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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 277 - (show annotations) (download)
Mon Nov 26 17:03:45 2007 UTC (7 years ago) by ph10
File size: 144764 byte(s)
Fix incorrect script bug in Unicode character table.

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 8. Change "alphameric" to the more common word "alphanumeric" in comments
40 and messages.
41
42 9. Fix two occurrences of "backslash" in comments that should have been
43 "backspace".
44
45 10. Remove two redundant lines of code that can never be obeyed (their function
46 was moved elsewhere).
47
48 11. The program that makes PCRE's Unicode character property table had a bug
49 which caused it to generate incorrect table entries for sequences of
50 characters that have the same character type, but are in different scripts.
51 It amalgamated them into a single range, with the script of the first of
52 them. In other words, some characters were in the wrong script. There were
53 thirteen such cases, affecting characters in the following ranges:
54
55 U+002b0 - U+002c1
56 U+0060c - U+0060d
57 U+0061e - U+00612
58 U+0064b - U+0065e
59 U+0074d - U+0076d
60 U+01800 - U+01805
61 U+01d00 - U+01d77
62 U+01d9b - U+01dbf
63 U+0200b - U+0200f
64 U+030fc - U+030fe
65 U+03260 - U+0327f
66 U+0fb46 - U+0fbb1
67 U+10450 - U+1049d
68
69
70 Version 7.4 21-Sep-07
71 ---------------------
72
73 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
74 means that a class such as [\s] counted as "explicit reference to CR or
75 LF". That isn't really right - the whole point of the change was to try to
76 help when there was an actual mention of one of the two characters. So now
77 the change happens only if \r or \n (or a literal CR or LF) character is
78 encountered.
79
80 2. The 32-bit options word was also used for 6 internal flags, but the numbers
81 of both had grown to the point where there were only 3 bits left.
82 Fortunately, there was spare space in the data structure, and so I have
83 moved the internal flags into a new 16-bit field to free up more option
84 bits.
85
86 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
87 but did not set the internal JCHANGED flag - either of these is enough to
88 control the way the "get" function works - but the PCRE_INFO_JCHANGED
89 facility is supposed to tell if (?J) was ever used, so now (?J) at the
90 start sets both bits.
91
92 4. Added options (at build time, compile time, exec time) to change \R from
93 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
94
95 5. doc/pcresyntax.html was missing from the distribution.
96
97 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
98 compatibility, even though it is no longer used.
99
100 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
101 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
102 windows.h file is present (where different names are used). [This was
103 reversed later after testing - see 16 below.]
104
105 8. Changed all #include <config.h> to #include "config.h". There were also
106 some further <pcre.h> cases that I changed to "pcre.h".
107
108 9. When pcregrep was used with the --colour option, it missed the line ending
109 sequence off the lines that it output.
110
111 10. It was pointed out to me that arrays of string pointers cause lots of
112 relocations when a shared library is dynamically loaded. A technique of
113 using a single long string with a table of offsets can drastically reduce
114 these. I have refactored PCRE in four places to do this. The result is
115 dramatic:
116
117 Originally: 290
118 After changing UCP table: 187
119 After changing error message table: 43
120 After changing table of "verbs" 36
121 After changing table of Posix names 22
122
123 Thanks to the folks working on Gregex for glib for this insight.
124
125 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
126 unicode-properties was also set.
127
128 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
129
130 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
131 checked only for CRLF.
132
133 14. Added casts to pcretest.c to avoid compiler warnings.
134
135 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
136
137 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
138 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
139 entirely. This removes changes made in 7 above.
140
141 17. The CMake files have been updated, and there is now more information about
142 building with CMake in the NON-UNIX-USE document.
143
144
145 Version 7.3 28-Aug-07
146 ---------------------
147
148 1. In the rejigging of the build system that eventually resulted in 7.1, the
149 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
150 brackets there is not right, since it causes compilers to look for an
151 installed pcre.h, not the version that is in the source that is being
152 compiled (which of course may be different). I have changed it back to:
153
154 #include "pcre.h"
155
156 I have a vague recollection that the change was concerned with compiling in
157 different directories, but in the new build system, that is taken care of
158 by the VPATH setting the Makefile.
159
160 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
161 when the subject happened to end in the byte 0x85 (e.g. if the last
162 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
163 characters but of course it shouldn't be taken as a newline when it is part
164 of another character. The bug was that, for an unlimited repeat of . in
165 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
166 characters when looking for a newline.
167
168 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
169
170 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
171 in debug output.
172
173 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
174 long printing in the pcrecpp unittest when running under MinGW.
175
176 6. ESC_K was left out of the EBCDIC table.
177
178 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
179 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
180 limit also applies to "virtual nesting" when a pattern is recursive, and in
181 this case 1000 isn't so big. I have been able to remove this limit at the
182 expense of backing off one optimization in certain circumstances. Normally,
183 when pcre_exec() would call its internal match() function recursively and
184 immediately return the result unconditionally, it uses a "tail recursion"
185 feature to save stack. However, when a subpattern that can match an empty
186 string has an unlimited repetition quantifier, it no longer makes this
187 optimization. That gives it a stack frame in which to save the data for
188 checking that an empty string has been matched. Previously this was taken
189 from the 1000-entry workspace that had been reserved. So now there is no
190 explicit limit, but more stack is used.
191
192 8. Applied Daniel's patches to solve problems with the import/export magic
193 syntax that is required for Windows, and which was going wrong for the
194 pcreposix and pcrecpp parts of the library. These were overlooked when this
195 problem was solved for the main library.
196
197 9. There were some crude static tests to avoid integer overflow when computing
198 the size of patterns that contain repeated groups with explicit upper
199 limits. As the maximum quantifier is 65535, the maximum group length was
200 set at 30,000 so that the product of these two numbers did not overflow a
201 32-bit integer. However, it turns out that people want to use groups that
202 are longer than 30,000 bytes (though not repeat them that many times).
203 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
204 made it possible to implement the integer overflow checks in a much more
205 dynamic way, which I have now done. The artificial limitation on group
206 length has been removed - we now have only the limit on the total length of
207 the compiled pattern, which depends on the LINK_SIZE setting.
208
209 10. Fixed a bug in the documentation for get/copy named substring when
210 duplicate names are permitted. If none of the named substrings are set, the
211 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
212 empty string.
213
214 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
215 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
216 because the ] is interpreted as the first data character and the
217 terminating ] is not found. PCRE has been made compatible with Perl in this
218 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
219 cause memory overwriting.
220
221 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
222 string has been matched (to stop an infinite loop). It was not recognizing
223 a conditional subpattern that could match an empty string if that
224 subpattern was within another subpattern. For example, it looped when
225 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
226 condition was not nested. This bug has been fixed.
227
228 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
229 past the start of the subject in the presence of bytes with the top bit
230 set, for example "\x8aBCD".
231
232 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
233 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
234
235 14. Optimized (?!) to (*FAIL).
236
237 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
238 This restricts code points to be within the range 0 to 0x10FFFF, excluding
239 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
240 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
241 does: it's just the validity check that is more restrictive.
242
243 16. Inserted checks for integer overflows during escape sequence (backslash)
244 processing, and also fixed erroneous offset values for syntax errors during
245 backslash processing.
246
247 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
248 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
249
250 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
251 caused an overrun.
252
253 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
254 something other than just ASCII characters) inside a group that had an
255 unlimited repeat caused a loop at compile time (while checking to see
256 whether the group could match an empty string).
257
258 20. Debugging a pattern containing \p or \P could cause a crash. For example,
259 [\P{Any}] did so. (Error in the code for printing property names.)
260
261 21. An orphan \E inside a character class could cause a crash.
262
263 22. A repeated capturing bracket such as (A)? could cause a wild memory
264 reference during compilation.
265
266 23. There are several functions in pcre_compile() that scan along a compiled
267 expression for various reasons (e.g. to see if it's fixed length for look
268 behind). There were bugs in these functions when a repeated \p or \P was
269 present in the pattern. These operators have additional parameters compared
270 with \d, etc, and these were not being taken into account when moving along
271 the compiled data. Specifically:
272
273 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
274 length.
275
276 (b) An item such as \pL+ within a repeated group could cause crashes or
277 loops.
278
279 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
280 "reference to non-existent subpattern" error.
281
282 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
283
284 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
285 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
286
287 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
288
289 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
290 character were causing crashes (broken optimization).
291
292 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
293 \p or \P) caused a compile-time loop.
294
295 28. More problems have arisen in unanchored patterns when CRLF is a valid line
296 break. For example, the unstudied pattern [\r\n]A does not match the string
297 "\r\nA" because change 7.0/46 below moves the current point on by two
298 characters after failing to match at the start. However, the pattern \nA
299 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
300 the same is true. There doesn't seem any very clean way out of this, but
301 what I have chosen to do makes the common cases work: PCRE now takes note
302 of whether there can be an explicit match for \r or \n anywhere in the
303 pattern, and if so, 7.0/46 no longer applies. As part of this change,
304 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
305 pattern has explicit CR or LF references.
306
307 29. Added (*CR) etc for changing newline setting at start of pattern.
308
309
310 Version 7.2 19-Jun-07
311 ---------------------
312
313 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
314 which is apparently normally available under Windows.
315
316 2. Re-jig the pcregrep tests with different newline settings in an attempt
317 to make them independent of the local environment's newline setting.
318
319 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
320
321 4. Some of the "internals" tests were previously cut out when the link size
322 was not 2, because the output contained actual offsets. The recent new
323 "Z" feature of pcretest means that these can be cut out, making the tests
324 usable with all link sizes.
325
326 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
327 stack recursion. This gives a massive performance boost under BSD, but just
328 a small improvement under Linux. However, it saves one field in the frame
329 in all cases.
330
331 6. Added more features from the forthcoming Perl 5.10:
332
333 (a) (?-n) (where n is a string of digits) is a relative subroutine or
334 recursion call. It refers to the nth most recently opened parentheses.
335
336 (b) (?+n) is also a relative subroutine call; it refers to the nth next
337 to be opened parentheses.
338
339 (c) Conditions that refer to capturing parentheses can be specified
340 relatively, for example, (?(-2)... or (?(+3)...
341
342 (d) \K resets the start of the current match so that everything before
343 is not part of it.
344
345 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
346
347 (f) \g{name} is another synonym - part of Perl 5.10's unification of
348 reference syntax.
349
350 (g) (?| introduces a group in which the numbering of parentheses in each
351 alternative starts with the same number.
352
353 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
354
355 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
356 PCRE_INFO_JCHANGED.
357
358 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
359 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
360 for detecting groups that can match an empty string.
361
362 9. A pattern with a very large number of alternatives (more than several
363 hundred) was running out of internal workspace during the pre-compile
364 phase, where pcre_compile() figures out how much memory will be needed. A
365 bit of new cunning has reduced the workspace needed for groups with
366 alternatives. The 1000-alternative test pattern now uses 12 bytes of
367 workspace instead of running out of the 4096 that are available.
368
369 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
370
371 11. Applied patch from Google to remove an optimization that didn't quite work.
372 The report of the bug said:
373
374 pcrecpp::RE("a*").FullMatch("aaa") matches, while
375 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
376 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
377
378 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
379 it matched the wrong number of bytes.
380
381
382 Version 7.1 24-Apr-07
383 ---------------------
384
385 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
386 that is more "standard", making use of automake and other Autotools. There
387 is some re-arrangement of the files and adjustment of comments consequent
388 on this.
389
390 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
391 for recursive directory scanning broke on some systems because the files
392 are not scanned in any specific order and on different systems the order
393 was different. A call to "sort" has been inserted into RunGrepTest for the
394 approprate test as a short-term fix. In the longer term there may be an
395 alternative.
396
397 3. I had an email from Eric Raymond about problems translating some of PCRE's
398 man pages to HTML (despite the fact that I distribute HTML pages, some
399 people do their own conversions for various reasons). The problems
400 concerned the use of low-level troff macros .br and .in. I have therefore
401 removed all such uses from the man pages (some were redundant, some could
402 be replaced by .nf/.fi pairs). The 132html script that I use to generate
403 HTML has been updated to handle .nf/.fi and to complain if it encounters
404 .br or .in.
405
406 4. Updated comments in configure.ac that get placed in config.h.in and also
407 arranged for config.h to be included in the distribution, with the name
408 config.h.generic, for the benefit of those who have to compile without
409 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
410
411 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
412 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
413 makevp.bat for the current PCRE, using the additional files
414 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
415
416 6. A Windows user reported a minor discrepancy with test 2, which turned out
417 to be caused by a trailing space on an input line that had got lost in his
418 copy. The trailing space was an accident, so I've just removed it.
419
420 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
421 that is needed.
422
423 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
424 as "const" (a) because they are and (b) because it helps the PHP
425 maintainers who have recently made a script to detect big data structures
426 in the php code that should be moved to the .rodata section. I remembered
427 to update Builducptable as well, so it won't revert if ucptable.h is ever
428 re-created.
429
430 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
431 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
432 order to be able to cut out the UTF-8 tables in the latter when UTF-8
433 support is not required. This saves 1.5-2K of code, which is important in
434 some applications.
435
436 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
437 so as not to refer to the tables, even though these functions will never be
438 called when UTF-8 support is disabled. Otherwise there are problems with a
439 shared library.
440
441 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
442
443 (a) It was defining its arguments as char * instead of void *.
444
445 (b) It was assuming that all moves were upwards in memory; this was true
446 a long time ago when I wrote it, but is no longer the case.
447
448 The emulated memove() is provided for those environments that have neither
449 memmove() nor bcopy(). I didn't think anyone used it these days, but that
450 is clearly not the case, as these two bugs were recently reported.
451
452 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
453 and Detrail to create the HTML documentation, the .txt form of the man
454 pages, and it removes trailing spaces from listed files. It also creates
455 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
456 case, it wraps all the #defines with #ifndefs. This script should be run
457 before "make dist".
458
459 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
460 with Unicode property support.
461
462 (a) For a maximizing quantifier, if the two different cases of the
463 character were of different lengths in their UTF-8 codings (there are
464 some cases like this - I found 11), and the matching function had to
465 back up over a mixture of the two cases, it incorrectly assumed they
466 were both the same length.
467
468 (b) When PCRE was configured to use the heap rather than the stack for
469 recursion during matching, it was not correctly preserving the data for
470 the other case of a UTF-8 character when checking ahead for a match
471 while processing a minimizing repeat. If the check also involved
472 matching a wide character, but failed, corruption could cause an
473 erroneous result when trying to check for a repeat of the original
474 character.
475
476 13. Some tidying changes to the testing mechanism:
477
478 (a) The RunTest script now detects the internal link size and whether there
479 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
480 values substituted by "configure". (The RunGrepTest script already did
481 this for UTF-8.) The configure.ac script no longer substitutes the
482 relevant variables.
483
484 (b) The debugging options /B and /D in pcretest show the compiled bytecode
485 with length and offset values. This means that the output is different
486 for different internal link sizes. Test 2 is skipped for link sizes
487 other than 2 because of this, bypassing the problem. Unfortunately,
488 there was also a test in test 3 (the locale tests) that used /B and
489 failed for link sizes other than 2. Rather than cut the whole test out,
490 I have added a new /Z option to pcretest that replaces the length and
491 offset values with spaces. This is now used to make test 3 independent
492 of link size. (Test 2 will be tidied up later.)
493
494 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
495 segmentation fault instead of returning the appropriate error message.
496
497 15. In multiline mode when the newline sequence was set to "any", the pattern
498 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
499 This doesn't seem right; it now treats the CRLF combination as the line
500 ending, and so does not match in that case. It's only a pattern such as ^$
501 that would hit this one: something like ^ABC$ would have failed after \r
502 and then tried again after \r\n.
503
504 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
505 in an attempt to make files that differ only in their line terminators
506 compare equal. This works on Linux.
507
508 17. Under certain error circumstances pcregrep might try to free random memory
509 as it exited. This is now fixed, thanks to valgrind.
510
511 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
512 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
513 was because its rules for how to advance for /g after matching an empty
514 string at the end of a line did not allow for this case. They now check for
515 it specially.
516
517 20. pcretest is supposed to handle patterns and data of any length, by
518 extending its buffers when necessary. It was getting this wrong when the
519 buffer for a data line had to be extended.
520
521 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
522 CRLF as a newline sequence.
523
524 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
525 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
526 I have nevertheless tidied it up.
527
528 23. Added some casts to kill warnings from HP-UX ia64 compiler.
529
530 24. Added a man page for pcre-config.
531
532
533 Version 7.0 19-Dec-06
534 ---------------------
535
536 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
537 moving to gcc 4.1.1.
538
539 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
540 sys/time.h, which is documented as needed for this function. It doesn't
541 seem to matter on Linux, but it showed up on some releases of OS X.
542
543 3. It seems that there are systems where bytes whose values are greater than
544 127 match isprint() in the "C" locale. The "C" locale should be the
545 default when a C program starts up. In most systems, only ASCII printing
546 characters match isprint(). This difference caused the output from pcretest
547 to vary, making some of the tests fail. I have changed pcretest so that:
548
549 (a) When it is outputting text in the compiled version of a pattern, bytes
550 other than 32-126 are always shown as hex escapes.
551
552 (b) When it is outputting text that is a matched part of a subject string,
553 it does the same, unless a different locale has been set for the match
554 (using the /L modifier). In this case, it uses isprint() to decide.
555
556 4. Fixed a major bug that caused incorrect computation of the amount of memory
557 required for a compiled pattern when options that changed within the
558 pattern affected the logic of the preliminary scan that determines the
559 length. The relevant options are -x, and -i in UTF-8 mode. The result was
560 that the computed length was too small. The symptoms of this bug were
561 either the PCRE error "internal error: code overflow" from pcre_compile(),
562 or a glibc crash with a message such as "pcretest: free(): invalid next
563 size (fast)". Examples of patterns that provoked this bug (shown in
564 pcretest format) are:
565
566 /(?-x: )/x
567 /(?x)(?-x: \s*#\s*)/
568 /((?i)[\x{c0}])/8
569 /(?i:[\x{c0}])/8
570
571 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
572 is now done differently.
573
574 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
575 wrapper classes; (b) implement a new function in the C++ scanner that is
576 more efficient than the old way of doing things because it avoids levels of
577 recursion in the regex matching; (c) add a paragraph to the documentation
578 for the FullMatch() function.
579
580 6. The escape sequence \n was being treated as whatever was defined as
581 "newline". Not only was this contrary to the documentation, which states
582 that \n is character 10 (hex 0A), but it also went horribly wrong when
583 "newline" was defined as CRLF. This has been fixed.
584
585 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
586 was being set to -1 for the "end of line" case (supposedly a value that no
587 character can have). Though this value is never used (the check for end of
588 line is "zero bytes in current character"), it caused compiler complaints.
589 I've changed it to 0xffffffff.
590
591 8. In pcre_version.c, the version string was being built by a sequence of
592 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
593 string (as it is for production releases) called a macro with an empty
594 argument. The C standard says the result of this is undefined. The gcc
595 compiler treats it as an empty string (which was what was wanted) but it is
596 reported that Visual C gives an error. The source has been hacked around to
597 avoid this problem.
598
599 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
600 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
601 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
602 of them did).
603
604 10. Originally, pcretest opened its input and output without "b"; then I was
605 told that "b" was needed in some environments, so it was added for release
606 5.0 to both the input and output. (It makes no difference on Unix-like
607 systems.) Later I was told that it is wrong for the input on Windows. I've
608 now abstracted the modes into two macros, to make it easier to fiddle with
609 them, and removed "b" from the input mode under Windows.
610
611 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
612
613 12. Added -help and --help to pcretest as an official way of being reminded
614 of the options.
615
616 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
617 and pcrecpp.cc because they annoy compilers at high warning levels.
618
619 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
620
621 15. Fixed an occurrence of == in configure.ac that should have been = (shell
622 scripts are not C programs :-) and which was not noticed because it works
623 on Linux.
624
625 16. pcretest is supposed to handle any length of pattern and data line (as one
626 line or as a continued sequence of lines) by extending its input buffer if
627 necessary. This feature was broken for very long pattern lines, leading to
628 a string of junk being passed to pcre_compile() if the pattern was longer
629 than about 50K.
630
631 17. I have done a major re-factoring of the way pcre_compile() computes the
632 amount of memory needed for a compiled pattern. Previously, there was code
633 that made a preliminary scan of the pattern in order to do this. That was
634 OK when PCRE was new, but as the facilities have expanded, it has become
635 harder and harder to keep it in step with the real compile phase, and there
636 have been a number of bugs (see for example, 4 above). I have now found a
637 cunning way of running the real compile function in a "fake" mode that
638 enables it to compute how much memory it would need, while actually only
639 ever using a few hundred bytes of working memory and without too many
640 tests of the mode. This should make future maintenance and development
641 easier. A side effect of this work is that the limit of 200 on the nesting
642 depth of parentheses has been removed (though this was never a serious
643 limitation, I suspect). However, there is a downside: pcre_compile() now
644 runs more slowly than before (30% or more, depending on the pattern). I
645 hope this isn't a big issue. There is no effect on runtime performance.
646
647 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
648 newline (only possible for the last line of a file) and it was a
649 pattern that set a locale (followed by /Lsomething), pcretest crashed.
650
651 19. Added additional timing features to pcretest. (1) The -tm option now times
652 matching only, not compiling. (2) Both -t and -tm can be followed, as a
653 separate command line item, by a number that specifies the number of
654 repeats to use when timing. The default is 50000; this gives better
655 precision, but takes uncomfortably long for very large patterns.
656
657 20. Extended pcre_study() to be more clever in cases where a branch of a
658 subpattern has no definite first character. For example, (a*|b*)[cd] would
659 previously give no result from pcre_study(). Now it recognizes that the
660 first character must be a, b, c, or d.
661
662 21. There was an incorrect error "recursive call could loop indefinitely" if
663 a subpattern (or the entire pattern) that was being tested for matching an
664 empty string contained only one non-empty item after a nested subpattern.
665 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
666 incorrectly, because the \d was being skipped in the check.
667
668 22. The pcretest program now has a new pattern option /B and a command line
669 option -b, which is equivalent to adding /B to every pattern. This causes
670 it to show the compiled bytecode, without the additional information that
671 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
672 is the same as /B/I).
673
674 23. A new optimization is now able automatically to treat some sequences such
675 as a*b as a*+b. More specifically, if something simple (such as a character
676 or a simple class like \d) has an unlimited quantifier, and is followed by
677 something that cannot possibly match the quantified thing, the quantifier
678 is automatically "possessified".
679
680 24. A recursive reference to a subpattern whose number was greater than 39
681 went wrong under certain circumstances in UTF-8 mode. This bug could also
682 have affected the operation of pcre_study().
683
684 25. Realized that a little bit of performance could be had by replacing
685 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
686
687 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
688
689 27. Possessive quantifiers such as a++ were previously implemented by turning
690 them into atomic groups such as ($>a+). Now they have their own opcodes,
691 which improves performance. This includes the automatically created ones
692 from 23 above.
693
694 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
695 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
696 the first matched character to be a colon. This applied both to named and
697 numbered groups.
698
699 29. The ucpinternal.h header file was missing its idempotency #ifdef.
700
701 30. I was sent a "project" file called libpcre.a.dev which I understand makes
702 building PCRE on Windows easier, so I have included it in the distribution.
703
704 31. There is now a check in pcretest against a ridiculously large number being
705 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
706 loop, the loop is abandoned.
707
708 32. Forward references to subpatterns in conditions such as (?(2)...) where
709 subpattern 2 is defined later cause pcre_compile() to search forwards in
710 the pattern for the relevant set of parentheses. This search went wrong
711 when there were unescaped parentheses in a character class, parentheses
712 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
713
714 33. "Subroutine" calls and backreferences were previously restricted to
715 referencing subpatterns earlier in the regex. This restriction has now
716 been removed.
717
718 34. Added a number of extra features that are going to be in Perl 5.10. On the
719 whole, these are just syntactic alternatives for features that PCRE had
720 previously implemented using the Python syntax or my own invention. The
721 other formats are all retained for compatibility.
722
723 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
724 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
725 also .NET compatible.
726
727 (b) A recursion or subroutine call to a named group can now be defined as
728 (?&name) as well as (?P>name).
729
730 (c) A backreference to a named group can now be defined as \k<name> or
731 \k'name' as well as (?P=name). The new forms, as well as being in Perl
732 5.10, are also .NET compatible.
733
734 (d) A conditional reference to a named group can now use the syntax
735 (?(<name>) or (?('name') as well as (?(name).
736
737 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
738 groups (named and numbered) that are never evaluated inline, but can be
739 called as "subroutines" from elsewhere. In effect, the DEFINE condition
740 is always false. There may be only one alternative in such a group.
741
742 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
743 as the simple (?(R). The condition is true only if the most recent
744 recursion is that of the given number or name. It does not search out
745 through the entire recursion stack.
746
747 (g) The escape \gN or \g{N} has been added, where N is a positive or
748 negative number, specifying an absolute or relative reference.
749
750 35. Tidied to get rid of some further signed/unsigned compiler warnings and
751 some "unreachable code" warnings.
752
753 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
754 things, this adds five new scripts.
755
756 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
757 There were also incompatibilities regarding the handling of \Q..\E inside
758 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
759 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
760
761 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
762 matches an empty string, and forcibly breaks the loop. There were bugs in
763 this code in non-simple cases. For a pattern such as ^(a()*)* matched
764 against aaaa the result was just "a" rather than "aaaa", for example. Two
765 separate and independent bugs (that affected different cases) have been
766 fixed.
767
768 39. Refactored the code to abolish the use of different opcodes for small
769 capturing bracket numbers. This is a tidy that I avoided doing when I
770 removed the limit on the number of capturing brackets for 3.5 back in 2001.
771 The new approach is not only tidier, it makes it possible to reduce the
772 memory needed to fix the previous bug (38).
773
774 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
775 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
776 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
777 mode.
778
779 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
780 report.
781
782 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
783 copy construction and assignment in the C++ wrapper.
784
785 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
786 couple of bugs that could have given wrong results in the "--newline=crlf"
787 case.
788
789 44. Added a number of casts and did some reorganization of signed/unsigned int
790 variables following suggestions from Dair Grant. Also renamed the variable
791 "this" as "item" because it is a C++ keyword.
792
793 45. Arranged for dftables to add
794
795 #include "pcre_internal.h"
796
797 to pcre_chartables.c because without it, gcc 4.x may remove the array
798 definition from the final binary if PCRE is built into a static library and
799 dead code stripping is activated.
800
801 46. For an unanchored pattern, if a match attempt fails at the start of a
802 newline sequence, and the newline setting is CRLF or ANY, and the next two
803 characters are CRLF, advance by two characters instead of one.
804
805
806 Version 6.7 04-Jul-06
807 ---------------------
808
809 1. In order to handle tests when input lines are enormously long, pcretest has
810 been re-factored so that it automatically extends its buffers when
811 necessary. The code is crude, but this _is_ just a test program. The
812 default size has been increased from 32K to 50K.
813
814 2. The code in pcre_study() was using the value of the re argument before
815 testing it for NULL. (Of course, in any sensible call of the function, it
816 won't be NULL.)
817
818 3. The memmove() emulation function in pcre_internal.h, which is used on
819 systems that lack both memmove() and bcopy() - that is, hardly ever -
820 was missing a "static" storage class specifier.
821
822 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
823 containing an extended class (one that cannot be represented by a bitmap
824 because it contains high-valued characters or Unicode property items, e.g.
825 [\pZ]). Almost always one would set UTF-8 mode when processing such a
826 pattern, but PCRE should not loop if you do not (it no longer does).
827 [Detail: two cases were found: (a) a repeated subpattern containing an
828 extended class; (b) a recursive reference to a subpattern that followed a
829 previous extended class. It wasn't skipping over the extended class
830 correctly when UTF-8 mode was not set.]
831
832 5. A negated single-character class was not being recognized as fixed-length
833 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
834 compile error "lookbehind assertion is not fixed length".
835
836 6. The RunPerlTest auxiliary script was showing an unexpected difference
837 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
838 write a Perl script that can interpret lines of an input file either as
839 byte characters or as UTF-8, which is what "perltest" was being required to
840 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
841 can't do is switch easily at run time between having the "use utf8;" pragma
842 or not. In the end, I fudged it by using the RunPerlTest script to insert
843 "use utf8;" explicitly for the UTF-8 tests.
844
845 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
846 the end of the subject string, contrary to the documentation and to what
847 Perl does. This was true of both matching functions. Now it matches only at
848 the start of the subject and immediately after *internal* newlines.
849
850 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
851 a pointer to an int instead of a pointer to an unsigned long int. This
852 caused problems on 64-bit systems.
853
854 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
855 instance of the 'standard' template library not being so standard".
856
857 10. There was no check on the number of named subpatterns nor the maximum
858 length of a subpattern name. The product of these values is used to compute
859 the size of the memory block for a compiled pattern. By supplying a very
860 long subpattern name and a large number of named subpatterns, the size
861 computation could be caused to overflow. This is now prevented by limiting
862 the length of names to 32 characters, and the number of named subpatterns
863 to 10,000.
864
865 11. Subpatterns that are repeated with specific counts have to be replicated in
866 the compiled pattern. The size of memory for this was computed from the
867 length of the subpattern and the repeat count. The latter is limited to
868 65535, but there was no limit on the former, meaning that integer overflow
869 could in principle occur. The compiled length of a repeated subpattern is
870 now limited to 30,000 bytes in order to prevent this.
871
872 12. Added the optional facility to have named substrings with the same name.
873
874 13. Added the ability to use a named substring as a condition, using the
875 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
876 are numbers (not recommended). Forward references are permitted.
877
878 14. Added forward references in named backreferences (if you see what I mean).
879
880 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
881 pattern could run off the end of the subject. For example, the pattern
882 "(?s)(.{1,5})"8 did this with the subject "ab".
883
884 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
885 PCRE_CASELESS was set when matching characters that were quantified with ?
886 or *.
887
888 17. A character class other than a single negated character that had a minimum
889 but no maximum quantifier - for example [ab]{6,} - was not handled
890 correctly by pce_dfa_exec(). It would match only one character.
891
892 18. A valid (though odd) pattern that looked like a POSIX character
893 class but used an invalid character after [ (for example [[,abc,]]) caused
894 pcre_compile() to give the error "Failed: internal error: code overflow" or
895 in some cases to crash with a glibc free() error. This could even happen if
896 the pattern terminated after [[ but there just happened to be a sequence of
897 letters, a binary zero, and a closing ] in the memory that followed.
898
899 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
900 over the years. Originally (before any Unicode support), just the bottom 8
901 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
902 output from "man perlunicode" includes this:
903
904 The regular expression compiler produces polymorphic opcodes. That
905 is, the pattern adapts to the data and automatically switches to
906 the Unicode character scheme when presented with Unicode data--or
907 instead uses a traditional byte scheme when presented with byte
908 data.
909
910 Sadly, a wide octal escape does not cause a switch, and in a string with
911 no other multibyte characters, these octal escapes are treated as before.
912 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
913 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
914 Unicode string.
915
916 I have not perpetrated such confusion in PCRE. Up till now, it took just
917 the bottom 8 bits, as in old Perl. I have now made octal escapes with
918 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
919 translate to the appropriate multibyte character.
920
921 29. Applied some refactoring to reduce the number of warnings from Microsoft
922 and Borland compilers. This has included removing the fudge introduced
923 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
924 a warning about an unused variable.
925
926 21. PCRE has not included VT (character 0x0b) in the set of whitespace
927 characters since release 4.0, because Perl (from release 5.004) does not.
928 [Or at least, is documented not to: some releases seem to be in conflict
929 with the documentation.] However, when a pattern was studied with
930 pcre_study() and all its branches started with \s, PCRE still included VT
931 as a possible starting character. Of course, this did no harm; it just
932 caused an unnecessary match attempt.
933
934 22. Removed a now-redundant internal flag bit that recorded the fact that case
935 dependency changed within the pattern. This was once needed for "required
936 byte" processing, but is no longer used. This recovers a now-scarce options
937 bit. Also moved the least significant internal flag bit to the most-
938 significant bit of the word, which was not previously used (hangover from
939 the days when it was an int rather than a uint) to free up another bit for
940 the future.
941
942 23. Added support for CRLF line endings as well as CR and LF. As well as the
943 default being selectable at build time, it can now be changed at runtime
944 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
945 specify that it is scanning data with non-default line endings.
946
947 24. Changed the definition of CXXLINK to make it agree with the definition of
948 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
949
950 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
951 recursions. This makes a big different to stack usage for some patterns.
952
953 26. If a subpattern containing a named recursion or subroutine reference such
954 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
955 the space required for the compiled pattern went wrong and gave too small a
956 value. Depending on the environment, this could lead to "Failed: internal
957 error: code overflow at offset 49" or "glibc detected double free or
958 corruption" errors.
959
960 27. Applied patches from Google (a) to support the new newline modes and (b) to
961 advance over multibyte UTF-8 characters in GlobalReplace.
962
963 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
964 difference for some implementation of PCRE in some Windows version.
965
966 29. Added some extra testing facilities to pcretest:
967
968 \q<number> in a data line sets the "match limit" value
969 \Q<number> in a data line sets the "match recursion limt" value
970 -S <number> sets the stack size, where <number> is in megabytes
971
972 The -S option isn't available for Windows.
973
974
975 Version 6.6 06-Feb-06
976 ---------------------
977
978 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
979 in pcreposix.h. I have copied the definition from pcre.h.
980
981 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
982 because pcre.h is no longer a built file.
983
984 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
985 not normally included in the compiled code.
986
987
988 Version 6.5 01-Feb-06
989 ---------------------
990
991 1. When using the partial match feature with pcre_dfa_exec(), it was not
992 anchoring the second and subsequent partial matches at the new starting
993 point. This could lead to incorrect results. For example, with the pattern
994 /1234/, partially matching against "123" and then "a4" gave a match.
995
996 2. Changes to pcregrep:
997
998 (a) All non-match returns from pcre_exec() were being treated as failures
999 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1000 error message is output. Some extra information is given for the
1001 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1002 probably the only errors that are likely to be caused by users (by
1003 specifying a regex that has nested indefinite repeats, for instance).
1004 If there are more than 20 of these errors, pcregrep is abandoned.
1005
1006 (b) A binary zero was treated as data while matching, but terminated the
1007 output line if it was written out. This has been fixed: binary zeroes
1008 are now no different to any other data bytes.
1009
1010 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1011 used to set a locale for matching. The --locale=xxxx long option has
1012 been added (no short equivalent) to specify a locale explicitly on the
1013 pcregrep command, overriding the environment variables.
1014
1015 (d) When -B was used with -n, some line numbers in the output were one less
1016 than they should have been.
1017
1018 (e) Added the -o (--only-matching) option.
1019
1020 (f) If -A or -C was used with -c (count only), some lines of context were
1021 accidentally printed for the final match.
1022
1023 (g) Added the -H (--with-filename) option.
1024
1025 (h) The combination of options -rh failed to suppress file names for files
1026 that were found from directory arguments.
1027
1028 (i) Added the -D (--devices) and -d (--directories) options.
1029
1030 (j) Added the -F (--fixed-strings) option.
1031
1032 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1033
1034 (l) Added the --colo(u)r option.
1035
1036 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1037 is not present by default.
1038
1039 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1040 items such as (?R) or (?1), when the recursion could match a number of
1041 alternatives. If it matched one of the alternatives, but subsequently,
1042 outside the recursion, there was a failure, the code tried to back up into
1043 the recursion. However, because of the way PCRE is implemented, this is not
1044 possible, and the result was an incorrect result from the match.
1045
1046 In order to prevent this happening, the specification of recursion has
1047 been changed so that all such subpatterns are automatically treated as
1048 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1049
1050 4. I had overlooked the fact that, in some locales, there are characters for
1051 which isalpha() is true but neither isupper() nor islower() are true. In
1052 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1053 and ordfeminine) are like this. This affected the treatment of \w and \W
1054 when they appeared in character classes, but not when they appeared outside
1055 a character class. The bit map for "word" characters is now created
1056 separately from the results of isalnum() instead of just taking it from the
1057 upper, lower, and digit maps. (Plus the underscore character, of course.)
1058
1059 5. The above bug also affected the handling of POSIX character classes such as
1060 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1061 permanent tables. Instead, the bit maps for such a class were previously
1062 created as the appropriate unions of the upper, lower, and digit bitmaps.
1063 Now they are created by subtraction from the [[:word:]] class, which has
1064 its own bitmap.
1065
1066 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1067 It is created by subtracting the vertical space characters (\x09, \x0a,
1068 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1069 subtraction was done in the overall bitmap for a character class, meaning
1070 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1071 be recognized. This bug has been fixed.
1072
1073 7. Patches from the folks at Google:
1074
1075 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1076 real life, but is still worth protecting against".
1077
1078 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1079 regular expressions".
1080
1081 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1082 have it.
1083
1084 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1085 "configure" and the latter not, in order to fix a problem somebody had
1086 with compiling the Arg class on HP-UX.
1087
1088 (e) Improve the error-handling of the C++ wrapper a little bit.
1089
1090 (f) New tests for checking recursion limiting.
1091
1092 8. The pcre_memmove() function, which is used only if the environment does not
1093 have a standard memmove() function (and is therefore rarely compiled),
1094 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1095 returning a result (though PCRE never actually uses the result).
1096
1097 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1098 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1099 returned instead of calling malloc() with an overflowing number that would
1100 most likely cause subsequent chaos.
1101
1102 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1103
1104 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1105 with this option is matched, the nmatch and pmatch options of regexec() are
1106 ignored.
1107
1108 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1109 provided in case anyone wants to the the POSIX interface with UTF-8
1110 strings.
1111
1112 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1113 C++ linking (needed for some HP-UX environments).
1114
1115 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1116 (unused parameter) and in the pcre_printint() function (omitted "default"
1117 switch label when the default is to do nothing).
1118
1119 15. Added some code to make it possible, when PCRE is compiled as a C++
1120 library, to replace subject pointers for pcre_exec() with a smart pointer
1121 class, thus making it possible to process discontinuous strings.
1122
1123 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1124 much the same function. They were added by different people who were trying
1125 to make PCRE easy to compile on non-Unix systems. It has been suggested
1126 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1127 for compiling on Windows systems. I have therefore replaced it with
1128 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1129 defaults to "extern" for C or "extern C" for C++, which works fine on
1130 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1131 SCOPE with something explicit in config.h. In addition:
1132
1133 (a) pcreposix.h still had just "extern" instead of either of these macros;
1134 I have replaced it with PCRE_DATA_SCOPE.
1135
1136 (b) Functions such as _pcre_xclass(), which are internal to the library,
1137 but external in the C sense, all had PCRE_EXPORT in their definitions.
1138 This is apparently wrong for the Windows case, so I have removed it.
1139 (It makes no difference on Unix-like systems.)
1140
1141 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1142 of recursive calls to match(). This is different to MATCH_LIMIT because
1143 that limits the total number of calls to match(), not all of which increase
1144 the depth of recursion. Limiting the recursion depth limits the amount of
1145 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1146 when PCRE is compiled, and changed at run time. A patch from Google adds
1147 this functionality to the C++ interface.
1148
1149 18. Changes to the handling of Unicode character properties:
1150
1151 (a) Updated the table to Unicode 4.1.0.
1152
1153 (b) Recognize characters that are not in the table as "Cn" (undefined).
1154
1155 (c) I revised the way the table is implemented to a much improved format
1156 which includes recognition of ranges. It now supports the ranges that
1157 are defined in UnicodeData.txt, and it also amalgamates other
1158 characters into ranges. This has reduced the number of entries in the
1159 table from around 16,000 to around 3,000, thus reducing its size
1160 considerably. I realized I did not need to use a tree structure after
1161 all - a binary chop search is just as efficient. Having reduced the
1162 number of entries, I extended their size from 6 bytes to 8 bytes to
1163 allow for more data.
1164
1165 (d) Added support for Unicode script names via properties such as \p{Han}.
1166
1167 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1168 matching that character.
1169
1170 20. When matching a repeated Unicode property with a minimum greater than zero,
1171 (for example \pL{2,}), PCRE could look past the end of the subject if it
1172 reached it while seeking the minimum number of characters. This could
1173 happen only if some of the characters were more than one byte long, because
1174 there is a check for at least the minimum number of bytes.
1175
1176 21. Refactored the implementation of \p and \P so as to be more general, to
1177 allow for more different types of property in future. This has changed the
1178 compiled form incompatibly. Anybody with saved compiled patterns that use
1179 \p or \P will have to recompile them.
1180
1181 22. Added "Any" and "L&" to the supported property types.
1182
1183 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1184 but give a compile time error if the value is greater than 0xff.
1185
1186 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1187 accidentally not being installed or uninstalled.
1188
1189 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1190 made were to insert the current release number. This seemed silly, because
1191 it made things harder for people building PCRE on systems that don't run
1192 "configure". I have turned pcre.h into a distributed file, no longer built
1193 by "configure", with the version identification directly included. There is
1194 no longer a pcre.h.in file.
1195
1196 However, this change necessitated a change to the pcre-config script as
1197 well. It is built from pcre-config.in, and one of the substitutions was the
1198 release number. I have updated configure.ac so that ./configure now finds
1199 the release number by grepping pcre.h.
1200
1201 26. Added the ability to run the tests under valgrind.
1202
1203
1204 Version 6.4 05-Sep-05
1205 ---------------------
1206
1207 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1208 "--" to be printed when multiple files were scanned, even when none of the
1209 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1210 consider it to be a bug, and have restored the previous behaviour.
1211
1212 2. A couple of code tidies to get rid of compiler warnings.
1213
1214 3. The pcretest program used to cheat by referring to symbols in the library
1215 whose names begin with _pcre_. These are internal symbols that are not
1216 really supposed to be visible externally, and in some environments it is
1217 possible to suppress them. The cheating is now confined to including
1218 certain files from the library's source, which is a bit cleaner.
1219
1220 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1221 file's purpose clearer.
1222
1223 5. Reorganized pcre_ucp_findchar().
1224
1225
1226 Version 6.3 15-Aug-05
1227 ---------------------
1228
1229 1. The file libpcre.pc.in did not have general read permission in the tarball.
1230
1231 2. There were some problems when building without C++ support:
1232
1233 (a) If C++ support was not built, "make install" and "make test" still
1234 tried to test it.
1235
1236 (b) There were problems when the value of CXX was explicitly set. Some
1237 changes have been made to try to fix these, and ...
1238
1239 (c) --disable-cpp can now be used to explicitly disable C++ support.
1240
1241 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1242 backslash in a target when C++ was disabled. This confuses some
1243 versions of "make", apparently. Using an intermediate variable solves
1244 this. (Same for CPP_LOBJ.)
1245
1246 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1247 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1248 necessary on certain architectures.
1249
1250 4. Added a setting of -export-symbols-regex to the link command to remove
1251 those symbols that are exported in the C sense, but actually are local
1252 within the library, and not documented. Their names all begin with
1253 "_pcre_". This is not a perfect job, because (a) we have to except some
1254 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1255 available (and never for static libraries). I have made a note to try to
1256 find a way round (a) in the future.
1257
1258
1259 Version 6.2 01-Aug-05
1260 ---------------------
1261
1262 1. There was no test for integer overflow of quantifier values. A construction
1263 such as {1111111111111111} would give undefined results. What is worse, if
1264 a minimum quantifier for a parenthesized subpattern overflowed and became
1265 negative, the calculation of the memory size went wrong. This could have
1266 led to memory overwriting.
1267
1268 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1269
1270 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1271 operating environments where this matters.
1272
1273 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1274 PCRE options from within the C++ wrapper.
1275
1276 5. Named capturing subpatterns were not being correctly counted when a pattern
1277 was compiled. This caused two problems: (a) If there were more than 100
1278 such subpatterns, the calculation of the memory needed for the whole
1279 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1280 back references of the form \12, where the number was greater than 9, were
1281 not recognized as back references, even though there were sufficient
1282 previous subpatterns.
1283
1284 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1285 versions of gcc, e.g. 2.95.4.
1286
1287
1288 Version 6.1 21-Jun-05
1289 ---------------------
1290
1291 1. There was one reference to the variable "posix" in pcretest.c that was not
1292 surrounded by "#if !defined NOPOSIX".
1293
1294 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1295 the cross-check on the old pcre_info() function, for the benefit of the
1296 cut-down version of PCRE that is currently imported into Exim.
1297
1298 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1299 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1300 patterns that start (?i) but I don't think that matters. The use of (?i) is
1301 just an example; this all applies to the other options as well.
1302
1303 4. Since libtool seems to echo the compile commands it is issuing, the output
1304 from "make" can be reduced a bit by putting "@" in front of each libtool
1305 compile command.
1306
1307 5. Patch from the folks at Google for configure.in to be a bit more thorough
1308 in checking for a suitable C++ installation before trying to compile the
1309 C++ stuff. This should fix a reported problem when a compiler was present,
1310 but no suitable headers.
1311
1312 6. The man pages all had just "PCRE" as their title. I have changed them to
1313 be the relevant file name. I have also arranged that these names are
1314 retained in the file doc/pcre.txt, which is a concatenation in text format
1315 of all the man pages except the little individual ones for each function.
1316
1317 7. The NON-UNIX-USE file had not been updated for the different set of source
1318 files that come with release 6. I also added a few comments about the C++
1319 wrapper.
1320
1321
1322 Version 6.0 07-Jun-05
1323 ---------------------
1324
1325 1. Some minor internal re-organization to help with my DFA experiments.
1326
1327 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1328 didn't matter for the library itself when fully configured, but did matter
1329 when compiling without UCP support, or within Exim, where the ucp files are
1330 not imported.
1331
1332 3. Refactoring of the library code to split up the various functions into
1333 different source modules. The addition of the new DFA matching code (see
1334 below) to a single monolithic source would have made it really too
1335 unwieldy, quite apart from causing all the code to be include in a
1336 statically linked application, when only some functions are used. This is
1337 relevant even without the DFA addition now that patterns can be compiled in
1338 one application and matched in another.
1339
1340 The downside of splitting up is that there have to be some external
1341 functions and data tables that are used internally in different modules of
1342 the library but which are not part of the API. These have all had their
1343 names changed to start with "_pcre_" so that they are unlikely to clash
1344 with other external names.
1345
1346 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1347 a different (DFA) algorithm. Although it is slower than the original
1348 function, it does have some advantages for certain types of matching
1349 problem.
1350
1351 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1352 including restarting after a partial match.
1353
1354 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1355 defined when compiling for Windows was sent to me. I have put it into the
1356 code, though I have no means of testing or verifying it.
1357
1358 7. Added the pcre_refcount() auxiliary function.
1359
1360 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1361 match before or at the first newline in the subject string. In pcretest,
1362 the /f option on a pattern can be used to set this.
1363
1364 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1365 would behave wrongly. This has been present in PCRE since release 4.0.
1366
1367 10. A number of changes to the pcregrep command:
1368
1369 (a) Refactored how -x works; insert ^(...)$ instead of setting
1370 PCRE_ANCHORED and checking the length, in preparation for adding
1371 something similar for -w.
1372
1373 (b) Added the -w (match as a word) option.
1374
1375 (c) Refactored the way lines are read and buffered so as to have more
1376 than one at a time available.
1377
1378 (d) Implemented a pcregrep test script.
1379
1380 (e) Added the -M (multiline match) option. This allows patterns to match
1381 over several lines of the subject. The buffering ensures that at least
1382 8K, or the rest of the document (whichever is the shorter) is available
1383 for matching (and similarly the previous 8K for lookbehind assertions).
1384
1385 (f) Changed the --help output so that it now says
1386
1387 -w, --word-regex(p)
1388
1389 instead of two lines, one with "regex" and the other with "regexp"
1390 because that confused at least one person since the short forms are the
1391 same. (This required a bit of code, as the output is generated
1392 automatically from a table. It wasn't just a text change.)
1393
1394 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1395 option but starts with a hyphen. Could be a pattern or a path name
1396 starting with a hyphen, for instance.
1397
1398 (h) "-" can be given as a file name to represent stdin.
1399
1400 (i) When file names are being printed, "(standard input)" is used for
1401 the standard input, for compatibility with GNU grep. Previously
1402 "<stdin>" was used.
1403
1404 (j) The option --label=xxx can be used to supply a name to be used for
1405 stdin when file names are being printed. There is no short form.
1406
1407 (k) Re-factored the options decoding logic because we are going to add
1408 two more options that take data. Such options can now be given in four
1409 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1410
1411 (l) Added the -A, -B, and -C options for requesting that lines of context
1412 around matches be printed.
1413
1414 (m) Added the -L option to print the names of files that do not contain
1415 any matching lines, that is, the complement of -l.
1416
1417 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1418 continue to scan other files.
1419
1420 (o) The -s option was incorrectly implemented. For compatibility with other
1421 greps, it now suppresses the error message for a non-existent or non-
1422 accessible file (but not the return code). There is a new option called
1423 -q that suppresses the output of matching lines, which was what -s was
1424 previously doing.
1425
1426 (p) Added --include and --exclude options to specify files for inclusion
1427 and exclusion when recursing.
1428
1429 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1430 Hopefully, it now does.
1431
1432 12. Missing cast in pcre_study().
1433
1434 13. Added an "uninstall" target to the makefile.
1435
1436 14. Replaced "extern" in the function prototypes in Makefile.in with
1437 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1438 world, but is set differently for Windows.
1439
1440 15. Added a second compiling function called pcre_compile2(). The only
1441 difference is that it has an extra argument, which is a pointer to an
1442 integer error code. When there is a compile-time failure, this is set
1443 non-zero, in addition to the error test pointer being set to point to an
1444 error message. The new argument may be NULL if no error number is required
1445 (but then you may as well call pcre_compile(), which is now just a
1446 wrapper). This facility is provided because some applications need a
1447 numeric error indication, but it has also enabled me to tidy up the way
1448 compile-time errors are handled in the POSIX wrapper.
1449
1450 16. Added VPATH=.libs to the makefile; this should help when building with one
1451 prefix path and installing with another. (Or so I'm told by someone who
1452 knows more about this stuff than I do.)
1453
1454 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1455 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1456 match everything, including newlines. This is not POSIX-compatible, but
1457 somebody wanted the feature. From pcretest it can be activated by using
1458 both the P and the s flags.
1459
1460 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1461
1462 19. libpcre.pc was being incorrectly installed as executable.
1463
1464 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1465 it now also looks for '\r' so that it will work unmodified on Windows.
1466
1467 21. Added Google's contributed C++ wrapper to the distribution.
1468
1469 22. Added some untidy missing memory free() calls in pcretest, to keep
1470 Electric Fence happy when testing.
1471
1472
1473
1474 Version 5.0 13-Sep-04
1475 ---------------------
1476
1477 1. Internal change: literal characters are no longer packed up into items
1478 containing multiple characters in a single byte-string. Each character
1479 is now matched using a separate opcode. However, there may be more than one
1480 byte in the character in UTF-8 mode.
1481
1482 2. The pcre_callout_block structure has two new fields: pattern_position and
1483 next_item_length. These contain the offset in the pattern to the next match
1484 item, and its length, respectively.
1485
1486 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1487 insertion of callouts before each pattern item. Added the /C option to
1488 pcretest to make use of this.
1489
1490 4. On the advice of a Windows user, the lines
1491
1492 #if defined(_WIN32) || defined(WIN32)
1493 _setmode( _fileno( stdout ), 0x8000 );
1494 #endif /* defined(_WIN32) || defined(WIN32) */
1495
1496 have been added to the source of pcretest. This apparently does useful
1497 magic in relation to line terminators.
1498
1499 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1500 for the benefit of those environments where the "b" makes a difference.
1501
1502 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1503 to know about it. I have put a hack into configure.in that adds in code
1504 to set GCC=yes if CC=icc. This seems to end up at a point in the
1505 generated configure script that is early enough to affect the setting of
1506 compiler options, which is what is needed, but I have no means of testing
1507 whether it really works. (The user who reported this had patched the
1508 generated configure script, which of course I cannot do.)
1509
1510 LATER: After change 22 below (new libtool files), the configure script
1511 seems to know about icc (and also ecc). Therefore, I have commented out
1512 this hack in configure.in.
1513
1514 7. Added support for pkg-config (2 patches were sent in).
1515
1516 8. Negated POSIX character classes that used a combination of internal tables
1517 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1518 [[:^ascii]]. Typically, they would match almost any characters. The other
1519 POSIX classes were not broken in this way.
1520
1521 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1522 to find the match, as PCRE was deluded into thinking that the match had to
1523 start at the start point or following a newline. The same bug applied to
1524 patterns with negative forward assertions or any backward assertions
1525 preceding ".*" at the start, unless the pattern required a fixed first
1526 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1527
1528 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1529 starting at the last subject character, bytes beyond the end of the subject
1530 string were read.
1531
1532 11. Renamed the variable "class" as "classbits" to make life easier for C++
1533 users. (Previously there was a macro definition, but it apparently wasn't
1534 enough.)
1535
1536 12. Added the new field "tables" to the extra data so that tables can be passed
1537 in at exec time, or the internal tables can be re-selected. This allows
1538 a compiled regex to be saved and re-used at a later time by a different
1539 program that might have everything at different addresses.
1540
1541 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1542 -R library as well as a -L library.
1543
1544 14. The debugging options of pcretest (-d on the command line or D on a
1545 pattern) showed incorrect output for anything following an extended class
1546 that contained multibyte characters and which was followed by a quantifier.
1547
1548 15. Added optional support for general category Unicode character properties
1549 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1550 support. It adds about 90K to the size of the library. The meanings of the
1551 inbuilt class escapes such as \d and \s have NOT been changed.
1552
1553 16. Updated pcredemo.c to include calls to free() to release the memory for the
1554 compiled pattern.
1555
1556 17. The generated file chartables.c was being created in the source directory
1557 instead of in the building directory. This caused the build to fail if the
1558 source directory was different from the building directory, and was
1559 read-only.
1560
1561 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1562 file. No doubt somebody will tell me if they don't make sense... Also added
1563 Dan Mooney's comments about building on OpenVMS.
1564
1565 19. Added support for partial matching via the PCRE_PARTIAL option for
1566 pcre_exec() and the \P data escape in pcretest.
1567
1568 20. Extended pcretest with 3 new pattern features:
1569
1570 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1571 write the compiled pattern to the file whose name is "rest-of-line".
1572 This is a straight binary dump of the data, with the saved pointer to
1573 the character tables forced to be NULL. The study data, if any, is
1574 written too. After writing, pcretest reads a new pattern.
1575
1576 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1577 compiled pattern from the given file. There must not be any
1578 occurrences of "<" in the file name (pretty unlikely); if there are,
1579 pcretest will instead treat the initial "<" as a pattern delimiter.
1580 After reading in the pattern, pcretest goes on to read data lines as
1581 usual.
1582
1583 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1584 and 16-bit fields in a compiled pattern, to simulate a pattern that
1585 was compiled on a host of opposite endianness.
1586
1587 21. The pcre-exec() function can now cope with patterns that were compiled on
1588 hosts of opposite endianness, with this restriction:
1589
1590 As for any compiled expression that is saved and used later, the tables
1591 pointer field cannot be preserved; the extra_data field in the arguments
1592 to pcre_exec() should be used to pass in a tables address if a value
1593 other than the default internal tables were used at compile time.
1594
1595 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1596 now diagnosed as an error. Previously, most of the time, a negative number
1597 would have been treated as zero, but if in addition "ovector" was passed as
1598 NULL, a crash could occur.
1599
1600 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1601 new versions from the libtool 1.5 distribution (the last one is a copy of
1602 a file called libtool.m4). This seems to have fixed the need to patch
1603 "configure" to support Darwin 1.3 (which I used to do). However, I still
1604 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1605 workstation).
1606
1607 24. Changed the PCRE licence to be the more standard "BSD" licence.
1608
1609
1610 Version 4.5 01-Dec-03
1611 ---------------------
1612
1613 1. There has been some re-arrangement of the code for the match() function so
1614 that it can be compiled in a version that does not call itself recursively.
1615 Instead, it keeps those local variables that need separate instances for
1616 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1617 needs to "recurse". Keeping track of where control must go is done by means
1618 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1619 hide most of the details from the main code, and operates only if
1620 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1621 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1622 operating.
1623
1624 To make it easier for callers to provide specially tailored get/free
1625 functions for this usage, two new functions, pcre_stack_malloc, and
1626 pcre_stack_free, are used. They are always called in strict stacking order,
1627 and the size of block requested is always the same.
1628
1629 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1630 PCRE has been compiled to use the stack or the heap for recursion. The
1631 -C option of pcretest uses this to show which version is compiled.
1632
1633 A new data escape \S, is added to pcretest; it causes the amounts of store
1634 obtained and freed by both kinds of malloc/free at match time to be added
1635 to the output.
1636
1637 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1638 what's available on my current Linux desktop machine.
1639
1640 3. When matching a UTF-8 string, the test for a valid string at the start has
1641 been extended. If start_offset is not zero, PCRE now checks that it points
1642 to a byte that is the start of a UTF-8 character. If not, it returns
1643 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1644 this is necessary because there may be backward assertions in the pattern.
1645 When matching the same subject several times, it may save resources to use
1646 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1647
1648 4. The code for checking the validity of UTF-8 strings has been tightened so
1649 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1650 containing "overlong sequences".
1651
1652 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1653 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1654 should have been "&", but it just so happened that all the cases this let
1655 through by mistake were picked up later in the function.
1656
1657 6. I had used a variable called "isblank" - this is a C99 function, causing
1658 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1659
1660 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1661 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1662 so the version is shown; (c) stop "make test" echoing "./RunTest".
1663
1664 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1665
1666 9. The prototype for memmove() for systems that don't have it was using
1667 size_t, but the inclusion of the header that defines size_t was later. I've
1668 moved the #includes for the C headers earlier to avoid this.
1669
1670 10. Added some adjustments to the code to make it easier to compiler on certain
1671 special systems:
1672
1673 (a) Some "const" qualifiers were missing.
1674 (b) Added the macro EXPORT before all exported functions; by default this
1675 is defined to be empty.
1676 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1677 that it reads its output file name as an argument instead of writing
1678 to the standard output and assuming this can be redirected.
1679
1680 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1681 class containing characters with values greater than 255, PCRE compilation
1682 went into a loop.
1683
1684 12. A recursive reference to a subpattern that was within another subpattern
1685 that had a minimum quantifier of zero caused PCRE to crash. For example,
1686 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1687 recursion. If the recursively-called subpattern itself had a zero repeat,
1688 that was OK.
1689
1690 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1691 buffer into which it was copied (for escape processing) was still set at
1692 1024, so long lines caused crashes.
1693
1694 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1695 "internal error: code overflow...". This applied to any character class
1696 that was followed by a possessive quantifier.
1697
1698 15. Modified the Makefile to add libpcre.la as a prerequisite for
1699 libpcreposix.la because I was told this is needed for a parallel build to
1700 work.
1701
1702 16. If a pattern that contained .* following optional items at the start was
1703 studied, the wrong optimizing data was generated, leading to matching
1704 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1705 matching string must start with a or b or c. The correct conclusion for
1706 this pattern is that a match can start with any character.
1707
1708
1709 Version 4.4 13-Aug-03
1710 ---------------------
1711
1712 1. In UTF-8 mode, a character class containing characters with values between
1713 127 and 255 was not handled correctly if the compiled pattern was studied.
1714 In fixing this, I have also improved the studying algorithm for such
1715 classes (slightly).
1716
1717 2. Three internal functions had redundant arguments passed to them. Removal
1718 might give a very teeny performance improvement.
1719
1720 3. Documentation bug: the value of the capture_top field in a callout is *one
1721 more than* the number of the hightest numbered captured substring.
1722
1723 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1724 in incorrectly linking with a previously installed version. They now link
1725 explicitly with libpcre.la.
1726
1727 5. configure.in no longer needs to recognize Cygwin specially.
1728
1729 6. A problem in pcre.in for Windows platforms is fixed.
1730
1731 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1732 pcretest, it used to include the size of the study block as part of its
1733 output. Unfortunately, the structure contains a field that has a different
1734 size on different hardware architectures. This meant that the tests that
1735 showed this size failed. As the block is currently always of a fixed size,
1736 this information isn't actually particularly useful in pcretest output, so
1737 I have just removed it.
1738
1739 8. Three pre-processor statements accidentally did not start in column 1.
1740 Sadly, there are *still* compilers around that complain, even though
1741 standard C has not required this for well over a decade. Sigh.
1742
1743 9. In pcretest, the code for checking callouts passed small integers in the
1744 callout_data field, which is a void * field. However, some picky compilers
1745 complained about the casts involved for this on 64-bit systems. Now
1746 pcretest passes the address of the small integer instead, which should get
1747 rid of the warnings.
1748
1749 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1750 both compile and run time, and gives an error if an invalid UTF-8 sequence
1751 is found. There is a option for disabling this check in cases where the
1752 string is known to be correct and/or the maximum performance is wanted.
1753
1754 11. In response to a bug report, I changed one line in Makefile.in from
1755
1756 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1757 to
1758 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1759
1760 to look similar to other lines, but I have no way of telling whether this
1761 is the right thing to do, as I do not use Windows. No doubt I'll get told
1762 if it's wrong...
1763
1764
1765 Version 4.3 21-May-03
1766 ---------------------
1767
1768 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1769 Makefile.
1770
1771 2. Some refactoring to improve the quality of the code:
1772
1773 (i) The utf8_table... variables are now declared "const".
1774
1775 (ii) The code for \cx, which used the "case flipping" table to upper case
1776 lower case letters, now just substracts 32. This is ASCII-specific,
1777 but the whole concept of \cx is ASCII-specific, so it seems
1778 reasonable.
1779
1780 (iii) PCRE was using its character types table to recognize decimal and
1781 hexadecimal digits in the pattern. This is silly, because it handles
1782 only 0-9, a-f, and A-F, but the character types table is locale-
1783 specific, which means strange things might happen. A private
1784 table is now used for this - though it costs 256 bytes, a table is
1785 much faster than multiple explicit tests. Of course, the standard
1786 character types table is still used for matching digits in subject
1787 strings against \d.
1788
1789 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1790 ending in _t are). So I've renamed it as ESC_tee.
1791
1792 3. The first argument for regexec() in the POSIX wrapper should have been
1793 defined as "const".
1794
1795 4. Changed pcretest to use malloc() for its buffers so that they can be
1796 Electric Fenced for debugging.
1797
1798 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1799 to read one or more bytes before the start of the subject string. Often this
1800 had no effect on PCRE's behaviour, but in some circumstances it could
1801 provoke a segmentation fault.
1802
1803 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1804 to try to read one or more bytes before the start of the subject string.
1805
1806 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1807 UTF-8 support could misbehave in various ways if the subject string
1808 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1809 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1810 back over UTF-8 characters.)
1811
1812
1813 Version 4.2 14-Apr-03
1814 ---------------------
1815
1816 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1817
1818 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1819 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1820 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1821 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1822 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1823 and BUILD_EXEEXT
1824 Note: automatic setting of the BUILD variables is not yet working
1825 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1826 compile-time but not at link-time
1827 [LINK]: use for linking executables only
1828 make different versions for Windows and non-Windows
1829 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1830 libraries
1831 [LINK_FOR_BUILD]: new variable
1832 [OBJEXT]: use throughout
1833 [EXEEXT]: use throughout
1834 <winshared>: new target
1835 <wininstall>: new target
1836 <dftables.o>: use native compiler
1837 <dftables>: use native linker
1838 <install>: handle Windows platform correctly
1839 <clean>: ditto
1840 <check>: ditto
1841 copy DLL to top builddir before testing
1842
1843 As part of these changes, -no-undefined was removed again. This was reported
1844 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1845 in any case.
1846
1847 3. Some tidies to get rid of compiler warnings:
1848
1849 . In the match_data structure, match_limit was an unsigned long int, whereas
1850 match_call_count was an int. I've made them both unsigned long ints.
1851
1852 . In pcretest the fact that a const uschar * doesn't automatically cast to
1853 a void * provoked a warning.
1854
1855 . Turning on some more compiler warnings threw up some "shadow" variables
1856 and a few more missing casts.
1857
1858 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1859 option, a class that contained a single character with a value between 128
1860 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1861
1862 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1863 option, a class that contained several characters, but with at least one
1864 whose value was between 128 and 255 caused PCRE to crash.
1865
1866
1867 Version 4.1 12-Mar-03
1868 ---------------------
1869
1870 1. Compiling with gcc -pedantic found a couple of places where casts were
1871 needed, and a string in dftables.c that was longer than standard compilers are
1872 required to support.
1873
1874 2. Compiling with Sun's compiler found a few more places where the code could
1875 be tidied up in order to avoid warnings.
1876
1877 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1878 first of these names is deprecated in the latest Autoconf in favour of the name
1879 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1880 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1881 analogy I have changed it to CFLAGS_FOR_BUILD.
1882
1883 4. Added -no-undefined to the linking command in the Makefile, because this is
1884 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1885 linking step for the pcreposix library.
1886
1887 5. PCRE was failing to diagnose the case of two named groups with the same
1888 name.
1889
1890 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1891 literal character that is needed in the subject for a match, and scans along to
1892 ensure that it is present before embarking on the full matching process. This
1893 saves time in cases of nested unlimited repeats that are never going to match.
1894 Problem: the scan can take a lot of time if the subject is very long (e.g.
1895 megabytes), thus penalizing straightforward matches. It is now done only if the
1896 amount of subject to be scanned is less than 1000 bytes.
1897
1898 7. A lesser problem with the same optimization is that it was recording the
1899 first character of an anchored pattern as "needed", thus provoking a search
1900 right along the subject, even when the first match of the pattern was going to
1901 fail. The "needed" character is now not set for anchored patterns, unless it
1902 follows something in the pattern that is of non-fixed length. Thus, it still
1903 fulfils its original purpose of finding quick non-matches in cases of nested
1904 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1905
1906
1907 Version 4.0 17-Feb-03
1908 ---------------------
1909
1910 1. If a comment in an extended regex that started immediately after a meta-item
1911 extended to the end of string, PCRE compiled incorrect data. This could lead to
1912 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1913
1914 2. Moved to autoconf 2.53 and libtool 1.4.2.
1915
1916 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1917 the special perltest8 script is no longer needed - all the tests can be run
1918 from a single perltest script.
1919
1920 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1921 by \s. It has now been removed in PCRE. This means it isn't recognized as
1922 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1923 class [:space:] *does* include VT, thereby creating a mess.
1924
1925 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1926 space and tab.
1927
1928 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1929 its new features into the main test script, reducing the number of scripts.
1930
1931 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1932 were backward compatible, and made the (?i) apply to the whole pattern, as if
1933 /i were given. Now it behaves more logically, and applies the option setting
1934 only to what follows. PCRE has been changed to follow suit. However, if it
1935 finds options settings right at the start of the pattern, it extracts them into
1936 the global options, as before. Thus, they show up in the info data.
1937
1938 8. Added support for the \Q...\E escape sequence. Characters in between are
1939 treated as literals. This is slightly different from Perl in that $ and @ are
1940 also handled as literals inside the quotes. In Perl, they will cause variable
1941 interpolation. Note the following examples:
1942
1943 Pattern PCRE matches Perl matches
1944
1945 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1946 \Qabc\$xyz\E abc\$xyz abc\$xyz
1947 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1948
1949 For compatibility with Perl, \Q...\E sequences are recognized inside character
1950 classes as well as outside them.
1951
1952 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1953 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1954 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1955 signed/unsigned warnings.
1956
1957 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1958 option for pcretest, so I've replaced it by a simple function that does just
1959 that job.
1960
1961 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1962 "pcregrep -".
1963
1964 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1965 Java package. This provides some syntactic sugar for simple cases of what my
1966 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1967 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1968 item, you can use this simplified notation. Note that only makes sense with
1969 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1970 greediness, whatever the setting of the PCRE_UNGREEDY option.
1971
1972 13. A change of greediness default within a pattern was not taking effect at
1973 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1974 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1975 was abstracted outside.
1976
1977 14. PCRE now supports the \G assertion. It is true when the current matching
1978 position is at the start point of the match. This differs from \A when the
1979 starting offset is non-zero. Used with the /g option of pcretest (or similar
1980 code), it works in the same way as it does for Perl's /g option. If all
1981 alternatives of a regex begin with \G, the expression is anchored to the start
1982 match position, and the "anchored" flag is set in the compiled expression.
1983
1984 15. Some bugs concerning the handling of certain option changes within patterns
1985 have been fixed. These applied to options other than (?ims). For example,
1986 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1987 been the other way round. Some of this was related to change 7 above.
1988
1989 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1990 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1991 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1992 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1993
1994 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1995 mode. Unlike ".", it always matches newline, whatever the setting of
1996 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1997 assertions. Perl allows it, but it doesn't (in general) work because it can't
1998 calculate the length of the lookbehind. At least, that's the case for Perl
1999 5.8.0 - I've been told they are going to document that it doesn't work in
2000 future.
2001
2002 18. Added an error diagnosis for escapes that PCRE does not support: these are
2003 \L, \l, \N, \P, \p, \U, \u, and \X.
2004
2005 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2006 reading past the end of the pattern in cases such as /[abcd/.
2007
2008 20. PCRE was getting more memory than necessary for patterns with classes that
2009 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2010
2011 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2012 compiling PCRE for use with Virtual Pascal.
2013
2014 22. Small fix to the Makefile to make it work properly if the build is done
2015 outside the source tree.
2016
2017 23. Added a new extension: a condition to go with recursion. If a conditional
2018 subpattern starts with (?(R) the "true" branch is used if recursion has
2019 happened, whereas the "false" branch is used only at the top level.
2020
2021 24. When there was a very long string of literal characters (over 255 bytes
2022 without UTF support, over 250 bytes with UTF support), the computation of how
2023 much memory was required could be incorrect, leading to segfaults or other
2024 strange effects.
2025
2026 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2027 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2028 there was a subsequent back reference to those brackets. This meant that, for
2029 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2030 possible to check for precisely this case. All we can do is abandon the
2031 optimization if .* occurs inside capturing brackets when there are any back
2032 references whatsoever. (See below for a better fix that came later.)
2033
2034 26. The handling of the optimization for finding the first character of a
2035 non-anchored pattern, and for finding a character that is required later in the
2036 match were failing in some cases. This didn't break the matching; it just
2037 failed to optimize when it could. The way this is done has been re-implemented.
2038
2039 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2040
2041 28. Added a new feature that provides some of the functionality that Perl
2042 provides with (?{...}). The facility is termed a "callout". The way it is done
2043 in PCRE is for the caller to provide an optional function, by setting
2044 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2045 global variable. By default it is unset, which disables all calling out. To get
2046 the function called, the regex must include (?C) at appropriate points. This
2047 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2048 This provides a means of identifying different callout points. When PCRE
2049 reaches such a point in the regex, if pcre_callout has been set, the external
2050 function is called. It is provided with data in a structure called
2051 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2052 matching continues; if it returns a non-zero value, the match at the current
2053 point fails. However, backtracking will occur if possible. [This was changed
2054 later and other features added - see item 49 below.]
2055
2056 29. pcretest is upgraded to test the callout functionality. It provides a
2057 callout function that displays information. By default, it shows the start of
2058 the match and the current position in the text. There are some new data escapes
2059 to vary what happens:
2060
2061 \C+ in addition, show current contents of captured substrings
2062 \C- do not supply a callout function
2063 \C!n return 1 when callout number n is reached
2064 \C!n!m return 1 when callout number n is reached for the mth time
2065
2066 30. If pcregrep was called with the -l option and just a single file name, it
2067 output "<stdin>" if a match was found, instead of the file name.
2068
2069 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2070 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2071 pcre_exec(). This saves a malloc/free per call. The default value of
2072 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2073 when configuring.
2074
2075 32. The default maximum size of a compiled pattern is 64K. There have been a
2076 few cases of people hitting this limit. The code now uses macros to handle the
2077 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2078 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2079 configuring. Tests 2 and 5 work only with 2-byte links because they output
2080 debugging information about compiled patterns.
2081
2082 33. Internal code re-arrangements:
2083
2084 (a) Moved the debugging function for printing out a compiled regex into
2085 its own source file (printint.c) and used #include to pull it into
2086 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2087 separate copies.
2088
2089 (b) Defined the list of op-code names for debugging as a macro in
2090 internal.h so that it is next to the definition of the opcodes.
2091
2092 (c) Defined a table of op-code lengths for simpler skipping along compiled
2093 code. This is again a macro in internal.h so that it is next to the
2094 definition of the opcodes.
2095
2096 34. Added support for recursive calls to individual subpatterns, along the
2097 lines of Robin Houston's patch (but implemented somewhat differently).
2098
2099 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2100 allow it to read and process whole directories in Win32. This code was
2101 contributed by Lionel Fourquaux; it has not been tested by me.
2102
2103 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2104 used to name a group. Names consist of alphanumerics and underscores, and must
2105 be unique. Back references use the syntax (?P=name) and recursive calls use
2106 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2107 numbers. The function pcre_fullinfo() can be used after compilation to extract
2108 a name/number map. There are three relevant calls:
2109
2110 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2111 PCRE_INFO_NAMECOUNT yields the number of entries
2112 PCRE_INFO_NAMETABLE yields a pointer to the map.
2113
2114 The map is a vector of fixed-size entries. The size of each entry depends on
2115 the length of the longest name used. The first two bytes of each entry are the
2116 group number, most significant byte first. There follows the corresponding
2117 name, zero terminated. The names are in alphabetical order.
2118
2119 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2120 case instead of 255. Making it the same both with and without UTF-8 support
2121 means that the same test output works with both.
2122
2123 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2124 calling malloc() with a zero argument.
2125
2126 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2127 optimization. I've improved things by keeping a bitmap of backreferences with
2128 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2129 fact referenced, the optimization can be applied. It is unlikely that a
2130 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2131 the match to follow \n) will appear inside brackets with a number greater than
2132 31, but if it does, any back reference > 31 suppresses the optimization.
2133
2134 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2135 of disabling numbered capturing parentheses. Any opening parenthesis that is
2136 not followed by ? behaves as if it were followed by ?: but named parentheses
2137 can still be used for capturing (and they will acquire numbers in the usual
2138 way).
2139
2140 41. Redesigned the return codes from the match() function into yes/no/error so
2141 that errors can be passed back from deep inside the nested calls. A malloc
2142 failure while inside a recursive subpattern call now causes the
2143 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2144
2145 42. It is now possible to set a limit on the number of times the match()
2146 function is called in a call to pcre_exec(). This facility makes it possible to
2147 limit the amount of recursion and backtracking, though not in a directly
2148 obvious way, because the match() function is used in a number of different
2149 circumstances. The count starts from zero for each position in the subject
2150 string (for non-anchored patterns). The default limit is, for compatibility, a
2151 large number, namely 10 000 000. You can change this in two ways:
2152
2153 (a) When configuring PCRE before making, you can use --with-match-limit=n
2154 to set a default value for the compiled library.
2155
2156 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2157 a different value is set. See 45 below.
2158
2159 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2160
2161 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2162 of things that can be changed at compile time. The first argument specifies
2163 what is wanted and the second points to where the information is to be placed.
2164 The current list of available information is:
2165
2166 PCRE_CONFIG_UTF8
2167
2168 The output is an integer that is set to one if UTF-8 support is available;
2169 otherwise it is set to zero.
2170
2171 PCRE_CONFIG_NEWLINE
2172
2173 The output is an integer that it set to the value of the code that is used for
2174 newline. It is either LF (10) or CR (13).
2175
2176 PCRE_CONFIG_LINK_SIZE
2177
2178 The output is an integer that contains the number of bytes used for internal
2179 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2180
2181 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2182
2183 The output is an integer that contains the threshold above which the POSIX
2184 interface uses malloc() for output vectors. See item 31 above.
2185
2186 PCRE_CONFIG_MATCH_LIMIT
2187
2188 The output is an unsigned integer that contains the default limit of the number
2189 of match() calls in a pcre_exec() execution. See 42 above.
2190
2191 44. pcretest has been upgraded by the addition of the -C option. This causes it
2192 to extract all the available output from the new pcre_config() function, and to
2193 output it. The program then exits immediately.
2194
2195 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2196 order to support additional features. One way would have been to define
2197 pcre_exec2() (for example) with extra arguments, but this would not have been
2198 extensible, and would also have required all calls to the original function to
2199 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2200 is used for passing in "extra" data from pcre_study().
2201
2202 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2203 contains the following fields:
2204
2205 flags a bitmap indicating which of the following fields are set
2206 study_data opaque data from pcre_study()
2207 match_limit a way of specifying a limit on match() calls for a specific
2208 call to pcre_exec()
2209 callout_data data for callouts (see 49 below)
2210
2211 The flag bits are also defined in pcre.h, and are
2212
2213 PCRE_EXTRA_STUDY_DATA
2214 PCRE_EXTRA_MATCH_LIMIT
2215 PCRE_EXTRA_CALLOUT_DATA
2216
2217 The pcre_study() function now returns one of these new pcre_extra blocks, with
2218 the actual study data pointed to by the study_data field, and the
2219 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2220 before. That is, this change is entirely upwards-compatible and requires no
2221 change to existing code.
2222
2223 If you want to pass in additional data to pcre_exec(), you can either place it
2224 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2225 block.
2226
2227 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2228 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2229 times with different match limits, until it finds the minimum value needed for
2230 pcre_exec() to complete. The value is then output. This can be instructive; for
2231 most simple matches the number is quite small, but for pathological cases it
2232 gets very large very quickly.
2233
2234 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2235 returns the size of the data block pointed to by the study_data field in a
2236 pcre_extra block, that is, the value that was passed as the argument to
2237 pcre_malloc() when PCRE was getting memory in which to place the information
2238 created by pcre_study(). The fourth argument should point to a size_t variable.
2239 pcretest has been extended so that this information is shown after a successful
2240 pcre_study() call when information about the compiled regex is being displayed.
2241
2242 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2243 because what follows is always an absolute path. (Later: it turns out that this
2244 is more than cosmetic for MinGW, because it doesn't like empty path
2245 components.)
2246
2247 49. Some changes have been made to the callout feature (see 28 above):
2248
2249 (i) A callout function now has three choices for what it returns:
2250
2251 0 => success, carry on matching
2252 > 0 => failure at this point, but backtrack if possible
2253 < 0 => serious error, return this value from pcre_exec()
2254
2255 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2256 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2257 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2258 use by callout functions. It will never be used by PCRE itself.
2259
2260 (ii) The pcre_extra structure (see 45 above) has a void * field called
2261 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2262 pcre_callout_block structure has a field of the same name. The contents of
2263 the field passed in the pcre_extra structure are passed to the callout
2264 function in the corresponding field in the callout block. This makes it
2265 easier to use the same callout-containing regex from multiple threads. For
2266 testing, the pcretest program has a new data escape
2267
2268 \C*n pass the number n (may be negative) as callout_data
2269
2270 If the callout function in pcretest receives a non-zero value as
2271 callout_data, it returns that value.
2272
2273 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2274 there were some redundant $(CFLAGS) in commands that are now specified as
2275 $(LINK), which already includes $(CFLAGS).
2276
2277 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2278 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2279 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2280 one-byte characters throughout. Note that case-insensitive matching applies
2281 only to characters whose values are less than 256. PCRE doesn't support the
2282 notion of cases for higher-valued characters.
2283
2284 (i) A character class whose characters are all within 0-255 is handled as
2285 a bit map, and the map is inverted for negative classes. Previously, a
2286 character > 255 always failed to match such a class; however it should
2287 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2288
2289 (ii) A negated character class with a single character < 255 is coded as
2290 "not this character" (OP_NOT). This wasn't working properly when the test
2291 character was multibyte, either singly or repeated.
2292
2293 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2294 mode, for example: \x{100}{2,3}.
2295
2296 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2297 singly or repeated) now correctly test multibyte characters. However,
2298 PCRE doesn't recognize any characters with values greater than 255 as
2299 digits, spaces, or word characters. Such characters always match \D, \S,
2300 and \W, and never match \d, \s, or \w.
2301
2302 (v) Classes may now contain characters and character ranges with values
2303 greater than 255. For example: [ab\x{100}-\x{400}].
2304
2305 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2306 PCRE in UTF-8 mode.
2307
2308 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2309 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2310 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2311 value.)
2312
2313 53. The single man page has become too large. I have therefore split it up into
2314 a number of separate man pages. These also give rise to individual HTML pages;
2315 these are now put in a separate directory, and there is an index.html page that
2316 lists them all. Some hyperlinking between the pages has been installed.
2317
2318 54. Added convenience functions for handling named capturing parentheses.
2319
2320 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2321 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2322 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2323 are faulted.
2324
2325 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2326 calling configure. These values are used when compiling the dftables.c program
2327 which is run to generate the source of the default character tables. They
2328 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2329 you will need to set these values.
2330
2331 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2332
2333
2334 Version 3.9 02-Jan-02
2335 ---------------------
2336
2337 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2338
2339 2. If --disable-static was given, the building process failed when trying to
2340 build pcretest and pcregrep. (For some reason it was using libtool to compile
2341 them, which is not right, as they aren't part of the library.)
2342
2343
2344 Version 3.8 18-Dec-01
2345 ---------------------
2346
2347 1. The experimental UTF-8 code was completely screwed up. It was packing the
2348 bytes in the wrong order. How dumb can you get?
2349
2350
2351 Version 3.7 29-Oct-01
2352 ---------------------
2353
2354 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2355 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2356 this didn't happen under Solaris 8, where I normally test things.
2357
2358 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2359 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2360 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2361 this fix an hour or so after the initial 3.7 release.)
2362
2363
2364 Version 3.6 23-Oct-01
2365 ---------------------
2366
2367 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2368 offsets passed as NULL with zero offset count.
2369
2370 2. The config.guess and config.sub files had not been updated when I moved to
2371 the latest autoconf.
2372
2373
2374 Version 3.5 15-Aug-01
2375 ---------------------
2376
2377 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2378 had been forgotten.
2379
2380 2. By using declared but undefined structures, we can avoid using "void"
2381 definitions in pcre.h while keeping the internal definitions of the structures
2382 private.
2383
2384 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2385 user point of view, this means that both static and shared libraries are built
2386 by default, but this can be individually controlled. More of the work of
2387 handling this static/shared cases is now inside libtool instead of PCRE's make
2388 file.
2389
2390 4. The pcretest utility is now installed along with pcregrep because it is
2391 useful for users (to test regexs) and by doing this, it automatically gets
2392 relinked by libtool. The documentation has been turned into a man page, so
2393 there are now .1, .txt, and .html versions in /doc.
2394
2395 5. Upgrades to pcregrep:
2396 (i) Added long-form option names like gnu grep.
2397 (ii) Added --help to list all options with an explanatory phrase.
2398 (iii) Added -r, --recursive to recurse into sub-directories.
2399 (iv) Added -f, --file to read patterns from a file.
2400
2401 6. pcre_exec() was referring to its "code" argument before testing that
2402 argument for NULL (and giving an error if it was NULL).
2403
2404 7. Upgraded Makefile.in to allow for compiling in a different directory from
2405 the source directory.
2406
2407 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2408 options bits, the pointer it was passed was to an int instead of to an unsigned
2409 long int. This mattered only on 64-bit systems.
2410
2411 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2412 generated) instead of pcre.in, which it its source. Also made the same change
2413 in several of the .c files.
2414
2415 10. A new release of gcc defines printf() as a macro, which broke pcretest
2416 because it had an ifdef in the middle of a string argument for printf(). Fixed
2417 by using separate calls to printf().
2418
2419 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2420 script, to force use of CR or LF instead of \n in the source. On non-Unix
2421 systems, the value can be set in config.h.
2422
2423 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2424 absolute limit. Changed the text of the error message to make this clear, and
2425 likewise updated the man page.
2426
2427 13. The limit of 99 on the number of capturing subpatterns has been removed.
2428 The new limit is 65535, which I hope will not be a "real" limit.
2429
2430
2431 Version 3.4 22-Aug-00
2432 ---------------------
2433
2434 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2435
2436 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2437
2438
2439 Version 3.3 01-Aug-00
2440 ---------------------
2441
2442 1. If an octal character was given, but the value was greater than \377, it
2443 was not getting masked to the least significant bits, as documented. This could
2444 lead to crashes in some systems.
2445
2446 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2447 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2448
2449 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2450 These just pass their arguments on to (pcre_free)(), but they are provided
2451 because some uses of PCRE bind it to non-C systems that can call its functions,
2452 but cannot call free() or pcre_free() directly.
2453
2454 4. Add "make test" as a synonym for "make check". Corrected some comments in
2455 the Makefile.
2456
2457 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2458 Makefile.
2459
2460 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2461 command called pgrep for grepping around the active processes.
2462
2463 7. Added the beginnings of support for UTF-8 character strings.
2464
2465 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2466 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2467 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2468 out for the ar command.)
2469
2470
2471 Version 3.2 12-May-00
2472 ---------------------
2473
2474 This is purely a bug fixing release.
2475
2476 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2477 of ZA. This was just one example of several cases that could provoke this bug,
2478 which was introduced by change 9 of version 2.00. The code for breaking
2479 infinite loops after an iteration that matches an empty string was't working
2480 correctly.
2481
2482 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2483 when matched against abbab (for example). After matching an empty string, it
2484 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2485 caused it to match further down the string than it should.
2486
2487 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2488 was there because it doesn't seem to be needed, and it causes trouble on some
2489 systems, as it is not a Standard C header. It has been removed.
2490
2491 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2492 were reported on the Macintosh. The changes were from
2493
2494 while ((c = *(++ptr)) != 0 && c != '\n');
2495 to
2496 while ((c = *(++ptr)) != 0 && c != '\n') ;
2497
2498 Totally extraordinary, but if that's what it takes...
2499
2500 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2501 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2502 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2503 assumes the way PCRE uses memmove() (always moving upwards).
2504
2505 6. PCRE is being used in one environment where strchr() is not available. There
2506 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2507 faster code anyway.
2508
2509
2510 Version 3.1 09-Feb-00
2511 ---------------------
2512
2513 The only change in this release is the fixing of some bugs in Makefile.in for
2514 the "install" target:
2515
2516 (1) It was failing to install pcreposix.h.
2517
2518 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2519
2520
2521 Version 3.0 01-Feb-00
2522 ---------------------
2523
2524 1. Add support for the /+ modifier to perltest (to output $` like it does in
2525 pcretest).
2526
2527 2. Add support for the /g modifier to perltest.
2528
2529 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2530 matches null strings.
2531
2532 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2533 pattern. Perl treats empty patterns specially - it reuses the most recent
2534 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2535 effect.
2536
2537 5. The POSIX interface was broken in that it was just handing over the POSIX
2538 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2539 required a bigger vector, with some working space on the end. This means that
2540 the POSIX wrapper now has to get and free some memory, and copy the results.
2541
2542 6. Added some simple autoconf support, placing the test data and the
2543 documentation in separate directories, re-organizing some of the
2544 information files, and making it build pcre-config (a GNU standard). Also added
2545 libtool support for building PCRE as a shared library, which is now the
2546 default.
2547
2548 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2549 09 are not valid octal constants. Single digits will be used for minor values
2550 less than 10.
2551
2552 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2553 existing programs that set these in the POSIX interface can use PCRE without
2554 modification.
2555
2556 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2557 return all that pcre_info() returns, plus additional data. The pcre_info()
2558 function is retained for compatibility, but is considered to be obsolete.
2559
2560 10. Added experimental recursion feature (?R) to handle one common case that
2561 Perl 5.6 will be able to do with (?p{...}).
2562
2563 11. Added support for POSIX character classes like [:alpha:], which Perl is
2564 adopting.
2565
2566
2567 Version 2.08 31-Aug-99
2568 ----------------------
2569
2570 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2571 trying to match at the startoffset position, but instead was moving forward to
2572 the next newline as if a previous match had failed.
2573
2574 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2575 and could get into a loop if a null string was matched other than at the start
2576 of the subject.
2577
2578 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2579 be distinguished at compile time, and for completeness also added PCRE_DATE.
2580
2581 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2582 in GnuWin32 environments.
2583
2584
2585 Version 2.07 29-Jul-99
2586 ----------------------
2587
2588 1. The documentation is now supplied in plain text form and HTML as well as in
2589 the form of man page sources.
2590
2591 2. C++ compilers don't like assigning (void *) values to other pointer types.
2592 In particular this affects malloc(). Although there is no problem in Standard
2593 C, I've put in casts to keep C++ compilers happy.
2594
2595 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2596 should be (const char *).
2597
2598 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2599 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2600 However, I haven't made this a standard facility. The documentation doesn't
2601 mention it, and the Makefile doesn't support it.
2602
2603 5. The Makefile now contains an "install" target, with editable destinations at
2604 the top of the file. The pcretest program is not installed.
2605
2606 6. pgrep -V now gives the PCRE version number and date.
2607
2608 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2609 causing the entire string to be ignored, instead of just the last character.
2610
2611 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2612 non-matching string, it can take a very, very long time, even for strings of
2613 quite modest length, because of the nested recursion. PCRE now does better in
2614 some of these cases. It does this by remembering the last required literal
2615 character in the pattern, and pre-searching the subject to ensure it is present
2616 before running the real match. In other words, it applies a heuristic to detect
2617 some types of certain failure quickly, and in the above example, if presented
2618 with a string that has no trailing " it gives "no match" very quickly.
2619
2620 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2621 other alternatives are tried instead.
2622
2623
2624 Version 2.06 09-Jun-99
2625 ----------------------
2626
2627 1. Change pcretest's output for amount of store used to show just the code
2628 space, because the remainder (the data block) varies in size between 32-bit and
2629 64-bit systems.
2630
2631 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2632 start matching at. This allows lookbehinds to work when searching for multiple
2633 occurrences in a string.
2634
2635 3. Added additional options to pcretest for testing multiple occurrences:
2636
2637 /+ outputs the rest of the string that follows a match
2638 /g loops for multiple occurrences, using the new startoffset argument
2639 /G loops for multiple occurrences by passing an incremented pointer
2640
2641 4. PCRE wasn't doing the "first character" optimization for patterns starting
2642 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2643 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2644 the letter 'x'. On long subject strings, this gives a significant speed-up.
2645
2646
2647 Version 2.05 21-Apr-99
2648 ----------------------
2649
2650 1. Changed the type of magic_number from int to long int so that it works
2651 properly on 16-bit systems.
2652
2653 2. Fixed a bug which caused patterns starting with .* not to work correctly
2654 when the subject string contained newline characters. PCRE was assuming
2655 anchoring for such patterns in all cases, which is not correct because .* will
2656 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2657 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2658 must be retried after every newline in the subject.
2659
2660
2661 Version 2.04 18-Feb-99
2662 ----------------------
2663
2664 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2665 computation of the store needed to hold the pattern was incorrect (too large).
2666 If such patterns were nested a few deep, this could multiply and become a real
2667 problem.
2668
2669 2. Added /M option to pcretest to show the memory requirement of a specific
2670 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2671
2672 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2673 compiled in such a way that the backtracking after subsequent failure was
2674 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2675 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2676
2677
2678 Version 2.03 02-Feb-99
2679 ----------------------
2680
2681 1. Fixed typo and small mistake in man page.
2682
2683 2. Added 4th condition (GPL supersedes if conflict) and created separate
2684 LICENCE file containing the conditions.
2685
2686 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2687 Perl, that is the internal \ allows the delimiter to be included in the
2688 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2689 the final delimiter, add \ to the end of the pattern (to test the error).
2690
2691 4. Added the convenience functions for extracting substrings after a successful
2692 match. Updated pcretest to make it able to test these functions.
2693
2694
2695 Version 2.02 14-Jan-99
2696 ----------------------
2697
2698 1. Initialized the working variables associated with each extraction so that
2699 their saving and restoring doesn't refer to uninitialized store.
2700
2701 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2702 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2703 fix the problem.
2704
2705 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2706 calls, and wasn't printing the correct value for compiling calls. Increased the
2707 default value of LOOPREPEAT, and the number of significant figures in the
2708 times.
2709
2710 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2711
2712 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2713 a building problem on Windows NT with a FAT file system.
2714
2715
2716 Version 2.01 21-Oct-98
2717 ----------------------
2718
2719 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2720 to character tables built by pcre_maketables() in the current locale. If NULL
2721 is passed, the default tables are used.
2722
2723
2724 Version 2.00 24-Sep-98
2725 ----------------------
2726
2727 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2728 it any more.
2729
2730 2. Allow quantification of (?>) groups, and make it work correctly.
2731
2732 3. The first character computation wasn't working for (?>) groups.
2733
2734 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2735 end of the subject) and add 5.005's \z, which really does match only at the
2736 very end of the subject.
2737
2738 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2739
2740 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2741 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2742 localized options. All options to pcre_study() were also removed.
2743
2744 7. Add other new features from 5.005:
2745
2746 $(?<= positive lookbehind
2747 $(?<! negative lookbehind
2748 (?imsx-imsx) added the unsetting capability
2749 such a setting is global if at outer level; local otherwise
2750 (?imsx-imsx:) non-capturing groups with option setting
2751 (?(cond)re|re) conditional pattern matching
2752
2753 A backreference to itself in a repeated group matches the previous
2754 captured string.
2755
2756 8. General tidying up of studying (both automatic and via "study")
2757 consequential on the addition of new assertions.
2758
2759 9. As in 5.005, unlimited repeated groups that could match an empty substring
2760 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2761 runtime if any iteration does actually match an empty substring.
2762
2763 10. Include the RunTest script in the distribution.
2764
2765 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2766 discrepancies, some of which were old and were also with respect to 5.004. They
2767 have now been fixed.
2768
2769
2770 Version 1.09 28-Apr-98
2771 ----------------------
2772
2773 1. A negated single character class followed by a quantifier with a minimum
2774 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2775 program crashes, or just wrong answers. This did not apply to negated classes
2776 containing more than one character, or to minima other than one.
2777
2778
2779 Version 1.08 27-Mar-98
2780 ----------------------
2781
2782 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2783
2784 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2785 latter must appear before anything that relies on it in the pattern.
2786
2787
2788 Version 1.07 16-Feb-98
2789 ----------------------
2790
2791 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2792 repeat of a potentially empty string).
2793
2794
2795 Version 1.06 23-Jan-98
2796 ----------------------
2797
2798 1. Added Markus Oberhumer's little patches for C++.
2799
2800 2. Literal strings longer than 255 characters were broken.
2801
2802
2803 Version 1.05 23-Dec-97
2804 ----------------------
2805
2806 1. Negated character classes containing more than one character were failing if
2807 PCRE_CASELESS was set at run time.
2808
2809
2810 Version 1.04 19-Dec-97
2811 ----------------------
2812
2813 1. Corrected the man page, where some "const" qualifiers had been omitted.
2814
2815 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2816 input syntax.
2817
2818 3. Fixed memory leak which occurred when a regex with back references was
2819 matched with an offsets vector that wasn't big enough. The temporary memory
2820 that is used in this case wasn't being freed if the match failed.
2821
2822 4. Tidied pcretest to ensure it frees memory that it gets.
2823
2824 5. Temporary memory was being obtained in the case where the passed offsets
2825 vector was exactly big enough.
2826
2827 6. Corrected definition of offsetof() from change 5 below.
2828
2829 7. I had screwed up change 6 below and broken the rules for the use of
2830 setjmp(). Now fixed.
2831
2832
2833 Version 1.03 18-Dec-97
2834 ----------------------
2835
2836 1. A erroneous regex with a missing opening parenthesis was correctly
2837 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2838 on some systems.
2839
2840 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2841 it was reported that one broken compiler failed on the former because "code" is
2842 also an independent variable.
2843
2844 3. The erroneous regex a[]b caused an array overrun reference.
2845
2846 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2847 fail on data ending with that character. (It was going on too far, and checking
2848 the next character, typically a binary zero.) This was specific to the
2849 optimized code for single-character negative classes.
2850
2851 5. Added a contributed patch from the TIN world which does the following:
2852
2853 + Add an undef for memmove, in case the the system defines a macro for it.
2854
2855 + Add a definition of offsetof(), in case there isn't one. (I don't know
2856 the reason behind this - offsetof() is part of the ANSI standard - but
2857 it does no harm).
2858
2859 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2860 most of the places where whitespace preceded '#'. I have given up and
2861 allowed the remaining 2 cases to be at the margin.
2862
2863 + Rename some variables in pcre to eliminate shadowing. This seems very
2864 pedantic, but does no harm, of course.
2865
2866 6. Moved the call to setjmp() into its own function, to get rid of warnings
2867 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2868
2869 7. Constructs such as \d{8,} were compiling into the equivalent of
2870 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2871 outcome, but in this particular case used more store than had been allocated,
2872 which caused the bug to be discovered because it threw up an internal error.
2873
2874 8. The debugging code in both pcre and pcretest for outputting the compiled
2875 form of a regex was going wrong in the case of back references followed by
2876 curly-bracketed repeats.
2877
2878
2879 Version 1.02 12-Dec-97
2880 ----------------------
2881
2882 1. Typos in pcre.3 and comments in the source fixed.
2883
2884 2. Applied a contributed patch to get rid of places where it used to remove
2885 'const' from variables, and fixed some signed/unsigned and uninitialized
2886 variable warnings.
2887
2888 3. Added the "runtest" target to Makefile.
2889
2890 4. Set default compiler flag to -O2 rather than just -O.
2891
2892
2893 Version 1.01 19-Nov-97
2894 ----------------------
2895
2896 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2897 like /([ab]*)*/, that is, for classes with more than one character in them.
2898
2899 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2900 as /((?>a*))*/ (a PCRE_EXTRA facility).
2901
2902
2903 Version 1.00 18-Nov-97
2904 ----------------------
2905
2906 1. Added compile-time macros to support systems such as SunOS4 which don't have
2907 memmove() or strerror() but have other things that can be used instead.
2908
2909 2. Arranged that "make clean" removes the executables.
2910
2911
2912 Version 0.99 27-Oct-97
2913 ----------------------
2914
2915 1. Fixed bug in code for optimizing classes with only one character. It was
2916 initializing a 32-byte map regardless, which could cause it to run off the end
2917 of the memory it had got.
2918
2919 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2920
2921
2922 Version 0.98 22-Oct-97
2923 ----------------------
2924
2925 1. Fixed bug in code for handling temporary memory usage when there are more
2926 back references than supplied space in the ovector. This could cause segfaults.
2927
2928
2929 Version 0.97 21-Oct-97
2930 ----------------------
2931
2932 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2933
2934 2. Optimized negated single characters not to use a bit map.
2935
2936 3. Brought error texts together as macro definitions; clarified some of them;
2937 fixed one that was wrong - it said "range out of order" when it meant "invalid
2938 escape sequence".
2939
2940 4. Changed some char * arguments to const char *.
2941
2942 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2943
2944 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2945 pcretest.
2946
2947
2948 Version 0.96 16-Oct-97
2949 ----------------------
2950
2951 1. Added a simple "pgrep" utility to the distribution.
2952
2953 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2954 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2955 where "ddd" means "one or more decimal digits".
2956
2957 3. Fixed serious bug. If a pattern had a back reference, but the call to
2958 pcre_exec() didn't supply a large enough ovector to record the related
2959 identifying subpattern, the match always failed. PCRE now remembers the number
2960 of the largest back reference, and gets some temporary memory in which to save
2961 the offsets during matching if necessary, in order to ensure that
2962 backreferences always work.
2963
2964 4. Increased the compatibility with Perl in a number of ways:
2965
2966 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2967 to request this handling. The option can be set at compile or exec time.
2968
2969 (b) $ matches before a terminating newline by default; an option
2970 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2971 mode). The option can be set at compile or exec time.
2972
2973 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2974 the same as Perl's. If the decimal number it represents is less than 10
2975 or there aren't that many previous left capturing parentheses, an octal
2976 escape is read. Inside a character class, it's always an octal escape,
2977 even if it is a single digit.
2978
2979 (d) An escaped but undefined alphabetic character is taken as a literal,
2980 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2981 escapes.
2982
2983 (e) {0} is now permitted. (The previous item is removed from the compiled
2984 pattern).
2985
2986 5. Changed all the names of code files so that the basic parts are no longer
2987 than 10 characters, and abolished the teeny "globals.c" file.
2988
2989 6. Changed the handling of character classes; they are now done with a 32-byte
2990 bit map always.
2991
2992 7. Added the -d and /D options to pcretest to make it possible to look at the
2993 internals of compilation without having to recompile pcre.
2994
2995
2996 Version 0.95 23-Sep-97
2997 ----------------------
2998
2999 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3000 \x20 at the start of a run of normal characters. These were being treated as
3001 real characters, instead of the source characters being re-checked.
3002
3003
3004 Version 0.94 18-Sep-97
3005 ----------------------
3006
3007 1. The functions are now thread-safe, with the caveat that the global variables
3008 containing pointers to malloc() and free() or alternative functions are the
3009 same for all threads.
3010
3011 2. Get pcre_study() to generate a bitmap of initial characters for non-
3012 anchored patterns when this is possible, and use it if passed to pcre_exec().
3013
3014
3015 Version 0.93 15-Sep-97
3016 ----------------------
3017
3018 1. /(b)|(:+)/ was computing an incorrect first character.
3019
3020 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3021 but not actually doing anything yet.
3022
3023 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3024 as Perl does (e.g. [-az] or [az-]).
3025
3026 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3027 all possible positions.
3028
3029 5. Split up into different modules to avoid including unneeded functions in a
3030 compiled binary. However, compile and exec are still in one module. The "study"
3031 function is split off.
3032
3033 6. The character tables are now in a separate module whose source is generated
3034 by an auxiliary program - but can then be edited by hand if required. There are
3035 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3036 toupper() in the code.
3037
3038 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3039 make them global. Abolish the function for setting them, as the caller can now
3040 set them directly.
3041
3042
3043 Version 0.92 11-Sep-97
3044 ----------------------
3045
3046 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3047 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3048
3049 2. Caseless matching was not working in character classes if the characters in
3050 the pattern were in upper case.
3051
3052 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3053
3054 4. Make PCRE_ANCHORED public and accept as a compile option.
3055
3056 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3057 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3058 pass them.
3059
3060 6. Give an error if bad option bits passed at compile or run time.
3061
3062 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3063 pcretest to cause it to pass that flag.
3064
3065 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3066 options, and the first character, if set.
3067
3068 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3069
3070
3071 Version 0.91 10-Sep-97
3072 ----------------------
3073
3074 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3075 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3076
3077 2. PCRE was looping on encountering an indefinitely repeated back reference to
3078 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3079 Perl does - treats the match as successful.
3080
3081 ****

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