() is provided for those environments that have neither
419 memmove() nor bcopy(). I didn't think anyone used it these days, but that
420 is clearly not the case, as these two bugs were recently reported.
421
422 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
423 and Detrail to create the HTML documentation, the .txt form of the man
424 pages, and it removes trailing spaces from listed files. It also creates
425 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
426 case, it wraps all the #defines with #ifndefs. This script should be run
427 before "make dist".
428
429 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
430 with Unicode property support.
431
432 (a) For a maximizing quantifier, if the two different cases of the
433 character were of different lengths in their UTF-8 codings (there are
434 some cases like this - I found 11), and the matching function had to
435 back up over a mixture of the two cases, it incorrectly assumed they
436 were both the same length.
437
438 (b) When PCRE was configured to use the heap rather than the stack for
439 recursion during matching, it was not correctly preserving the data for
440 the other case of a UTF-8 character when checking ahead for a match
441 while processing a minimizing repeat. If the check also involved
442 matching a wide character, but failed, corruption could cause an
443 erroneous result when trying to check for a repeat of the original
444 character.
445
446 13. Some tidying changes to the testing mechanism:
447
448 (a) The RunTest script now detects the internal link size and whether there
449 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
450 values substituted by "configure". (The RunGrepTest script already did
451 this for UTF-8.) The configure.ac script no longer substitutes the
452 relevant variables.
453
454 (b) The debugging options /B and /D in pcretest show the compiled bytecode
455 with length and offset values. This means that the output is different
456 for different internal link sizes. Test 2 is skipped for link sizes
457 other than 2 because of this, bypassing the problem. Unfortunately,
458 there was also a test in test 3 (the locale tests) that used /B and
459 failed for link sizes other than 2. Rather than cut the whole test out,
460 I have added a new /Z option to pcretest that replaces the length and
461 offset values with spaces. This is now used to make test 3 independent
462 of link size. (Test 2 will be tidied up later.)
463
464 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
465 segmentation fault instead of returning the appropriate error message.
466
467 15. In multiline mode when the newline sequence was set to "any", the pattern
468 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
469 This doesn't seem right; it now treats the CRLF combination as the line
470 ending, and so does not match in that case. It's only a pattern such as ^$
471 that would hit this one: something like ^ABC$ would have failed after \r
472 and then tried again after \r\n.
473
474 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
475 in an attempt to make files that differ only in their line terminators
476 compare equal. This works on Linux.
477
478 17. Under certain error circumstances pcregrep might try to free random memory
479 as it exited. This is now fixed, thanks to valgrind.
480
481 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
482 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
483 was because its rules for how to advance for /g after matching an empty
484 string at the end of a line did not allow for this case. They now check for
485 it specially.
486
487 20. pcretest is supposed to handle patterns and data of any length, by
488 extending its buffers when necessary. It was getting this wrong when the
489 buffer for a data line had to be extended.
490
491 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
492 CRLF as a newline sequence.
493
494 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
495 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
496 I have nevertheless tidied it up.
497
498 23. Added some casts to kill warnings from HP-UX ia64 compiler.
499
500 24. Added a man page for pcre-config.
501
502
503 Version 7.0 19-Dec-06
504 ---------------------
505
506 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
507 moving to gcc 4.1.1.
508
509 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
510 sys/time.h, which is documented as needed for this function. It doesn't
511 seem to matter on Linux, but it showed up on some releases of OS X.
512
513 3. It seems that there are systems where bytes whose values are greater than
514 127 match isprint() in the "C" locale. The "C" locale should be the
515 default when a C program starts up. In most systems, only ASCII printing
516 characters match isprint(). This difference caused the output from pcretest
517 to vary, making some of the tests fail. I have changed pcretest so that:
518
519 (a) When it is outputting text in the compiled version of a pattern, bytes
520 other than 32-126 are always shown as hex escapes.
521
522 (b) When it is outputting text that is a matched part of a subject string,
523 it does the same, unless a different locale has been set for the match
524 (using the /L modifier). In this case, it uses isprint() to decide.
525
526 4. Fixed a major bug that caused incorrect computation of the amount of memory
527 required for a compiled pattern when options that changed within the
528 pattern affected the logic of the preliminary scan that determines the
529 length. The relevant options are -x, and -i in UTF-8 mode. The result was
530 that the computed length was too small. The symptoms of this bug were
531 either the PCRE error "internal error: code overflow" from pcre_compile(),
532 or a glibc crash with a message such as "pcretest: free(): invalid next
533 size (fast)". Examples of patterns that provoked this bug (shown in
534 pcretest format) are:
535
536 /(?-x: )/x
537 /(?x)(?-x: \s*#\s*)/
538 /((?i)[\x{c0}])/8
539 /(?i:[\x{c0}])/8
540
541 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
542 is now done differently.
543
544 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
545 wrapper classes; (b) implement a new function in the C++ scanner that is
546 more efficient than the old way of doing things because it avoids levels of
547 recursion in the regex matching; (c) add a paragraph to the documentation
548 for the FullMatch() function.
549
550 6. The escape sequence \n was being treated as whatever was defined as
551 "newline". Not only was this contrary to the documentation, which states
552 that \n is character 10 (hex 0A), but it also went horribly wrong when
553 "newline" was defined as CRLF. This has been fixed.
554
555 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
556 was being set to -1 for the "end of line" case (supposedly a value that no
557 character can have). Though this value is never used (the check for end of
558 line is "zero bytes in current character"), it caused compiler complaints.
559 I've changed it to 0xffffffff.
560
561 8. In pcre_version.c, the version string was being built by a sequence of
562 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
563 string (as it is for production releases) called a macro with an empty
564 argument. The C standard says the result of this is undefined. The gcc
565 compiler treats it as an empty string (which was what was wanted) but it is
566 reported that Visual C gives an error. The source has been hacked around to
567 avoid this problem.
568
569 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
570 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
571 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
572 of them did).
573
574 10. Originally, pcretest opened its input and output without "b"; then I was
575 told that "b" was needed in some environments, so it was added for release
576 5.0 to both the input and output. (It makes no difference on Unix-like
577 systems.) Later I was told that it is wrong for the input on Windows. I've
578 now abstracted the modes into two macros, to make it easier to fiddle with
579 them, and removed "b" from the input mode under Windows.
580
581 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
582
583 12. Added -help and --help to pcretest as an official way of being reminded
584 of the options.
585
586 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
587 and pcrecpp.cc because they annoy compilers at high warning levels.
588
589 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
590
591 15. Fixed an occurrence of == in configure.ac that should have been = (shell
592 scripts are not C programs :-) and which was not noticed because it works
593 on Linux.
594
595 16. pcretest is supposed to handle any length of pattern and data line (as one
596 line or as a continued sequence of lines) by extending its input buffer if
597 necessary. This feature was broken for very long pattern lines, leading to
598 a string of junk being passed to pcre_compile() if the pattern was longer
599 than about 50K.
600
601 17. I have done a major re-factoring of the way pcre_compile() computes the
602 amount of memory needed for a compiled pattern. Previously, there was code
603 that made a preliminary scan of the pattern in order to do this. That was
604 OK when PCRE was new, but as the facilities have expanded, it has become
605 harder and harder to keep it in step with the real compile phase, and there
606 have been a number of bugs (see for example, 4 above). I have now found a
607 cunning way of running the real compile function in a "fake" mode that
608 enables it to compute how much memory it would need, while actually only
609 ever using a few hundred bytes of working memory and without too many
610 tests of the mode. This should make future maintenance and development
611 easier. A side effect of this work is that the limit of 200 on the nesting
612 depth of parentheses has been removed (though this was never a serious
613 limitation, I suspect). However, there is a downside: pcre_compile() now
614 runs more slowly than before (30% or more, depending on the pattern). I
615 hope this isn't a big issue. There is no effect on runtime performance.
616
617 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
618 newline (only possible for the last line of a file) and it was a
619 pattern that set a locale (followed by /Lsomething), pcretest crashed.
620
621 19. Added additional timing features to pcretest. (1) The -tm option now times
622 matching only, not compiling. (2) Both -t and -tm can be followed, as a
623 separate command line item, by a number that specifies the number of
624 repeats to use when timing. The default is 50000; this gives better
625 precision, but takes uncomfortably long for very large patterns.
626
627 20. Extended pcre_study() to be more clever in cases where a branch of a
628 subpattern has no definite first character. For example, (a*|b*)[cd] would
629 previously give no result from pcre_study(). Now it recognizes that the
630 first character must be a, b, c, or d.
631
632 21. There was an incorrect error "recursive call could loop indefinitely" if
633 a subpattern (or the entire pattern) that was being tested for matching an
634 empty string contained only one non-empty item after a nested subpattern.
635 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
636 incorrectly, because the \d was being skipped in the check.
637
638 22. The pcretest program now has a new pattern option /B and a command line
639 option -b, which is equivalent to adding /B to every pattern. This causes
640 it to show the compiled bytecode, without the additional information that
641 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
642 is the same as /B/I).
643
644 23. A new optimization is now able automatically to treat some sequences such
645 as a*b as a*+b. More specifically, if something simple (such as a character
646 or a simple class like \d) has an unlimited quantifier, and is followed by
647 something that cannot possibly match the quantified thing, the quantifier
648 is automatically "possessified".
649
650 24. A recursive reference to a subpattern whose number was greater than 39
651 went wrong under certain circumstances in UTF-8 mode. This bug could also
652 have affected the operation of pcre_study().
653
654 25. Realized that a little bit of performance could be had by replacing
655 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
656
657 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
658
659 27. Possessive quantifiers such as a++ were previously implemented by turning
660 them into atomic groups such as ($>a+). Now they have their own opcodes,
661 which improves performance. This includes the automatically created ones
662 from 23 above.
663
664 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
665 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
666 the first matched character to be a colon. This applied both to named and
667 numbered groups.
668
669 29. The ucpinternal.h header file was missing its idempotency #ifdef.
670
671 30. I was sent a "project" file called libpcre.a.dev which I understand makes
672 building PCRE on Windows easier, so I have included it in the distribution.
673
674 31. There is now a check in pcretest against a ridiculously large number being
675 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
676 loop, the loop is abandoned.
677
678 32. Forward references to subpatterns in conditions such as (?(2)...) where
679 subpattern 2 is defined later cause pcre_compile() to search forwards in
680 the pattern for the relevant set of parentheses. This search went wrong
681 when there were unescaped parentheses in a character class, parentheses
682 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
683
684 33. "Subroutine" calls and backreferences were previously restricted to
685 referencing subpatterns earlier in the regex. This restriction has now
686 been removed.
687
688 34. Added a number of extra features that are going to be in Perl 5.10. On the
689 whole, these are just syntactic alternatives for features that PCRE had
690 previously implemented using the Python syntax or my own invention. The
691 other formats are all retained for compatibility.
692
693 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
694 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
695 also .NET compatible.
696
697 (b) A recursion or subroutine call to a named group can now be defined as
698 (?&name) as well as (?P>name).
699
700 (c) A backreference to a named group can now be defined as \k<name> or
701 \k'name' as well as (?P=name). The new forms, as well as being in Perl
702 5.10, are also .NET compatible.
703
704 (d) A conditional reference to a named group can now use the syntax
705 (?(<name>) or (?('name') as well as (?(name).
706
707 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
708 groups (named and numbered) that are never evaluated inline, but can be
709 called as "subroutines" from elsewhere. In effect, the DEFINE condition
710 is always false. There may be only one alternative in such a group.
711
712 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
713 as the simple (?(R). The condition is true only if the most recent
714 recursion is that of the given number or name. It does not search out
715 through the entire recursion stack.
716
717 (g) The escape \gN or \g{N} has been added, where N is a positive or
718 negative number, specifying an absolute or relative reference.
719
720 35. Tidied to get rid of some further signed/unsigned compiler warnings and
721 some "unreachable code" warnings.
722
723 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
724 things, this adds five new scripts.
725
726 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
727 There were also incompatibilities regarding the handling of \Q..\E inside
728 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
729 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
730
731 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
732 matches an empty string, and forcibly breaks the loop. There were bugs in
733 this code in non-simple cases. For a pattern such as ^(a()*)* matched
734 against aaaa the result was just "a" rather than "aaaa", for example. Two
735 separate and independent bugs (that affected different cases) have been
736 fixed.
737
738 39. Refactored the code to abolish the use of different opcodes for small
739 capturing bracket numbers. This is a tidy that I avoided doing when I
740 removed the limit on the number of capturing brackets for 3.5 back in 2001.
741 The new approach is not only tidier, it makes it possible to reduce the
742 memory needed to fix the previous bug (38).
743
744 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
745 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
746 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
747 mode.
748
749 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
750 report.
751
752 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
753 copy construction and assignment in the C++ wrapper.
754
755 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
756 couple of bugs that could have given wrong results in the "--newline=crlf"
757 case.
758
759 44. Added a number of casts and did some reorganization of signed/unsigned int
760 variables following suggestions from Dair Grant. Also renamed the variable
761 "this" as "item" because it is a C++ keyword.
762
763 45. Arranged for dftables to add
764
765 #include "pcre_internal.h"
766
767 to pcre_chartables.c because without it, gcc 4.x may remove the array
768 definition from the final binary if PCRE is built into a static library and
769 dead code stripping is activated.
770
771 46. For an unanchored pattern, if a match attempt fails at the start of a
772 newline sequence, and the newline setting is CRLF or ANY, and the next two
773 characters are CRLF, advance by two characters instead of one.
774
775
776 Version 6.7 04-Jul-06
777 ---------------------
778
779 1. In order to handle tests when input lines are enormously long, pcretest has
780 been re-factored so that it automatically extends its buffers when
781 necessary. The code is crude, but this _is_ just a test program. The
782 default size has been increased from 32K to 50K.
783
784 2. The code in pcre_study() was using the value of the re argument before
785 testing it for NULL. (Of course, in any sensible call of the function, it
786 won't be NULL.)
787
788 3. The memmove() emulation function in pcre_internal.h, which is used on
789 systems that lack both memmove() and bcopy() - that is, hardly ever -
790 was missing a "static" storage class specifier.
791
792 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
793 containing an extended class (one that cannot be represented by a bitmap
794 because it contains high-valued characters or Unicode property items, e.g.
795 [\pZ]). Almost always one would set UTF-8 mode when processing such a
796 pattern, but PCRE should not loop if you do not (it no longer does).
797 [Detail: two cases were found: (a) a repeated subpattern containing an
798 extended class; (b) a recursive reference to a subpattern that followed a
799 previous extended class. It wasn't skipping over the extended class
800 correctly when UTF-8 mode was not set.]
801
802 5. A negated single-character class was not being recognized as fixed-length
803 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
804 compile error "lookbehind assertion is not fixed length".
805
806 6. The RunPerlTest auxiliary script was showing an unexpected difference
807 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
808 write a Perl script that can interpret lines of an input file either as
809 byte characters or as UTF-8, which is what "perltest" was being required to
810 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
811 can't do is switch easily at run time between having the "use utf8;" pragma
812 or not. In the end, I fudged it by using the RunPerlTest script to insert
813 "use utf8;" explicitly for the UTF-8 tests.
814
815 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
816 the end of the subject string, contrary to the documentation and to what
817 Perl does. This was true of both matching functions. Now it matches only at
818 the start of the subject and immediately after *internal* newlines.
819
820 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
821 a pointer to an int instead of a pointer to an unsigned long int. This
822 caused problems on 64-bit systems.
823
824 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
825 instance of the 'standard' template library not being so standard".
826
827 10. There was no check on the number of named subpatterns nor the maximum
828 length of a subpattern name. The product of these values is used to compute
829 the size of the memory block for a compiled pattern. By supplying a very
830 long subpattern name and a large number of named subpatterns, the size
831 computation could be caused to overflow. This is now prevented by limiting
832 the length of names to 32 characters, and the number of named subpatterns
833 to 10,000.
834
835 11. Subpatterns that are repeated with specific counts have to be replicated in
836 the compiled pattern. The size of memory for this was computed from the
837 length of the subpattern and the repeat count. The latter is limited to
838 65535, but there was no limit on the former, meaning that integer overflow
839 could in principle occur. The compiled length of a repeated subpattern is
840 now limited to 30,000 bytes in order to prevent this.
841
842 12. Added the optional facility to have named substrings with the same name.
843
844 13. Added the ability to use a named substring as a condition, using the
845 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
846 are numbers (not recommended). Forward references are permitted.
847
848 14. Added forward references in named backreferences (if you see what I mean).
849
850 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
851 pattern could run off the end of the subject. For example, the pattern
852 "(?s)(.{1,5})"8 did this with the subject "ab".
853
854 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
855 PCRE_CASELESS was set when matching characters that were quantified with ?
856 or *.
857
858 17. A character class other than a single negated character that had a minimum
859 but no maximum quantifier - for example [ab]{6,} - was not handled
860 correctly by pce_dfa_exec(). It would match only one character.
861
862 18. A valid (though odd) pattern that looked like a POSIX character
863 class but used an invalid character after [ (for example [[,abc,]]) caused
864 pcre_compile() to give the error "Failed: internal error: code overflow" or
865 in some cases to crash with a glibc free() error. This could even happen if
866 the pattern terminated after [[ but there just happened to be a sequence of
867 letters, a binary zero, and a closing ] in the memory that followed.
868
869 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
870 over the years. Originally (before any Unicode support), just the bottom 8
871 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
872 output from "man perlunicode" includes this:
873
874 The regular expression compiler produces polymorphic opcodes. That
875 is, the pattern adapts to the data and automatically switches to
876 the Unicode character scheme when presented with Unicode data--or
877 instead uses a traditional byte scheme when presented with byte
878 data.
879
880 Sadly, a wide octal escape does not cause a switch, and in a string with
881 no other multibyte characters, these octal escapes are treated as before.
882 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
883 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
884 Unicode string.
885
886 I have not perpetrated such confusion in PCRE. Up till now, it took just
887 the bottom 8 bits, as in old Perl. I have now made octal escapes with
888 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
889 translate to the appropriate multibyte character.
890
891 29. Applied some refactoring to reduce the number of warnings from Microsoft
892 and Borland compilers. This has included removing the fudge introduced
893 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
894 a warning about an unused variable.
895
896 21. PCRE has not included VT (character 0x0b) in the set of whitespace
897 characters since release 4.0, because Perl (from release 5.004) does not.
898 [Or at least, is documented not to: some releases seem to be in conflict
899 with the documentation.] However, when a pattern was studied with
900 pcre_study() and all its branches started with \s, PCRE still included VT
901 as a possible starting character. Of course, this did no harm; it just
902 caused an unnecessary match attempt.
903
904 22. Removed a now-redundant internal flag bit that recorded the fact that case
905 dependency changed within the pattern. This was once needed for "required
906 byte" processing, but is no longer used. This recovers a now-scarce options
907 bit. Also moved the least significant internal flag bit to the most-
908 significant bit of the word, which was not previously used (hangover from
909 the days when it was an int rather than a uint) to free up another bit for
910 the future.
911
912 23. Added support for CRLF line endings as well as CR and LF. As well as the
913 default being selectable at build time, it can now be changed at runtime
914 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
915 specify that it is scanning data with non-default line endings.
916
917 24. Changed the definition of CXXLINK to make it agree with the definition of
918 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
919
920 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
921 recursions. This makes a big different to stack usage for some patterns.
922
923 26. If a subpattern containing a named recursion or subroutine reference such
924 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
925 the space required for the compiled pattern went wrong and gave too small a
926 value. Depending on the environment, this could lead to "Failed: internal
927 error: code overflow at offset 49" or "glibc detected double free or
928 corruption" errors.
929
930 27. Applied patches from Google (a) to support the new newline modes and (b) to
931 advance over multibyte UTF-8 characters in GlobalReplace.
932
933 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
934 difference for some implementation of PCRE in some Windows version.
935
936 29. Added some extra testing facilities to pcretest:
937
938 \q<number> in a data line sets the "match limit" value
939 \Q<number> in a data line sets the "match recursion limt" value
940 -S <number> sets the stack size, where <number> is in megabytes
941
942 The -S option isn't available for Windows.
943
944
945 Version 6.6 06-Feb-06
946 ---------------------
947
948 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
949 in pcreposix.h. I have copied the definition from pcre.h.
950
951 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
952 because pcre.h is no longer a built file.
953
954 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
955 not normally included in the compiled code.
956
957
958 Version 6.5 01-Feb-06
959 ---------------------
960
961 1. When using the partial match feature with pcre_dfa_exec(), it was not
962 anchoring the second and subsequent partial matches at the new starting
963 point. This could lead to incorrect results. For example, with the pattern
964 /1234/, partially matching against "123" and then "a4" gave a match.
965
966 2. Changes to pcregrep:
967
968 (a) All non-match returns from pcre_exec() were being treated as failures
969 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
970 error message is output. Some extra information is given for the
971 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
972 probably the only errors that are likely to be caused by users (by
973 specifying a regex that has nested indefinite repeats, for instance).
974 If there are more than 20 of these errors, pcregrep is abandoned.
975
976 (b) A binary zero was treated as data while matching, but terminated the
977 output line if it was written out. This has been fixed: binary zeroes
978 are now no different to any other data bytes.
979
980 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
981 used to set a locale for matching. The --locale=xxxx long option has
982 been added (no short equivalent) to specify a locale explicitly on the
983 pcregrep command, overriding the environment variables.
984
985 (d) When -B was used with -n, some line numbers in the output were one less
986 than they should have been.
987
988 (e) Added the -o (--only-matching) option.
989
990 (f) If -A or -C was used with -c (count only), some lines of context were
991 accidentally printed for the final match.
992
993 (g) Added the -H (--with-filename) option.
994
995 (h) The combination of options -rh failed to suppress file names for files
996 that were found from directory arguments.
997
998 (i) Added the -D (--devices) and -d (--directories) options.
999
1000 (j) Added the -F (--fixed-strings) option.
1001
1002 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1003
1004 (l) Added the --colo(u)r option.
1005
1006 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1007 is not present by default.
1008
1009 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1010 items such as (?R) or (?1), when the recursion could match a number of
1011 alternatives. If it matched one of the alternatives, but subsequently,
1012 outside the recursion, there was a failure, the code tried to back up into
1013 the recursion. However, because of the way PCRE is implemented, this is not
1014 possible, and the result was an incorrect result from the match.
1015
1016 In order to prevent this happening, the specification of recursion has
1017 been changed so that all such subpatterns are automatically treated as
1018 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1019
1020 4. I had overlooked the fact that, in some locales, there are characters for
1021 which isalpha() is true but neither isupper() nor islower() are true. In
1022 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1023 and ordfeminine) are like this. This affected the treatment of \w and \W
1024 when they appeared in character classes, but not when they appeared outside
1025 a character class. The bit map for "word" characters is now created
1026 separately from the results of isalnum() instead of just taking it from the
1027 upper, lower, and digit maps. (Plus the underscore character, of course.)
1028
1029 5. The above bug also affected the handling of POSIX character classes such as
1030 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1031 permanent tables. Instead, the bit maps for such a class were previously
1032 created as the appropriate unions of the upper, lower, and digit bitmaps.
1033 Now they are created by subtraction from the [[:word:]] class, which has
1034 its own bitmap.
1035
1036 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1037 It is created by subtracting the vertical space characters (\x09, \x0a,
1038 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1039 subtraction was done in the overall bitmap for a character class, meaning
1040 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1041 be recognized. This bug has been fixed.
1042
1043 7. Patches from the folks at Google:
1044
1045 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1046 real life, but is still worth protecting against".
1047
1048 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1049 regular expressions".
1050
1051 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1052 have it.
1053
1054 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1055 "configure" and the latter not, in order to fix a problem somebody had
1056 with compiling the Arg class on HP-UX.
1057
1058 (e) Improve the error-handling of the C++ wrapper a little bit.
1059
1060 (f) New tests for checking recursion limiting.
1061
1062 8. The pcre_memmove() function, which is used only if the environment does not
1063 have a standard memmove() function (and is therefore rarely compiled),
1064 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1065 returning a result (though PCRE never actually uses the result).
1066
1067 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1068 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1069 returned instead of calling malloc() with an overflowing number that would
1070 most likely cause subsequent chaos.
1071
1072 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1073
1074 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1075 with this option is matched, the nmatch and pmatch options of regexec() are
1076 ignored.
1077
1078 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1079 provided in case anyone wants to the the POSIX interface with UTF-8
1080 strings.
1081
1082 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1083 C++ linking (needed for some HP-UX environments).
1084
1085 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1086 (unused parameter) and in the pcre_printint() function (omitted "default"
1087 switch label when the default is to do nothing).
1088
1089 15. Added some code to make it possible, when PCRE is compiled as a C++
1090 library, to replace subject pointers for pcre_exec() with a smart pointer
1091 class, thus making it possible to process discontinuous strings.
1092
1093 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1094 much the same function. They were added by different people who were trying
1095 to make PCRE easy to compile on non-Unix systems. It has been suggested
1096 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1097 for compiling on Windows systems. I have therefore replaced it with
1098 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1099 defaults to "extern" for C or "extern C" for C++, which works fine on
1100 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1101 SCOPE with something explicit in config.h. In addition:
1102
1103 (a) pcreposix.h still had just "extern" instead of either of these macros;
1104 I have replaced it with PCRE_DATA_SCOPE.
1105
1106 (b) Functions such as _pcre_xclass(), which are internal to the library,
1107 but external in the C sense, all had PCRE_EXPORT in their definitions.
1108 This is apparently wrong for the Windows case, so I have removed it.
1109 (It makes no difference on Unix-like systems.)
1110
1111 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1112 of recursive calls to match(). This is different to MATCH_LIMIT because
1113 that limits the total number of calls to match(), not all of which increase
1114 the depth of recursion. Limiting the recursion depth limits the amount of
1115 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1116 when PCRE is compiled, and changed at run time. A patch from Google adds
1117 this functionality to the C++ interface.
1118
1119 18. Changes to the handling of Unicode character properties:
1120
1121 (a) Updated the table to Unicode 4.1.0.
1122
1123 (b) Recognize characters that are not in the table as "Cn" (undefined).
1124
1125 (c) I revised the way the table is implemented to a much improved format
1126 which includes recognition of ranges. It now supports the ranges that
1127 are defined in UnicodeData.txt, and it also amalgamates other
1128 characters into ranges. This has reduced the number of entries in the
1129 table from around 16,000 to around 3,000, thus reducing its size
1130 considerably. I realized I did not need to use a tree structure after
1131 all - a binary chop search is just as efficient. Having reduced the
1132 number of entries, I extended their size from 6 bytes to 8 bytes to
1133 allow for more data.
1134
1135 (d) Added support for Unicode script names via properties such as \p{Han}.
1136
1137 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1138 matching that character.
1139
1140 20. When matching a repeated Unicode property with a minimum greater than zero,
1141 (for example \pL{2,}), PCRE could look past the end of the subject if it
1142 reached it while seeking the minimum number of characters. This could
1143 happen only if some of the characters were more than one byte long, because
1144 there is a check for at least the minimum number of bytes.
1145
1146 21. Refactored the implementation of \p and \P so as to be more general, to
1147 allow for more different types of property in future. This has changed the
1148 compiled form incompatibly. Anybody with saved compiled patterns that use
1149 \p or \P will have to recompile them.
1150
1151 22. Added "Any" and "L&" to the supported property types.
1152
1153 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1154 but give a compile time error if the value is greater than 0xff.
1155
1156 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1157 accidentally not being installed or uninstalled.
1158
1159 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1160 made were to insert the current release number. This seemed silly, because
1161 it made things harder for people building PCRE on systems that don't run
1162 "configure". I have turned pcre.h into a distributed file, no longer built
1163 by "configure", with the version identification directly included. There is
1164 no longer a pcre.h.in file.
1165
1166 However, this change necessitated a change to the pcre-config script as
1167 well. It is built from pcre-config.in, and one of the substitutions was the
1168 release number. I have updated configure.ac so that ./configure now finds
1169 the release number by grepping pcre.h.
1170
1171 26. Added the ability to run the tests under valgrind.
1172
1173
1174 Version 6.4 05-Sep-05
1175 ---------------------
1176
1177 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1178 "--" to be printed when multiple files were scanned, even when none of the
1179 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1180 consider it to be a bug, and have restored the previous behaviour.
1181
1182 2. A couple of code tidies to get rid of compiler warnings.
1183
1184 3. The pcretest program used to cheat by referring to symbols in the library
1185 whose names begin with _pcre_. These are internal symbols that are not
1186 really supposed to be visible externally, and in some environments it is
1187 possible to suppress them. The cheating is now confined to including
1188 certain files from the library's source, which is a bit cleaner.
1189
1190 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1191 file's purpose clearer.
1192
1193 5. Reorganized pcre_ucp_findchar().
1194
1195
1196 Version 6.3 15-Aug-05
1197 ---------------------
1198
1199 1. The file libpcre.pc.in did not have general read permission in the tarball.
1200
1201 2. There were some problems when building without C++ support:
1202
1203 (a) If C++ support was not built, "make install" and "make test" still
1204 tried to test it.
1205
1206 (b) There were problems when the value of CXX was explicitly set. Some
1207 changes have been made to try to fix these, and ...
1208
1209 (c) --disable-cpp can now be used to explicitly disable C++ support.
1210
1211 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1212 backslash in a target when C++ was disabled. This confuses some
1213 versions of "make", apparently. Using an intermediate variable solves
1214 this. (Same for CPP_LOBJ.)
1215
1216 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1217 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1218 necessary on certain architectures.
1219
1220 4. Added a setting of -export-symbols-regex to the link command to remove
1221 those symbols that are exported in the C sense, but actually are local
1222 within the library, and not documented. Their names all begin with
1223 "_pcre_". This is not a perfect job, because (a) we have to except some
1224 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1225 available (and never for static libraries). I have made a note to try to
1226 find a way round (a) in the future.
1227
1228
1229 Version 6.2 01-Aug-05
1230 ---------------------
1231
1232 1. There was no test for integer overflow of quantifier values. A construction
1233 such as {1111111111111111} would give undefined results. What is worse, if
1234 a minimum quantifier for a parenthesized subpattern overflowed and became
1235 negative, the calculation of the memory size went wrong. This could have
1236 led to memory overwriting.
1237
1238 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1239
1240 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1241 operating environments where this matters.
1242
1243 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1244 PCRE options from within the C++ wrapper.
1245
1246 5. Named capturing subpatterns were not being correctly counted when a pattern
1247 was compiled. This caused two problems: (a) If there were more than 100
1248 such subpatterns, the calculation of the memory needed for the whole
1249 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1250 back references of the form \12, where the number was greater than 9, were
1251 not recognized as back references, even though there were sufficient
1252 previous subpatterns.
1253
1254 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1255 versions of gcc, e.g. 2.95.4.
1256
1257
1258 Version 6.1 21-Jun-05
1259 ---------------------
1260
1261 1. There was one reference to the variable "posix" in pcretest.c that was not
1262 surrounded by "#if !defined NOPOSIX".
1263
1264 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1265 the cross-check on the old pcre_info() function, for the benefit of the
1266 cut-down version of PCRE that is currently imported into Exim.
1267
1268 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1269 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1270 patterns that start (?i) but I don't think that matters. The use of (?i) is
1271 just an example; this all applies to the other options as well.
1272
1273 4. Since libtool seems to echo the compile commands it is issuing, the output
1274 from "make" can be reduced a bit by putting "@" in front of each libtool
1275 compile command.
1276
1277 5. Patch from the folks at Google for configure.in to be a bit more thorough
1278 in checking for a suitable C++ installation before trying to compile the
1279 C++ stuff. This should fix a reported problem when a compiler was present,
1280 but no suitable headers.
1281
1282 6. The man pages all had just "PCRE" as their title. I have changed them to
1283 be the relevant file name. I have also arranged that these names are
1284 retained in the file doc/pcre.txt, which is a concatenation in text format
1285 of all the man pages except the little individual ones for each function.
1286
1287 7. The NON-UNIX-USE file had not been updated for the different set of source
1288 files that come with release 6. I also added a few comments about the C++
1289 wrapper.
1290
1291
1292 Version 6.0 07-Jun-05
1293 ---------------------
1294
1295 1. Some minor internal re-organization to help with my DFA experiments.
1296
1297 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1298 didn't matter for the library itself when fully configured, but did matter
1299 when compiling without UCP support, or within Exim, where the ucp files are
1300 not imported.
1301
1302 3. Refactoring of the library code to split up the various functions into
1303 different source modules. The addition of the new DFA matching code (see
1304 below) to a single monolithic source would have made it really too
1305 unwieldy, quite apart from causing all the code to be include in a
1306 statically linked application, when only some functions are used. This is
1307 relevant even without the DFA addition now that patterns can be compiled in
1308 one application and matched in another.
1309
1310 The downside of splitting up is that there have to be some external
1311 functions and data tables that are used internally in different modules of
1312 the library but which are not part of the API. These have all had their
1313 names changed to start with "_pcre_" so that they are unlikely to clash
1314 with other external names.
1315
1316 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1317 a different (DFA) algorithm. Although it is slower than the original
1318 function, it does have some advantages for certain types of matching
1319 problem.
1320
1321 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1322 including restarting after a partial match.
1323
1324 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1325 defined when compiling for Windows was sent to me. I have put it into the
1326 code, though I have no means of testing or verifying it.
1327
1328 7. Added the pcre_refcount() auxiliary function.
1329
1330 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1331 match before or at the first newline in the subject string. In pcretest,
1332 the /f option on a pattern can be used to set this.
1333
1334 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1335 would behave wrongly. This has been present in PCRE since release 4.0.
1336
1337 10. A number of changes to the pcregrep command:
1338
1339 (a) Refactored how -x works; insert ^(...)$ instead of setting
1340 PCRE_ANCHORED and checking the length, in preparation for adding
1341 something similar for -w.
1342
1343 (b) Added the -w (match as a word) option.
1344
1345 (c) Refactored the way lines are read and buffered so as to have more
1346 than one at a time available.
1347
1348 (d) Implemented a pcregrep test script.
1349
1350 (e) Added the -M (multiline match) option. This allows patterns to match
1351 over several lines of the subject. The buffering ensures that at least
1352 8K, or the rest of the document (whichever is the shorter) is available
1353 for matching (and similarly the previous 8K for lookbehind assertions).
1354
1355 (f) Changed the --help output so that it now says
1356
1357 -w, --word-regex(p)
1358
1359 instead of two lines, one with "regex" and the other with "regexp"
1360 because that confused at least one person since the short forms are the
1361 same. (This required a bit of code, as the output is generated
1362 automatically from a table. It wasn't just a text change.)
1363
1364 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1365 option but starts with a hyphen. Could be a pattern or a path name
1366 starting with a hyphen, for instance.
1367
1368 (h) "-" can be given as a file name to represent stdin.
1369
1370 (i) When file names are being printed, "(standard input)" is used for
1371 the standard input, for compatibility with GNU grep. Previously
1372 "<stdin>" was used.
1373
1374 (j) The option --label=xxx can be used to supply a name to be used for
1375 stdin when file names are being printed. There is no short form.
1376
1377 (k) Re-factored the options decoding logic because we are going to add
1378 two more options that take data. Such options can now be given in four
1379 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1380
1381 (l) Added the -A, -B, and -C options for requesting that lines of context
1382 around matches be printed.
1383
1384 (m) Added the -L option to print the names of files that do not contain
1385 any matching lines, that is, the complement of -l.
1386
1387 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1388 continue to scan other files.
1389
1390 (o) The -s option was incorrectly implemented. For compatibility with other
1391 greps, it now suppresses the error message for a non-existent or non-
1392 accessible file (but not the return code). There is a new option called
1393 -q that suppresses the output of matching lines, which was what -s was
1394 previously doing.
1395
1396 (p) Added --include and --exclude options to specify files for inclusion
1397 and exclusion when recursing.
1398
1399 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1400 Hopefully, it now does.
1401
1402 12. Missing cast in pcre_study().
1403
1404 13. Added an "uninstall" target to the makefile.
1405
1406 14. Replaced "extern" in the function prototypes in Makefile.in with
1407 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1408 world, but is set differently for Windows.
1409
1410 15. Added a second compiling function called pcre_compile2(). The only
1411 difference is that it has an extra argument, which is a pointer to an
1412 integer error code. When there is a compile-time failure, this is set
1413 non-zero, in addition to the error test pointer being set to point to an
1414 error message. The new argument may be NULL if no error number is required
1415 (but then you may as well call pcre_compile(), which is now just a
1416 wrapper). This facility is provided because some applications need a
1417 numeric error indication, but it has also enabled me to tidy up the way
1418 compile-time errors are handled in the POSIX wrapper.
1419
1420 16. Added VPATH=.libs to the makefile; this should help when building with one
1421 prefix path and installing with another. (Or so I'm told by someone who
1422 knows more about this stuff than I do.)
1423
1424 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1425 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1426 match everything, including newlines. This is not POSIX-compatible, but
1427 somebody wanted the feature. From pcretest it can be activated by using
1428 both the P and the s flags.
1429
1430 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1431
1432 19. libpcre.pc was being incorrectly installed as executable.
1433
1434 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1435 it now also looks for '\r' so that it will work unmodified on Windows.
1436
1437 21. Added Google's contributed C++ wrapper to the distribution.
1438
1439 22. Added some untidy missing memory free() calls in pcretest, to keep
1440 Electric Fence happy when testing.
1441
1442
1443
1444 Version 5.0 13-Sep-04
1445 ---------------------
1446
1447 1. Internal change: literal characters are no longer packed up into items
1448 containing multiple characters in a single byte-string. Each character
1449 is now matched using a separate opcode. However, there may be more than one
1450 byte in the character in UTF-8 mode.
1451
1452 2. The pcre_callout_block structure has two new fields: pattern_position and
1453 next_item_length. These contain the offset in the pattern to the next match
1454 item, and its length, respectively.
1455
1456 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1457 insertion of callouts before each pattern item. Added the /C option to
1458 pcretest to make use of this.
1459
1460 4. On the advice of a Windows user, the lines
1461
1462 #if defined(_WIN32) || defined(WIN32)
1463 _setmode( _fileno( stdout ), 0x8000 );
1464 #endif /* defined(_WIN32) || defined(WIN32) */
1465
1466 have been added to the source of pcretest. This apparently does useful
1467 magic in relation to line terminators.
1468
1469 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1470 for the benefit of those environments where the "b" makes a difference.
1471
1472 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1473 to know about it. I have put a hack into configure.in that adds in code
1474 to set GCC=yes if CC=icc. This seems to end up at a point in the
1475 generated configure script that is early enough to affect the setting of
1476 compiler options, which is what is needed, but I have no means of testing
1477 whether it really works. (The user who reported this had patched the
1478 generated configure script, which of course I cannot do.)
1479
1480 LATER: After change 22 below (new libtool files), the configure script
1481 seems to know about icc (and also ecc). Therefore, I have commented out
1482 this hack in configure.in.
1483
1484 7. Added support for pkg-config (2 patches were sent in).
1485
1486 8. Negated POSIX character classes that used a combination of internal tables
1487 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1488 [[:^ascii]]. Typically, they would match almost any characters. The other
1489 POSIX classes were not broken in this way.
1490
1491 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1492 to find the match, as PCRE was deluded into thinking that the match had to
1493 start at the start point or following a newline. The same bug applied to
1494 patterns with negative forward assertions or any backward assertions
1495 preceding ".*" at the start, unless the pattern required a fixed first
1496 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1497
1498 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1499 starting at the last subject character, bytes beyond the end of the subject
1500 string were read.
1501
1502 11. Renamed the variable "class" as "classbits" to make life easier for C++
1503 users. (Previously there was a macro definition, but it apparently wasn't
1504 enough.)
1505
1506 12. Added the new field "tables" to the extra data so that tables can be passed
1507 in at exec time, or the internal tables can be re-selected. This allows
1508 a compiled regex to be saved and re-used at a later time by a different
1509 program that might have everything at different addresses.
1510
1511 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1512 -R library as well as a -L library.
1513
1514 14. The debugging options of pcretest (-d on the command line or D on a
1515 pattern) showed incorrect output for anything following an extended class
1516 that contained multibyte characters and which was followed by a quantifier.
1517
1518 15. Added optional support for general category Unicode character properties
1519 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1520 support. It adds about 90K to the size of the library. The meanings of the
1521 inbuilt class escapes such as \d and \s have NOT been changed.
1522
1523 16. Updated pcredemo.c to include calls to free() to release the memory for the
1524 compiled pattern.
1525
1526 17. The generated file chartables.c was being created in the source directory
1527 instead of in the building directory. This caused the build to fail if the
1528 source directory was different from the building directory, and was
1529 read-only.
1530
1531 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1532 file. No doubt somebody will tell me if they don't make sense... Also added
1533 Dan Mooney's comments about building on OpenVMS.
1534
1535 19. Added support for partial matching via the PCRE_PARTIAL option for
1536 pcre_exec() and the \P data escape in pcretest.
1537
1538 20. Extended pcretest with 3 new pattern features:
1539
1540 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1541 write the compiled pattern to the file whose name is "rest-of-line".
1542 This is a straight binary dump of the data, with the saved pointer to
1543 the character tables forced to be NULL. The study data, if any, is
1544 written too. After writing, pcretest reads a new pattern.
1545
1546 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1547 compiled pattern from the given file. There must not be any
1548 occurrences of "<" in the file name (pretty unlikely); if there are,
1549 pcretest will instead treat the initial "<" as a pattern delimiter.
1550 After reading in the pattern, pcretest goes on to read data lines as
1551 usual.
1552
1553 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1554 and 16-bit fields in a compiled pattern, to simulate a pattern that
1555 was compiled on a host of opposite endianness.
1556
1557 21. The pcre-exec() function can now cope with patterns that were compiled on
1558 hosts of opposite endianness, with this restriction:
1559
1560 As for any compiled expression that is saved and used later, the tables
1561 pointer field cannot be preserved; the extra_data field in the arguments
1562 to pcre_exec() should be used to pass in a tables address if a value
1563 other than the default internal tables were used at compile time.
1564
1565 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1566 now diagnosed as an error. Previously, most of the time, a negative number
1567 would have been treated as zero, but if in addition "ovector" was passed as
1568 NULL, a crash could occur.
1569
1570 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1571 new versions from the libtool 1.5 distribution (the last one is a copy of
1572 a file called libtool.m4). This seems to have fixed the need to patch
1573 "configure" to support Darwin 1.3 (which I used to do). However, I still
1574 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1575 workstation).
1576
1577 24. Changed the PCRE licence to be the more standard "BSD" licence.
1578
1579
1580 Version 4.5 01-Dec-03
1581 ---------------------
1582
1583 1. There has been some re-arrangement of the code for the match() function so
1584 that it can be compiled in a version that does not call itself recursively.
1585 Instead, it keeps those local variables that need separate instances for
1586 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1587 needs to "recurse". Keeping track of where control must go is done by means
1588 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1589 hide most of the details from the main code, and operates only if
1590 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1591 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1592 operating.
1593
1594 To make it easier for callers to provide specially tailored get/free
1595 functions for this usage, two new functions, pcre_stack_malloc, and
1596 pcre_stack_free, are used. They are always called in strict stacking order,
1597 and the size of block requested is always the same.
1598
1599 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1600 PCRE has been compiled to use the stack or the heap for recursion. The
1601 -C option of pcretest uses this to show which version is compiled.
1602
1603 A new data escape \S, is added to pcretest; it causes the amounts of store
1604 obtained and freed by both kinds of malloc/free at match time to be added
1605 to the output.
1606
1607 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1608 what's available on my current Linux desktop machine.
1609
1610 3. When matching a UTF-8 string, the test for a valid string at the start has
1611 been extended. If start_offset is not zero, PCRE now checks that it points
1612 to a byte that is the start of a UTF-8 character. If not, it returns
1613 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1614 this is necessary because there may be backward assertions in the pattern.
1615 When matching the same subject several times, it may save resources to use
1616 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1617
1618 4. The code for checking the validity of UTF-8 strings has been tightened so
1619 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1620 containing "overlong sequences".
1621
1622 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1623 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1624 should have been "&", but it just so happened that all the cases this let
1625 through by mistake were picked up later in the function.
1626
1627 6. I had used a variable called "isblank" - this is a C99 function, causing
1628 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1629
1630 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1631 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1632 so the version is shown; (c) stop "make test" echoing "./RunTest".
1633
1634 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1635
1636 9. The prototype for memmove() for systems that don't have it was using
1637 size_t, but the inclusion of the header that defines size_t was later. I've
1638 moved the #includes for the C headers earlier to avoid this.
1639
1640 10. Added some adjustments to the code to make it easier to compiler on certain
1641 special systems:
1642
1643 (a) Some "const" qualifiers were missing.
1644 (b) Added the macro EXPORT before all exported functions; by default this
1645 is defined to be empty.
1646 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1647 that it reads its output file name as an argument instead of writing
1648 to the standard output and assuming this can be redirected.
1649
1650 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1651 class containing characters with values greater than 255, PCRE compilation
1652 went into a loop.
1653
1654 12. A recursive reference to a subpattern that was within another subpattern
1655 that had a minimum quantifier of zero caused PCRE to crash. For example,
1656 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1657 recursion. If the recursively-called subpattern itself had a zero repeat,
1658 that was OK.
1659
1660 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1661 buffer into which it was copied (for escape processing) was still set at
1662 1024, so long lines caused crashes.
1663
1664 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1665 "internal error: code overflow...". This applied to any character class
1666 that was followed by a possessive quantifier.
1667
1668 15. Modified the Makefile to add libpcre.la as a prerequisite for
1669 libpcreposix.la because I was told this is needed for a parallel build to
1670 work.
1671
1672 16. If a pattern that contained .* following optional items at the start was
1673 studied, the wrong optimizing data was generated, leading to matching
1674 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1675 matching string must start with a or b or c. The correct conclusion for
1676 this pattern is that a match can start with any character.
1677
1678
1679 Version 4.4 13-Aug-03
1680 ---------------------
1681
1682 1. In UTF-8 mode, a character class containing characters with values between
1683 127 and 255 was not handled correctly if the compiled pattern was studied.
1684 In fixing this, I have also improved the studying algorithm for such
1685 classes (slightly).
1686
1687 2. Three internal functions had redundant arguments passed to them. Removal
1688 might give a very teeny performance improvement.
1689
1690 3. Documentation bug: the value of the capture_top field in a callout is *one
1691 more than* the number of the hightest numbered captured substring.
1692
1693 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1694 in incorrectly linking with a previously installed version. They now link
1695 explicitly with libpcre.la.
1696
1697 5. configure.in no longer needs to recognize Cygwin specially.
1698
1699 6. A problem in pcre.in for Windows platforms is fixed.
1700
1701 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1702 pcretest, it used to include the size of the study block as part of its
1703 output. Unfortunately, the structure contains a field that has a different
1704 size on different hardware architectures. This meant that the tests that
1705 showed this size failed. As the block is currently always of a fixed size,
1706 this information isn't actually particularly useful in pcretest output, so
1707 I have just removed it.
1708
1709 8. Three pre-processor statements accidentally did not start in column 1.
1710 Sadly, there are *still* compilers around that complain, even though
1711 standard C has not required this for well over a decade. Sigh.
1712
1713 9. In pcretest, the code for checking callouts passed small integers in the
1714 callout_data field, which is a void * field. However, some picky compilers
1715 complained about the casts involved for this on 64-bit systems. Now
1716 pcretest passes the address of the small integer instead, which should get
1717 rid of the warnings.
1718
1719 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1720 both compile and run time, and gives an error if an invalid UTF-8 sequence
1721 is found. There is a option for disabling this check in cases where the
1722 string is known to be correct and/or the maximum performance is wanted.
1723
1724 11. In response to a bug report, I changed one line in Makefile.in from
1725
1726 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1727 to
1728 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1729
1730 to look similar to other lines, but I have no way of telling whether this
1731 is the right thing to do, as I do not use Windows. No doubt I'll get told
1732 if it's wrong...
1733
1734
1735 Version 4.3 21-May-03
1736 ---------------------
1737
1738 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1739 Makefile.
1740
1741 2. Some refactoring to improve the quality of the code:
1742
1743 (i) The utf8_table... variables are now declared "const".
1744
1745 (ii) The code for \cx, which used the "case flipping" table to upper case
1746 lower case letters, now just substracts 32. This is ASCII-specific,
1747 but the whole concept of \cx is ASCII-specific, so it seems
1748 reasonable.
1749
1750 (iii) PCRE was using its character types table to recognize decimal and
1751 hexadecimal digits in the pattern. This is silly, because it handles
1752 only 0-9, a-f, and A-F, but the character types table is locale-
1753 specific, which means strange things might happen. A private
1754 table is now used for this - though it costs 256 bytes, a table is
1755 much faster than multiple explicit tests. Of course, the standard
1756 character types table is still used for matching digits in subject
1757 strings against \d.
1758
1759 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1760 ending in _t are). So I've renamed it as ESC_tee.
1761
1762 3. The first argument for regexec() in the POSIX wrapper should have been
1763 defined as "const".
1764
1765 4. Changed pcretest to use malloc() for its buffers so that they can be
1766 Electric Fenced for debugging.
1767
1768 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1769 to read one or more bytes before the start of the subject string. Often this
1770 had no effect on PCRE's behaviour, but in some circumstances it could
1771 provoke a segmentation fault.
1772
1773 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1774 to try to read one or more bytes before the start of the subject string.
1775
1776 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1777 UTF-8 support could misbehave in various ways if the subject string
1778 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1779 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1780 back over UTF-8 characters.)
1781
1782
1783 Version 4.2 14-Apr-03
1784 ---------------------
1785
1786 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1787
1788 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1789 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1790 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1791 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1792 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1793 and BUILD_EXEEXT
1794 Note: automatic setting of the BUILD variables is not yet working
1795 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1796 compile-time but not at link-time
1797 [LINK]: use for linking executables only
1798 make different versions for Windows and non-Windows
1799 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1800 libraries
1801 [LINK_FOR_BUILD]: new variable
1802 [OBJEXT]: use throughout
1803 [EXEEXT]: use throughout
1804 <winshared>: new target
1805 <wininstall>: new target
1806 <dftables.o>: use native compiler
1807 <dftables>: use native linker
1808 <install>: handle Windows platform correctly
1809 <clean>: ditto
1810 <check>: ditto
1811 copy DLL to top builddir before testing
1812
1813 As part of these changes, -no-undefined was removed again. This was reported
1814 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1815 in any case.
1816
1817 3. Some tidies to get rid of compiler warnings:
1818
1819 . In the match_data structure, match_limit was an unsigned long int, whereas
1820 match_call_count was an int. I've made them both unsigned long ints.
1821
1822 . In pcretest the fact that a const uschar * doesn't automatically cast to
1823 a void * provoked a warning.
1824
1825 . Turning on some more compiler warnings threw up some "shadow" variables
1826 and a few more missing casts.
1827
1828 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1829 option, a class that contained a single character with a value between 128
1830 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1831
1832 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1833 option, a class that contained several characters, but with at least one
1834 whose value was between 128 and 255 caused PCRE to crash.
1835
1836
1837 Version 4.1 12-Mar-03
1838 ---------------------
1839
1840 1. Compiling with gcc -pedantic found a couple of places where casts were
1841 needed, and a string in dftables.c that was longer than standard compilers are
1842 required to support.
1843
1844 2. Compiling with Sun's compiler found a few more places where the code could
1845 be tidied up in order to avoid warnings.
1846
1847 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1848 first of these names is deprecated in the latest Autoconf in favour of the name
1849 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1850 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1851 analogy I have changed it to CFLAGS_FOR_BUILD.
1852
1853 4. Added -no-undefined to the linking command in the Makefile, because this is
1854 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1855 linking step for the pcreposix library.
1856
1857 5. PCRE was failing to diagnose the case of two named groups with the same
1858 name.
1859
1860 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1861 literal character that is needed in the subject for a match, and scans along to
1862 ensure that it is present before embarking on the full matching process. This
1863 saves time in cases of nested unlimited repeats that are never going to match.
1864 Problem: the scan can take a lot of time if the subject is very long (e.g.
1865 megabytes), thus penalizing straightforward matches. It is now done only if the
1866 amount of subject to be scanned is less than 1000 bytes.
1867
1868 7. A lesser problem with the same optimization is that it was recording the
1869 first character of an anchored pattern as "needed", thus provoking a search
1870 right along the subject, even when the first match of the pattern was going to
1871 fail. The "needed" character is now not set for anchored patterns, unless it
1872 follows something in the pattern that is of non-fixed length. Thus, it still
1873 fulfils its original purpose of finding quick non-matches in cases of nested
1874 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1875
1876
1877 Version 4.0 17-Feb-03
1878 ---------------------
1879
1880 1. If a comment in an extended regex that started immediately after a meta-item
1881 extended to the end of string, PCRE compiled incorrect data. This could lead to
1882 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1883
1884 2. Moved to autoconf 2.53 and libtool 1.4.2.
1885
1886 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1887 the special perltest8 script is no longer needed - all the tests can be run
1888 from a single perltest script.
1889
1890 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1891 by \s. It has now been removed in PCRE. This means it isn't recognized as
1892 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1893 class [:space:] *does* include VT, thereby creating a mess.
1894
1895 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1896 space and tab.
1897
1898 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1899 its new features into the main test script, reducing the number of scripts.
1900
1901 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1902 were backward compatible, and made the (?i) apply to the whole pattern, as if
1903 /i were given. Now it behaves more logically, and applies the option setting
1904 only to what follows. PCRE has been changed to follow suit. However, if it
1905 finds options settings right at the start of the pattern, it extracts them into
1906 the global options, as before. Thus, they show up in the info data.
1907
1908 8. Added support for the \Q...\E escape sequence. Characters in between are
1909 treated as literals. This is slightly different from Perl in that $ and @ are
1910 also handled as literals inside the quotes. In Perl, they will cause variable
1911 interpolation. Note the following examples:
1912
1913 Pattern PCRE matches Perl matches
1914
1915 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1916 \Qabc\$xyz\E abc\$xyz abc\$xyz
1917 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1918
1919 For compatibility with Perl, \Q...\E sequences are recognized inside character
1920 classes as well as outside them.
1921
1922 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1923 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1924 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1925 signed/unsigned warnings.
1926
1927 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1928 option for pcretest, so I've replaced it by a simple function that does just
1929 that job.
1930
1931 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1932 "pcregrep -".
1933
1934 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1935 Java package. This provides some syntactic sugar for simple cases of what my
1936 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1937 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1938 item, you can use this simplified notation. Note that only makes sense with
1939 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1940 greediness, whatever the setting of the PCRE_UNGREEDY option.
1941
1942 13. A change of greediness default within a pattern was not taking effect at
1943 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1944 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1945 was abstracted outside.
1946
1947 14. PCRE now supports the \G assertion. It is true when the current matching
1948 position is at the start point of the match. This differs from \A when the
1949 starting offset is non-zero. Used with the /g option of pcretest (or similar
1950 code), it works in the same way as it does for Perl's /g option. If all
1951 alternatives of a regex begin with \G, the expression is anchored to the start
1952 match position, and the "anchored" flag is set in the compiled expression.
1953
1954 15. Some bugs concerning the handling of certain option changes within patterns
1955 have been fixed. These applied to options other than (?ims). For example,
1956 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1957 been the other way round. Some of this was related to change 7 above.
1958
1959 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1960 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1961 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1962 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1963
1964 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1965 mode. Unlike ".", it always matches newline, whatever the setting of
1966 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1967 assertions. Perl allows it, but it doesn't (in general) work because it can't
1968 calculate the length of the lookbehind. At least, that's the case for Perl
1969 5.8.0 - I've been told they are going to document that it doesn't work in
1970 future.
1971
1972 18. Added an error diagnosis for escapes that PCRE does not support: these are
1973 \L, \l, \N, \P, \p, \U, \u, and \X.
1974
1975 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1976 reading past the end of the pattern in cases such as /[abcd/.
1977
1978 20. PCRE was getting more memory than necessary for patterns with classes that
1979 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1980
1981 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1982 compiling PCRE for use with Virtual Pascal.
1983
1984 22. Small fix to the Makefile to make it work properly if the build is done
1985 outside the source tree.
1986
1987 23. Added a new extension: a condition to go with recursion. If a conditional
1988 subpattern starts with (?(R) the "true" branch is used if recursion has
1989 happened, whereas the "false" branch is used only at the top level.
1990
1991 24. When there was a very long string of literal characters (over 255 bytes
1992 without UTF support, over 250 bytes with UTF support), the computation of how
1993 much memory was required could be incorrect, leading to segfaults or other
1994 strange effects.
1995
1996 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1997 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1998 there was a subsequent back reference to those brackets. This meant that, for
1999 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2000 possible to check for precisely this case. All we can do is abandon the
2001 optimization if .* occurs inside capturing brackets when there are any back
2002 references whatsoever. (See below for a better fix that came later.)
2003
2004 26. The handling of the optimization for finding the first character of a
2005 non-anchored pattern, and for finding a character that is required later in the
2006 match were failing in some cases. This didn't break the matching; it just
2007 failed to optimize when it could. The way this is done has been re-implemented.
2008
2009 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2010
2011 28. Added a new feature that provides some of the functionality that Perl
2012 provides with (?{...}). The facility is termed a "callout". The way it is done
2013 in PCRE is for the caller to provide an optional function, by setting
2014 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2015 global variable. By default it is unset, which disables all calling out. To get
2016 the function called, the regex must include (?C) at appropriate points. This
2017 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2018 This provides a means of identifying different callout points. When PCRE
2019 reaches such a point in the regex, if pcre_callout has been set, the external
2020 function is called. It is provided with data in a structure called
2021 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2022 matching continues; if it returns a non-zero value, the match at the current
2023 point fails. However, backtracking will occur if possible. [This was changed
2024 later and other features added - see item 49 below.]
2025
2026 29. pcretest is upgraded to test the callout functionality. It provides a
2027 callout function that displays information. By default, it shows the start of
2028 the match and the current position in the text. There are some new data escapes
2029 to vary what happens:
2030
2031 \C+ in addition, show current contents of captured substrings
2032 \C- do not supply a callout function
2033 \C!n return 1 when callout number n is reached
2034 \C!n!m return 1 when callout number n is reached for the mth time
2035
2036 30. If pcregrep was called with the -l option and just a single file name, it
2037 output "<stdin>" if a match was found, instead of the file name.
2038
2039 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2040 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2041 pcre_exec(). This saves a malloc/free per call. The default value of
2042 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2043 when configuring.
2044
2045 32. The default maximum size of a compiled pattern is 64K. There have been a
2046 few cases of people hitting this limit. The code now uses macros to handle the
2047 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2048 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2049 configuring. Tests 2 and 5 work only with 2-byte links because they output
2050 debugging information about compiled patterns.
2051
2052 33. Internal code re-arrangements:
2053
2054 (a) Moved the debugging function for printing out a compiled regex into
2055 its own source file (printint.c) and used #include to pull it into
2056 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2057 separate copies.
2058
2059 (b) Defined the list of op-code names for debugging as a macro in
2060 internal.h so that it is next to the definition of the opcodes.
2061
2062 (c) Defined a table of op-code lengths for simpler skipping along compiled
2063 code. This is again a macro in internal.h so that it is next to the
2064 definition of the opcodes.
2065
2066 34. Added support for recursive calls to individual subpatterns, along the
2067 lines of Robin Houston's patch (but implemented somewhat differently).
2068
2069 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2070 allow it to read and process whole directories in Win32. This code was
2071 contributed by Lionel Fourquaux; it has not been tested by me.
2072
2073 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2074 used to name a group. Names consist of alphanumerics and underscores, and must
2075 be unique. Back references use the syntax (?P=name) and recursive calls use
2076 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2077 numbers. The function pcre_fullinfo() can be used after compilation to extract
2078 a name/number map. There are three relevant calls:
2079
2080 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2081 PCRE_INFO_NAMECOUNT yields the number of entries
2082 PCRE_INFO_NAMETABLE yields a pointer to the map.
2083
2084 The map is a vector of fixed-size entries. The size of each entry depends on
2085 the length of the longest name used. The first two bytes of each entry are the
2086 group number, most significant byte first. There follows the corresponding
2087 name, zero terminated. The names are in alphabetical order.
2088
2089 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2090 case instead of 255. Making it the same both with and without UTF-8 support
2091 means that the same test output works with both.
2092
2093 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2094 calling malloc() with a zero argument.
2095
2096 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2097 optimization. I've improved things by keeping a bitmap of backreferences with
2098 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2099 fact referenced, the optimization can be applied. It is unlikely that a
2100 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2101 the match to follow \n) will appear inside brackets with a number greater than
2102 31, but if it does, any back reference > 31 suppresses the optimization.
2103
2104 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2105 of disabling numbered capturing parentheses. Any opening parenthesis that is
2106 not followed by ? behaves as if it were followed by ?: but named parentheses
2107 can still be used for capturing (and they will acquire numbers in the usual
2108 way).
2109
2110 41. Redesigned the return codes from the match() function into yes/no/error so
2111 that errors can be passed back from deep inside the nested calls. A malloc
2112 failure while inside a recursive subpattern call now causes the
2113 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2114
2115 42. It is now possible to set a limit on the number of times the match()
2116 function is called in a call to pcre_exec(). This facility makes it possible to
2117 limit the amount of recursion and backtracking, though not in a directly
2118 obvious way, because the match() function is used in a number of different
2119 circumstances. The count starts from zero for each position in the subject
2120 string (for non-anchored patterns). The default limit is, for compatibility, a
2121 large number, namely 10 000 000. You can change this in two ways:
2122
2123 (a) When configuring PCRE before making, you can use --with-match-limit=n
2124 to set a default value for the compiled library.
2125
2126 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2127 a different value is set. See 45 below.
2128
2129 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2130
2131 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2132 of things that can be changed at compile time. The first argument specifies
2133 what is wanted and the second points to where the information is to be placed.
2134 The current list of available information is:
2135
2136 PCRE_CONFIG_UTF8
2137
2138 The output is an integer that is set to one if UTF-8 support is available;
2139 otherwise it is set to zero.
2140
2141 PCRE_CONFIG_NEWLINE
2142
2143 The output is an integer that it set to the value of the code that is used for
2144 newline. It is either LF (10) or CR (13).
2145
2146 PCRE_CONFIG_LINK_SIZE
2147
2148 The output is an integer that contains the number of bytes used for internal
2149 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2150
2151 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2152
2153 The output is an integer that contains the threshold above which the POSIX
2154 interface uses malloc() for output vectors. See item 31 above.
2155
2156 PCRE_CONFIG_MATCH_LIMIT
2157
2158 The output is an unsigned integer that contains the default limit of the number
2159 of match() calls in a pcre_exec() execution. See 42 above.
2160
2161 44. pcretest has been upgraded by the addition of the -C option. This causes it
2162 to extract all the available output from the new pcre_config() function, and to
2163 output it. The program then exits immediately.
2164
2165 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2166 order to support additional features. One way would have been to define
2167 pcre_exec2() (for example) with extra arguments, but this would not have been
2168 extensible, and would also have required all calls to the original function to
2169 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2170 is used for passing in "extra" data from pcre_study().
2171
2172 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2173 contains the following fields:
2174
2175 flags a bitmap indicating which of the following fields are set
2176 study_data opaque data from pcre_study()
2177 match_limit a way of specifying a limit on match() calls for a specific
2178 call to pcre_exec()
2179 callout_data data for callouts (see 49 below)
2180
2181 The flag bits are also defined in pcre.h, and are
2182
2183 PCRE_EXTRA_STUDY_DATA
2184 PCRE_EXTRA_MATCH_LIMIT
2185 PCRE_EXTRA_CALLOUT_DATA
2186
2187 The pcre_study() function now returns one of these new pcre_extra blocks, with
2188 the actual study data pointed to by the study_data field, and the
2189 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2190 before. That is, this change is entirely upwards-compatible and requires no
2191 change to existing code.
2192
2193 If you want to pass in additional data to pcre_exec(), you can either place it
2194 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2195 block.
2196
2197 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2198 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2199 times with different match limits, until it finds the minimum value needed for
2200 pcre_exec() to complete. The value is then output. This can be instructive; for
2201 most simple matches the number is quite small, but for pathological cases it
2202 gets very large very quickly.
2203
2204 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2205 returns the size of the data block pointed to by the study_data field in a
2206 pcre_extra block, that is, the value that was passed as the argument to
2207 pcre_malloc() when PCRE was getting memory in which to place the information
2208 created by pcre_study(). The fourth argument should point to a size_t variable.
2209 pcretest has been extended so that this information is shown after a successful
2210 pcre_study() call when information about the compiled regex is being displayed.
2211
2212 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2213 because what follows is always an absolute path. (Later: it turns out that this
2214 is more than cosmetic for MinGW, because it doesn't like empty path
2215 components.)
2216
2217 49. Some changes have been made to the callout feature (see 28 above):
2218
2219 (i) A callout function now has three choices for what it returns:
2220
2221 0 => success, carry on matching
2222 > 0 => failure at this point, but backtrack if possible
2223 < 0 => serious error, return this value from pcre_exec()
2224
2225 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2226 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2227 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2228 use by callout functions. It will never be used by PCRE itself.
2229
2230 (ii) The pcre_extra structure (see 45 above) has a void * field called
2231 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2232 pcre_callout_block structure has a field of the same name. The contents of
2233 the field passed in the pcre_extra structure are passed to the callout
2234 function in the corresponding field in the callout block. This makes it
2235 easier to use the same callout-containing regex from multiple threads. For
2236 testing, the pcretest program has a new data escape
2237
2238 \C*n pass the number n (may be negative) as callout_data
2239
2240 If the callout function in pcretest receives a non-zero value as
2241 callout_data, it returns that value.
2242
2243 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2244 there were some redundant $(CFLAGS) in commands that are now specified as
2245 $(LINK), which already includes $(CFLAGS).
2246
2247 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2248 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2249 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2250 one-byte characters throughout. Note that case-insensitive matching applies
2251 only to characters whose values are less than 256. PCRE doesn't support the
2252 notion of cases for higher-valued characters.
2253
2254 (i) A character class whose characters are all within 0-255 is handled as
2255 a bit map, and the map is inverted for negative classes. Previously, a
2256 character > 255 always failed to match such a class; however it should
2257 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2258
2259 (ii) A negated character class with a single character < 255 is coded as
2260 "not this character" (OP_NOT). This wasn't working properly when the test
2261 character was multibyte, either singly or repeated.
2262
2263 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2264 mode, for example: \x{100}{2,3}.
2265
2266 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2267 singly or repeated) now correctly test multibyte characters. However,
2268 PCRE doesn't recognize any characters with values greater than 255 as
2269 digits, spaces, or word characters. Such characters always match \D, \S,
2270 and \W, and never match \d, \s, or \w.
2271
2272 (v) Classes may now contain characters and character ranges with values
2273 greater than 255. For example: [ab\x{100}-\x{400}].
2274
2275 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2276 PCRE in UTF-8 mode.
2277
2278 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2279 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2280 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2281 value.)
2282
2283 53. The single man page has become too large. I have therefore split it up into
2284 a number of separate man pages. These also give rise to individual HTML pages;
2285 these are now put in a separate directory, and there is an index.html page that
2286 lists them all. Some hyperlinking between the pages has been installed.
2287
2288 54. Added convenience functions for handling named capturing parentheses.
2289
2290 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2291 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2292 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2293 are faulted.
2294
2295 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2296 calling configure. These values are used when compiling the dftables.c program
2297 which is run to generate the source of the default character tables. They
2298 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2299 you will need to set these values.
2300
2301 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2302
2303
2304 Version 3.9 02-Jan-02
2305 ---------------------
2306
2307 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2308
2309 2. If --disable-static was given, the building process failed when trying to
2310 build pcretest and pcregrep. (For some reason it was using libtool to compile
2311 them, which is not right, as they aren't part of the library.)
2312
2313
2314 Version 3.8 18-Dec-01
2315 ---------------------
2316
2317 1. The experimental UTF-8 code was completely screwed up. It was packing the
2318 bytes in the wrong order. How dumb can you get?
2319
2320
2321 Version 3.7 29-Oct-01
2322 ---------------------
2323
2324 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2325 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2326 this didn't happen under Solaris 8, where I normally test things.
2327
2328 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2329 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2330 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2331 this fix an hour or so after the initial 3.7 release.)
2332
2333
2334 Version 3.6 23-Oct-01
2335 ---------------------
2336
2337 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2338 offsets passed as NULL with zero offset count.
2339
2340 2. The config.guess and config.sub files had not been updated when I moved to
2341 the latest autoconf.
2342
2343
2344 Version 3.5 15-Aug-01
2345 ---------------------
2346
2347 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2348 had been forgotten.
2349
2350 2. By using declared but undefined structures, we can avoid using "void"
2351 definitions in pcre.h while keeping the internal definitions of the structures
2352 private.
2353
2354 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2355 user point of view, this means that both static and shared libraries are built
2356 by default, but this can be individually controlled. More of the work of
2357 handling this static/shared cases is now inside libtool instead of PCRE's make
2358 file.
2359
2360 4. The pcretest utility is now installed along with pcregrep because it is
2361 useful for users (to test regexs) and by doing this, it automatically gets
2362 relinked by libtool. The documentation has been turned into a man page, so
2363 there are now .1, .txt, and .html versions in /doc.
2364
2365 5. Upgrades to pcregrep:
2366 (i) Added long-form option names like gnu grep.
2367 (ii) Added --help to list all options with an explanatory phrase.
2368 (iii) Added -r, --recursive to recurse into sub-directories.
2369 (iv) Added -f, --file to read patterns from a file.
2370
2371 6. pcre_exec() was referring to its "code" argument before testing that
2372 argument for NULL (and giving an error if it was NULL).
2373
2374 7. Upgraded Makefile.in to allow for compiling in a different directory from
2375 the source directory.
2376
2377 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2378 options bits, the pointer it was passed was to an int instead of to an unsigned
2379 long int. This mattered only on 64-bit systems.
2380
2381 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2382 generated) instead of pcre.in, which it its source. Also made the same change
2383 in several of the .c files.
2384
2385 10. A new release of gcc defines printf() as a macro, which broke pcretest
2386 because it had an ifdef in the middle of a string argument for printf(). Fixed
2387 by using separate calls to printf().
2388
2389 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2390 script, to force use of CR or LF instead of \n in the source. On non-Unix
2391 systems, the value can be set in config.h.
2392
2393 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2394 absolute limit. Changed the text of the error message to make this clear, and
2395 likewise updated the man page.
2396
2397 13. The limit of 99 on the number of capturing subpatterns has been removed.
2398 The new limit is 65535, which I hope will not be a "real" limit.
2399
2400
2401 Version 3.4 22-Aug-00
2402 ---------------------
2403
2404 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2405
2406 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2407
2408
2409 Version 3.3 01-Aug-00
2410 ---------------------
2411
2412 1. If an octal character was given, but the value was greater than \377, it
2413 was not getting masked to the least significant bits, as documented. This could
2414 lead to crashes in some systems.
2415
2416 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2417 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2418
2419 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2420 These just pass their arguments on to (pcre_free)(), but they are provided
2421 because some uses of PCRE bind it to non-C systems that can call its functions,
2422 but cannot call free() or pcre_free() directly.
2423
2424 4. Add "make test" as a synonym for "make check". Corrected some comments in
2425 the Makefile.
2426
2427 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2428 Makefile.
2429
2430 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2431 command called pgrep for grepping around the active processes.
2432
2433 7. Added the beginnings of support for UTF-8 character strings.
2434
2435 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2436 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2437 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2438 out for the ar command.)
2439
2440
2441 Version 3.2 12-May-00
2442 ---------------------
2443
2444 This is purely a bug fixing release.
2445
2446 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2447 of ZA. This was just one example of several cases that could provoke this bug,
2448 which was introduced by change 9 of version 2.00. The code for breaking
2449 infinite loops after an iteration that matches an empty string was't working
2450 correctly.
2451
2452 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2453 when matched against abbab (for example). After matching an empty string, it
2454 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2455 caused it to match further down the string than it should.
2456
2457 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2458 was there because it doesn't seem to be needed, and it causes trouble on some
2459 systems, as it is not a Standard C header. It has been removed.
2460
2461 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2462 were reported on the Macintosh. The changes were from
2463
2464 while ((c = *(++ptr)) != 0 && c != '\n');
2465 to
2466 while ((c = *(++ptr)) != 0 && c != '\n') ;
2467
2468 Totally extraordinary, but if that's what it takes...
2469
2470 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2471 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2472 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2473 assumes the way PCRE uses memmove() (always moving upwards).
2474
2475 6. PCRE is being used in one environment where strchr() is not available. There
2476 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2477 faster code anyway.
2478
2479
2480 Version 3.1 09-Feb-00
2481 ---------------------
2482
2483 The only change in this release is the fixing of some bugs in Makefile.in for
2484 the "install" target:
2485
2486 (1) It was failing to install pcreposix.h.
2487
2488 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2489
2490
2491 Version 3.0 01-Feb-00
2492 ---------------------
2493
2494 1. Add support for the /+ modifier to perltest (to output $` like it does in
2495 pcretest).
2496
2497 2. Add support for the /g modifier to perltest.
2498
2499 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2500 matches null strings.
2501
2502 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2503 pattern. Perl treats empty patterns specially - it reuses the most recent
2504 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2505 effect.
2506
2507 5. The POSIX interface was broken in that it was just handing over the POSIX
2508 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2509 required a bigger vector, with some working space on the end. This means that
2510 the POSIX wrapper now has to get and free some memory, and copy the results.
2511
2512 6. Added some simple autoconf support, placing the test data and the
2513 documentation in separate directories, re-organizing some of the
2514 information files, and making it build pcre-config (a GNU standard). Also added
2515 libtool support for building PCRE as a shared library, which is now the
2516 default.
2517
2518 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2519 09 are not valid octal constants. Single digits will be used for minor values
2520 less than 10.
2521
2522 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2523 existing programs that set these in the POSIX interface can use PCRE without
2524 modification.
2525
2526 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2527 return all that pcre_info() returns, plus additional data. The pcre_info()
2528 function is retained for compatibility, but is considered to be obsolete.
2529
2530 10. Added experimental recursion feature (?R) to handle one common case that
2531 Perl 5.6 will be able to do with (?p{...}).
2532
2533 11. Added support for POSIX character classes like [:alpha:], which Perl is
2534 adopting.
2535
2536
2537 Version 2.08 31-Aug-99
2538 ----------------------
2539
2540 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2541 trying to match at the startoffset position, but instead was moving forward to
2542 the next newline as if a previous match had failed.
2543
2544 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2545 and could get into a loop if a null string was matched other than at the start
2546 of the subject.
2547
2548 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2549 be distinguished at compile time, and for completeness also added PCRE_DATE.
2550
2551 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2552 in GnuWin32 environments.
2553
2554
2555 Version 2.07 29-Jul-99
2556 ----------------------
2557
2558 1. The documentation is now supplied in plain text form and HTML as well as in
2559 the form of man page sources.
2560
2561 2. C++ compilers don't like assigning (void *) values to other pointer types.
2562 In particular this affects malloc(). Although there is no problem in Standard
2563 C, I've put in casts to keep C++ compilers happy.
2564
2565 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2566 should be (const char *).
2567
2568 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2569 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2570 However, I haven't made this a standard facility. The documentation doesn't
2571 mention it, and the Makefile doesn't support it.
2572
2573 5. The Makefile now contains an "install" target, with editable destinations at
2574 the top of the file. The pcretest program is not installed.
2575
2576 6. pgrep -V now gives the PCRE version number and date.
2577
2578 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2579 causing the entire string to be ignored, instead of just the last character.
2580
2581 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2582 non-matching string, it can take a very, very long time, even for strings of
2583 quite modest length, because of the nested recursion. PCRE now does better in
2584 some of these cases. It does this by remembering the last required literal
2585 character in the pattern, and pre-searching the subject to ensure it is present
2586 before running the real match. In other words, it applies a heuristic to detect
2587 some types of certain failure quickly, and in the above example, if presented
2588 with a string that has no trailing " it gives "no match" very quickly.
2589
2590 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2591 other alternatives are tried instead.
2592
2593
2594 Version 2.06 09-Jun-99
2595 ----------------------
2596
2597 1. Change pcretest's output for amount of store used to show just the code
2598 space, because the remainder (the data block) varies in size between 32-bit and
2599 64-bit systems.
2600
2601 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2602 start matching at. This allows lookbehinds to work when searching for multiple
2603 occurrences in a string.
2604
2605 3. Added additional options to pcretest for testing multiple occurrences:
2606
2607 /+ outputs the rest of the string that follows a match
2608 /g loops for multiple occurrences, using the new startoffset argument
2609 /G loops for multiple occurrences by passing an incremented pointer
2610
2611 4. PCRE wasn't doing the "first character" optimization for patterns starting
2612 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2613 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2614 the letter 'x'. On long subject strings, this gives a significant speed-up.
2615
2616
2617 Version 2.05 21-Apr-99
2618 ----------------------
2619
2620 1. Changed the type of magic_number from int to long int so that it works
2621 properly on 16-bit systems.
2622
2623 2. Fixed a bug which caused patterns starting with .* not to work correctly
2624 when the subject string contained newline characters. PCRE was assuming
2625 anchoring for such patterns in all cases, which is not correct because .* will
2626 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2627 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2628 must be retried after every newline in the subject.
2629
2630
2631 Version 2.04 18-Feb-99
2632 ----------------------
2633
2634 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2635 computation of the store needed to hold the pattern was incorrect (too large).
2636 If such patterns were nested a few deep, this could multiply and become a real
2637 problem.
2638
2639 2. Added /M option to pcretest to show the memory requirement of a specific
2640 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2641
2642 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2643 compiled in such a way that the backtracking after subsequent failure was
2644 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2645 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2646
2647
2648 Version 2.03 02-Feb-99
2649 ----------------------
2650
2651 1. Fixed typo and small mistake in man page.
2652
2653 2. Added 4th condition (GPL supersedes if conflict) and created separate
2654 LICENCE file containing the conditions.
2655
2656 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2657 Perl, that is the internal \ allows the delimiter to be included in the
2658 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2659 the final delimiter, add \ to the end of the pattern (to test the error).
2660
2661 4. Added the convenience functions for extracting substrings after a successful
2662 match. Updated pcretest to make it able to test these functions.
2663
2664
2665 Version 2.02 14-Jan-99
2666 ----------------------
2667
2668 1. Initialized the working variables associated with each extraction so that
2669 their saving and restoring doesn't refer to uninitialized store.
2670
2671 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2672 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2673 fix the problem.
2674
2675 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2676 calls, and wasn't printing the correct value for compiling calls. Increased the
2677 default value of LOOPREPEAT, and the number of significant figures in the
2678 times.
2679
2680 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2681
2682 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2683 a building problem on Windows NT with a FAT file system.
2684
2685
2686 Version 2.01 21-Oct-98
2687 ----------------------
2688
2689 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2690 to character tables built by pcre_maketables() in the current locale. If NULL
2691 is passed, the default tables are used.
2692
2693
2694 Version 2.00 24-Sep-98
2695 ----------------------
2696
2697 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2698 it any more.
2699
2700 2. Allow quantification of (?>) groups, and make it work correctly.
2701
2702 3. The first character computation wasn't working for (?>) groups.
2703
2704 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2705 end of the subject) and add 5.005's \z, which really does match only at the
2706 very end of the subject.
2707
2708 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2709
2710 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2711 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2712 localized options. All options to pcre_study() were also removed.
2713
2714 7. Add other new features from 5.005:
2715
2716 $(?<= positive lookbehind
2717 $(?<! negative lookbehind
2718 (?imsx-imsx) added the unsetting capability
2719 such a setting is global if at outer level; local otherwise
2720 (?imsx-imsx:) non-capturing groups with option setting
2721 (?(cond)re|re) conditional pattern matching
2722
2723 A backreference to itself in a repeated group matches the previous
2724 captured string.
2725
2726 8. General tidying up of studying (both automatic and via "study")
2727 consequential on the addition of new assertions.
2728
2729 9. As in 5.005, unlimited repeated groups that could match an empty substring
2730 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2731 runtime if any iteration does actually match an empty substring.
2732
2733 10. Include the RunTest script in the distribution.
2734
2735 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2736 discrepancies, some of which were old and were also with respect to 5.004. They
2737 have now been fixed.
2738
2739
2740 Version 1.09 28-Apr-98
2741 ----------------------
2742
2743 1. A negated single character class followed by a quantifier with a minimum
2744 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2745 program crashes, or just wrong answers. This did not apply to negated classes
2746 containing more than one character, or to minima other than one.
2747
2748
2749 Version 1.08 27-Mar-98
2750 ----------------------
2751
2752 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2753
2754 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2755 latter must appear before anything that relies on it in the pattern.
2756
2757
2758 Version 1.07 16-Feb-98
2759 ----------------------
2760
2761 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2762 repeat of a potentially empty string).
2763
2764
2765 Version 1.06 23-Jan-98
2766 ----------------------
2767
2768 1. Added Markus Oberhumer's little patches for C++.
2769
2770 2. Literal strings longer than 255 characters were broken.
2771
2772
2773 Version 1.05 23-Dec-97
2774 ----------------------
2775
2776 1. Negated character classes containing more than one character were failing if
2777 PCRE_CASELESS was set at run time.
2778
2779
2780 Version 1.04 19-Dec-97
2781 ----------------------
2782
2783 1. Corrected the man page, where some "const" qualifiers had been omitted.
2784
2785 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2786 input syntax.
2787
2788 3. Fixed memory leak which occurred when a regex with back references was
2789 matched with an offsets vector that wasn't big enough. The temporary memory
2790 that is used in this case wasn't being freed if the match failed.
2791
2792 4. Tidied pcretest to ensure it frees memory that it gets.
2793
2794 5. Temporary memory was being obtained in the case where the passed offsets
2795 vector was exactly big enough.
2796
2797 6. Corrected definition of offsetof() from change 5 below.
2798
2799 7. I had screwed up change 6 below and broken the rules for the use of
2800 setjmp(). Now fixed.
2801
2802
2803 Version 1.03 18-Dec-97
2804 ----------------------
2805
2806 1. A erroneous regex with a missing opening parenthesis was correctly
2807 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2808 on some systems.
2809
2810 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2811 it was reported that one broken compiler failed on the former because "code" is
2812 also an independent variable.
2813
2814 3. The erroneous regex a[]b caused an array overrun reference.
2815
2816 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2817 fail on data ending with that character. (It was going on too far, and checking
2818 the next character, typically a binary zero.) This was specific to the
2819 optimized code for single-character negative classes.
2820
2821 5. Added a contributed patch from the TIN world which does the following:
2822
2823 + Add an undef for memmove, in case the the system defines a macro for it.
2824
2825 + Add a definition of offsetof(), in case there isn't one. (I don't know
2826 the reason behind this - offsetof() is part of the ANSI standard - but
2827 it does no harm).
2828
2829 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2830 most of the places where whitespace preceded '#'. I have given up and
2831 allowed the remaining 2 cases to be at the margin.
2832
2833 + Rename some variables in pcre to eliminate shadowing. This seems very
2834 pedantic, but does no harm, of course.
2835
2836 6. Moved the call to setjmp() into its own function, to get rid of warnings
2837 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2838
2839 7. Constructs such as \d{8,} were compiling into the equivalent of
2840 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2841 outcome, but in this particular case used more store than had been allocated,
2842 which caused the bug to be discovered because it threw up an internal error.
2843
2844 8. The debugging code in both pcre and pcretest for outputting the compiled
2845 form of a regex was going wrong in the case of back references followed by
2846 curly-bracketed repeats.
2847
2848
2849 Version 1.02 12-Dec-97
2850 ----------------------
2851
2852 1. Typos in pcre.3 and comments in the source fixed.
2853
2854 2. Applied a contributed patch to get rid of places where it used to remove
2855 'const' from variables, and fixed some signed/unsigned and uninitialized
2856 variable warnings.
2857
2858 3. Added the "runtest" target to Makefile.
2859
2860 4. Set default compiler flag to -O2 rather than just -O.
2861
2862
2863 Version 1.01 19-Nov-97
2864 ----------------------
2865
2866 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2867 like /([ab]*)*/, that is, for classes with more than one character in them.
2868
2869 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2870 as /((?>a*))*/ (a PCRE_EXTRA facility).
2871
2872
2873 Version 1.00 18-Nov-97
2874 ----------------------
2875
2876 1. Added compile-time macros to support systems such as SunOS4 which don't have
2877 memmove() or strerror() but have other things that can be used instead.
2878
2879 2. Arranged that "make clean" removes the executables.
2880
2881
2882 Version 0.99 27-Oct-97
2883 ----------------------
2884
2885 1. Fixed bug in code for optimizing classes with only one character. It was
2886 initializing a 32-byte map regardless, which could cause it to run off the end
2887 of the memory it had got.
2888
2889 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2890
2891
2892 Version 0.98 22-Oct-97
2893 ----------------------
2894
2895 1. Fixed bug in code for handling temporary memory usage when there are more
2896 back references than supplied space in the ovector. This could cause segfaults.
2897
2898
2899 Version 0.97 21-Oct-97
2900 ----------------------
2901
2902 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2903
2904 2. Optimized negated single characters not to use a bit map.
2905
2906 3. Brought error texts together as macro definitions; clarified some of them;
2907 fixed one that was wrong - it said "range out of order" when it meant "invalid
2908 escape sequence".
2909
2910 4. Changed some char * arguments to const char *.
2911
2912 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2913
2914 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2915 pcretest.
2916
2917
2918 Version 0.96 16-Oct-97
2919 ----------------------
2920
2921 1. Added a simple "pgrep" utility to the distribution.
2922
2923 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2924 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2925 where "ddd" means "one or more decimal digits".
2926
2927 3. Fixed serious bug. If a pattern had a back reference, but the call to
2928 pcre_exec() didn't supply a large enough ovector to record the related
2929 identifying subpattern, the match always failed. PCRE now remembers the number
2930 of the largest back reference, and gets some temporary memory in which to save
2931 the offsets during matching if necessary, in order to ensure that
2932 backreferences always work.
2933
2934 4. Increased the compatibility with Perl in a number of ways:
2935
2936 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2937 to request this handling. The option can be set at compile or exec time.
2938
2939 (b) $ matches before a terminating newline by default; an option
2940 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2941 mode). The option can be set at compile or exec time.
2942
2943 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2944 the same as Perl's. If the decimal number it represents is less than 10
2945 or there aren't that many previous left capturing parentheses, an octal
2946 escape is read. Inside a character class, it's always an octal escape,
2947 even if it is a single digit.
2948
2949 (d) An escaped but undefined alphabetic character is taken as a literal,
2950 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2951 escapes.
2952
2953 (e) {0} is now permitted. (The previous item is removed from the compiled
2954 pattern).
2955
2956 5. Changed all the names of code files so that the basic parts are no longer
2957 than 10 characters, and abolished the teeny "globals.c" file.
2958
2959 6. Changed the handling of character classes; they are now done with a 32-byte
2960 bit map always.
2961
2962 7. Added the -d and /D options to pcretest to make it possible to look at the
2963 internals of compilation without having to recompile pcre.
2964
2965
2966 Version 0.95 23-Sep-97
2967 ----------------------
2968
2969 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2970 \x20 at the start of a run of normal characters. These were being treated as
2971 real characters, instead of the source characters being re-checked.
2972
2973
2974 Version 0.94 18-Sep-97
2975 ----------------------
2976
2977 1. The functions are now thread-safe, with the caveat that the global variables
2978 containing pointers to malloc() and free() or alternative functions are the
2979 same for all threads.
2980
2981 2. Get pcre_study() to generate a bitmap of initial characters for non-
2982 anchored patterns when this is possible, and use it if passed to pcre_exec().
2983
2984
2985 Version 0.93 15-Sep-97
2986 ----------------------
2987
2988 1. /(b)|(:+)/ was computing an incorrect first character.
2989
2990 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2991 but not actually doing anything yet.
2992
2993 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2994 as Perl does (e.g. [-az] or [az-]).
2995
2996 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2997 all possible positions.
2998
2999 5. Split up into different modules to avoid including unneeded functions in a
3000 compiled binary. However, compile and exec are still in one module. The "study"
3001 function is split off.
3002
3003 6. The character tables are now in a separate module whose source is generated
3004 by an auxiliary program - but can then be edited by hand if required. There are
3005 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3006 toupper() in the code.
3007
3008 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3009 make them global. Abolish the function for setting them, as the caller can now
3010 set them directly.
3011
3012
3013 Version 0.92 11-Sep-97
3014 ----------------------
3015
3016 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3017 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3018
3019 2. Caseless matching was not working in character classes if the characters in
3020 the pattern were in upper case.
3021
3022 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3023
3024 4. Make PCRE_ANCHORED public and accept as a compile option.
3025
3026 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3027 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3028 pass them.
3029
3030 6. Give an error if bad option bits passed at compile or run time.
3031
3032 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3033 pcretest to cause it to pass that flag.
3034
3035 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3036 options, and the first character, if set.
3037
3038 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3039
3040
3041 Version 0.91 10-Sep-97
3042 ----------------------
3043
3044 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3045 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3046
3047 2. PCRE was looping on encountering an indefinitely repeated back reference to
3048 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3049 Perl does - treats the match as successful.
3050
3051 ****

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