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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 302 - (show annotations) (download)
Sun Jan 20 19:12:46 2008 UTC (6 years, 7 months ago) by ph10
File size: 148842 byte(s)
Exclude "long long" test in pcrecpp_unittest when HAVE_LONG_LONG is not 
defined.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.6 19-Jan-08
5 ---------------------
6
7 1. A character class containing a very large number of characters with
8 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
9 overflow.
10
11 2. Patch to cut out the "long long" test in pcrecpp_unittest when
12 HAVE_LONG_LONG is not defined.
13
14
15 Version 7.5 10-Jan-08
16 ---------------------
17
18 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
19 values in parens when parsing an RE using the C++ wrapper."
20
21 2. Negative specials like \S did not work in character classes in UTF-8 mode.
22 Characters greater than 255 were excluded from the class instead of being
23 included.
24
25 3. The same bug as (2) above applied to negated POSIX classes such as
26 [:^space:].
27
28 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
29 defined or documented. It seems to have been a typo for PCRE_STATIC, so
30 I have changed it.
31
32 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
33 first named subpattern) and a construct such as (?&a) would reference the
34 first named subpattern whose name started with "a" (in other words, the
35 length check was missing). Both these problems are fixed. "Subpattern name
36 expected" is now given for (?&) (a zero-length name), and this patch also
37 makes it give the same error for \k'' (previously it complained that that
38 was a reference to a non-existent subpattern).
39
40 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
41 this is right because (?- can be followed by option settings as well as by
42 digits. I have, however, made the messages clearer.
43
44 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
45 than the number used in the conditional) now cause a compile-time error.
46 This is actually not compatible with Perl, which accepts such patterns, but
47 treats the conditional as always being FALSE (as PCRE used to), but it
48 seems to me that giving a diagnostic is better.
49
50 8. Change "alphameric" to the more common word "alphanumeric" in comments
51 and messages.
52
53 9. Fix two occurrences of "backslash" in comments that should have been
54 "backspace".
55
56 10. Remove two redundant lines of code that can never be obeyed (their function
57 was moved elsewhere).
58
59 11. The program that makes PCRE's Unicode character property table had a bug
60 which caused it to generate incorrect table entries for sequences of
61 characters that have the same character type, but are in different scripts.
62 It amalgamated them into a single range, with the script of the first of
63 them. In other words, some characters were in the wrong script. There were
64 thirteen such cases, affecting characters in the following ranges:
65
66 U+002b0 - U+002c1
67 U+0060c - U+0060d
68 U+0061e - U+00612
69 U+0064b - U+0065e
70 U+0074d - U+0076d
71 U+01800 - U+01805
72 U+01d00 - U+01d77
73 U+01d9b - U+01dbf
74 U+0200b - U+0200f
75 U+030fc - U+030fe
76 U+03260 - U+0327f
77 U+0fb46 - U+0fbb1
78 U+10450 - U+1049d
79
80 12. The -o option (show only the matching part of a line) for pcregrep was not
81 compatible with GNU grep in that, if there was more than one match in a
82 line, it showed only the first of them. It now behaves in the same way as
83 GNU grep.
84
85 13. If the -o and -v options were combined for pcregrep, it printed a blank
86 line for every non-matching line. GNU grep prints nothing, and pcregrep now
87 does the same. The return code can be used to tell if there were any
88 non-matching lines.
89
90 14. Added --file-offsets and --line-offsets to pcregrep.
91
92 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
93 infinitely looping recursion. The bug was that positive lookaheads were not
94 being skipped when checking for a possible empty match (negative lookaheads
95 and both kinds of lookbehind were skipped).
96
97 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
98 inclusion of <windows.h> to before rather than after the definition of
99 INVALID_FILE_ATTRIBUTES (patch from David Byron).
100
101 17. Specifying a possessive quantifier with a specific limit for a Unicode
102 character property caused pcre_compile() to compile bad code, which led at
103 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
104 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
105 caused the error; without that there was no problem.
106
107 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
108
109 19. Added --enable-pcretest-libreadline.
110
111 20. In pcrecpp.cc, the variable 'count' was incremented twice in
112 RE::GlobalReplace(). As a result, the number of replacements returned was
113 double what it should be. I removed one of the increments, but Craig sent a
114 later patch that removed the other one (the right fix) and added unit tests
115 that check the return values (which was not done before).
116
117 21. Several CMake things:
118
119 (1) Arranged that, when cmake is used on Unix, the libraries end up with
120 the names libpcre and libpcreposix, not just pcre and pcreposix.
121
122 (2) The above change means that pcretest and pcregrep are now correctly
123 linked with the newly-built libraries, not previously installed ones.
124
125 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
126
127 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
128 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
129 UTF-8 newline character). The key issue is that the pattern starts .*;
130 this means that the match must be either at the beginning, or after a
131 newline. The bug was in the code for advancing after a failed match and
132 checking that the new position followed a newline. It was not taking
133 account of UTF-8 characters correctly.
134
135 23. PCRE was behaving differently from Perl in the way it recognized POSIX
136 character classes. PCRE was not treating the sequence [:...:] as a
137 character class unless the ... were all letters. Perl, however, seems to
138 allow any characters between [: and :], though of course it rejects as
139 unknown any "names" that contain non-letters, because all the known class
140 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
141 for example, whereas PCRE did not - it did not recognize a POSIX character
142 class. This seemed a bit dangerous, so the code has been changed to be
143 closer to Perl. The behaviour is not identical to Perl, because PCRE will
144 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
145 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
146 Perl does, and where it didn't before.
147
148 24. Rewrite so as to remove the single use of %n from pcregrep because in some
149 Windows environments %n is disabled by default.
150
151
152 Version 7.4 21-Sep-07
153 ---------------------
154
155 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
156 means that a class such as [\s] counted as "explicit reference to CR or
157 LF". That isn't really right - the whole point of the change was to try to
158 help when there was an actual mention of one of the two characters. So now
159 the change happens only if \r or \n (or a literal CR or LF) character is
160 encountered.
161
162 2. The 32-bit options word was also used for 6 internal flags, but the numbers
163 of both had grown to the point where there were only 3 bits left.
164 Fortunately, there was spare space in the data structure, and so I have
165 moved the internal flags into a new 16-bit field to free up more option
166 bits.
167
168 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
169 but did not set the internal JCHANGED flag - either of these is enough to
170 control the way the "get" function works - but the PCRE_INFO_JCHANGED
171 facility is supposed to tell if (?J) was ever used, so now (?J) at the
172 start sets both bits.
173
174 4. Added options (at build time, compile time, exec time) to change \R from
175 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
176
177 5. doc/pcresyntax.html was missing from the distribution.
178
179 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
180 compatibility, even though it is no longer used.
181
182 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
183 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
184 windows.h file is present (where different names are used). [This was
185 reversed later after testing - see 16 below.]
186
187 8. Changed all #include <config.h> to #include "config.h". There were also
188 some further <pcre.h> cases that I changed to "pcre.h".
189
190 9. When pcregrep was used with the --colour option, it missed the line ending
191 sequence off the lines that it output.
192
193 10. It was pointed out to me that arrays of string pointers cause lots of
194 relocations when a shared library is dynamically loaded. A technique of
195 using a single long string with a table of offsets can drastically reduce
196 these. I have refactored PCRE in four places to do this. The result is
197 dramatic:
198
199 Originally: 290
200 After changing UCP table: 187
201 After changing error message table: 43
202 After changing table of "verbs" 36
203 After changing table of Posix names 22
204
205 Thanks to the folks working on Gregex for glib for this insight.
206
207 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
208 unicode-properties was also set.
209
210 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
211
212 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
213 checked only for CRLF.
214
215 14. Added casts to pcretest.c to avoid compiler warnings.
216
217 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
218
219 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
220 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
221 entirely. This removes changes made in 7 above.
222
223 17. The CMake files have been updated, and there is now more information about
224 building with CMake in the NON-UNIX-USE document.
225
226
227 Version 7.3 28-Aug-07
228 ---------------------
229
230 1. In the rejigging of the build system that eventually resulted in 7.1, the
231 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
232 brackets there is not right, since it causes compilers to look for an
233 installed pcre.h, not the version that is in the source that is being
234 compiled (which of course may be different). I have changed it back to:
235
236 #include "pcre.h"
237
238 I have a vague recollection that the change was concerned with compiling in
239 different directories, but in the new build system, that is taken care of
240 by the VPATH setting the Makefile.
241
242 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
243 when the subject happened to end in the byte 0x85 (e.g. if the last
244 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
245 characters but of course it shouldn't be taken as a newline when it is part
246 of another character. The bug was that, for an unlimited repeat of . in
247 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
248 characters when looking for a newline.
249
250 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
251
252 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
253 in debug output.
254
255 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
256 long printing in the pcrecpp unittest when running under MinGW.
257
258 6. ESC_K was left out of the EBCDIC table.
259
260 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
261 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
262 limit also applies to "virtual nesting" when a pattern is recursive, and in
263 this case 1000 isn't so big. I have been able to remove this limit at the
264 expense of backing off one optimization in certain circumstances. Normally,
265 when pcre_exec() would call its internal match() function recursively and
266 immediately return the result unconditionally, it uses a "tail recursion"
267 feature to save stack. However, when a subpattern that can match an empty
268 string has an unlimited repetition quantifier, it no longer makes this
269 optimization. That gives it a stack frame in which to save the data for
270 checking that an empty string has been matched. Previously this was taken
271 from the 1000-entry workspace that had been reserved. So now there is no
272 explicit limit, but more stack is used.
273
274 8. Applied Daniel's patches to solve problems with the import/export magic
275 syntax that is required for Windows, and which was going wrong for the
276 pcreposix and pcrecpp parts of the library. These were overlooked when this
277 problem was solved for the main library.
278
279 9. There were some crude static tests to avoid integer overflow when computing
280 the size of patterns that contain repeated groups with explicit upper
281 limits. As the maximum quantifier is 65535, the maximum group length was
282 set at 30,000 so that the product of these two numbers did not overflow a
283 32-bit integer. However, it turns out that people want to use groups that
284 are longer than 30,000 bytes (though not repeat them that many times).
285 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
286 made it possible to implement the integer overflow checks in a much more
287 dynamic way, which I have now done. The artificial limitation on group
288 length has been removed - we now have only the limit on the total length of
289 the compiled pattern, which depends on the LINK_SIZE setting.
290
291 10. Fixed a bug in the documentation for get/copy named substring when
292 duplicate names are permitted. If none of the named substrings are set, the
293 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
294 empty string.
295
296 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
297 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
298 because the ] is interpreted as the first data character and the
299 terminating ] is not found. PCRE has been made compatible with Perl in this
300 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
301 cause memory overwriting.
302
303 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
304 string has been matched (to stop an infinite loop). It was not recognizing
305 a conditional subpattern that could match an empty string if that
306 subpattern was within another subpattern. For example, it looped when
307 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
308 condition was not nested. This bug has been fixed.
309
310 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
311 past the start of the subject in the presence of bytes with the top bit
312 set, for example "\x8aBCD".
313
314 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
315 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
316
317 14. Optimized (?!) to (*FAIL).
318
319 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
320 This restricts code points to be within the range 0 to 0x10FFFF, excluding
321 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
322 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
323 does: it's just the validity check that is more restrictive.
324
325 16. Inserted checks for integer overflows during escape sequence (backslash)
326 processing, and also fixed erroneous offset values for syntax errors during
327 backslash processing.
328
329 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
330 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
331
332 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
333 caused an overrun.
334
335 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
336 something other than just ASCII characters) inside a group that had an
337 unlimited repeat caused a loop at compile time (while checking to see
338 whether the group could match an empty string).
339
340 20. Debugging a pattern containing \p or \P could cause a crash. For example,
341 [\P{Any}] did so. (Error in the code for printing property names.)
342
343 21. An orphan \E inside a character class could cause a crash.
344
345 22. A repeated capturing bracket such as (A)? could cause a wild memory
346 reference during compilation.
347
348 23. There are several functions in pcre_compile() that scan along a compiled
349 expression for various reasons (e.g. to see if it's fixed length for look
350 behind). There were bugs in these functions when a repeated \p or \P was
351 present in the pattern. These operators have additional parameters compared
352 with \d, etc, and these were not being taken into account when moving along
353 the compiled data. Specifically:
354
355 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
356 length.
357
358 (b) An item such as \pL+ within a repeated group could cause crashes or
359 loops.
360
361 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
362 "reference to non-existent subpattern" error.
363
364 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
365
366 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
367 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
368
369 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
370
371 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
372 character were causing crashes (broken optimization).
373
374 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
375 \p or \P) caused a compile-time loop.
376
377 28. More problems have arisen in unanchored patterns when CRLF is a valid line
378 break. For example, the unstudied pattern [\r\n]A does not match the string
379 "\r\nA" because change 7.0/46 below moves the current point on by two
380 characters after failing to match at the start. However, the pattern \nA
381 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
382 the same is true. There doesn't seem any very clean way out of this, but
383 what I have chosen to do makes the common cases work: PCRE now takes note
384 of whether there can be an explicit match for \r or \n anywhere in the
385 pattern, and if so, 7.0/46 no longer applies. As part of this change,
386 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
387 pattern has explicit CR or LF references.
388
389 29. Added (*CR) etc for changing newline setting at start of pattern.
390
391
392 Version 7.2 19-Jun-07
393 ---------------------
394
395 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
396 which is apparently normally available under Windows.
397
398 2. Re-jig the pcregrep tests with different newline settings in an attempt
399 to make them independent of the local environment's newline setting.
400
401 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
402
403 4. Some of the "internals" tests were previously cut out when the link size
404 was not 2, because the output contained actual offsets. The recent new
405 "Z" feature of pcretest means that these can be cut out, making the tests
406 usable with all link sizes.
407
408 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
409 stack recursion. This gives a massive performance boost under BSD, but just
410 a small improvement under Linux. However, it saves one field in the frame
411 in all cases.
412
413 6. Added more features from the forthcoming Perl 5.10:
414
415 (a) (?-n) (where n is a string of digits) is a relative subroutine or
416 recursion call. It refers to the nth most recently opened parentheses.
417
418 (b) (?+n) is also a relative subroutine call; it refers to the nth next
419 to be opened parentheses.
420
421 (c) Conditions that refer to capturing parentheses can be specified
422 relatively, for example, (?(-2)... or (?(+3)...
423
424 (d) \K resets the start of the current match so that everything before
425 is not part of it.
426
427 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
428
429 (f) \g{name} is another synonym - part of Perl 5.10's unification of
430 reference syntax.
431
432 (g) (?| introduces a group in which the numbering of parentheses in each
433 alternative starts with the same number.
434
435 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
436
437 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
438 PCRE_INFO_JCHANGED.
439
440 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
441 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
442 for detecting groups that can match an empty string.
443
444 9. A pattern with a very large number of alternatives (more than several
445 hundred) was running out of internal workspace during the pre-compile
446 phase, where pcre_compile() figures out how much memory will be needed. A
447 bit of new cunning has reduced the workspace needed for groups with
448 alternatives. The 1000-alternative test pattern now uses 12 bytes of
449 workspace instead of running out of the 4096 that are available.
450
451 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
452
453 11. Applied patch from Google to remove an optimization that didn't quite work.
454 The report of the bug said:
455
456 pcrecpp::RE("a*").FullMatch("aaa") matches, while
457 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
458 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
459
460 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
461 it matched the wrong number of bytes.
462
463
464 Version 7.1 24-Apr-07
465 ---------------------
466
467 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
468 that is more "standard", making use of automake and other Autotools. There
469 is some re-arrangement of the files and adjustment of comments consequent
470 on this.
471
472 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
473 for recursive directory scanning broke on some systems because the files
474 are not scanned in any specific order and on different systems the order
475 was different. A call to "sort" has been inserted into RunGrepTest for the
476 approprate test as a short-term fix. In the longer term there may be an
477 alternative.
478
479 3. I had an email from Eric Raymond about problems translating some of PCRE's
480 man pages to HTML (despite the fact that I distribute HTML pages, some
481 people do their own conversions for various reasons). The problems
482 concerned the use of low-level troff macros .br and .in. I have therefore
483 removed all such uses from the man pages (some were redundant, some could
484 be replaced by .nf/.fi pairs). The 132html script that I use to generate
485 HTML has been updated to handle .nf/.fi and to complain if it encounters
486 .br or .in.
487
488 4. Updated comments in configure.ac that get placed in config.h.in and also
489 arranged for config.h to be included in the distribution, with the name
490 config.h.generic, for the benefit of those who have to compile without
491 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
492
493 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
494 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
495 makevp.bat for the current PCRE, using the additional files
496 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
497
498 6. A Windows user reported a minor discrepancy with test 2, which turned out
499 to be caused by a trailing space on an input line that had got lost in his
500 copy. The trailing space was an accident, so I've just removed it.
501
502 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
503 that is needed.
504
505 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
506 as "const" (a) because they are and (b) because it helps the PHP
507 maintainers who have recently made a script to detect big data structures
508 in the php code that should be moved to the .rodata section. I remembered
509 to update Builducptable as well, so it won't revert if ucptable.h is ever
510 re-created.
511
512 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
513 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
514 order to be able to cut out the UTF-8 tables in the latter when UTF-8
515 support is not required. This saves 1.5-2K of code, which is important in
516 some applications.
517
518 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
519 so as not to refer to the tables, even though these functions will never be
520 called when UTF-8 support is disabled. Otherwise there are problems with a
521 shared library.
522
523 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
524
525 (a) It was defining its arguments as char * instead of void *.
526
527 (b) It was assuming that all moves were upwards in memory; this was true
528 a long time ago when I wrote it, but is no longer the case.
529
530 The emulated memove() is provided for those environments that have neither
531 memmove() nor bcopy(). I didn't think anyone used it these days, but that
532 is clearly not the case, as these two bugs were recently reported.
533
534 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
535 and Detrail to create the HTML documentation, the .txt form of the man
536 pages, and it removes trailing spaces from listed files. It also creates
537 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
538 case, it wraps all the #defines with #ifndefs. This script should be run
539 before "make dist".
540
541 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
542 with Unicode property support.
543
544 (a) For a maximizing quantifier, if the two different cases of the
545 character were of different lengths in their UTF-8 codings (there are
546 some cases like this - I found 11), and the matching function had to
547 back up over a mixture of the two cases, it incorrectly assumed they
548 were both the same length.
549
550 (b) When PCRE was configured to use the heap rather than the stack for
551 recursion during matching, it was not correctly preserving the data for
552 the other case of a UTF-8 character when checking ahead for a match
553 while processing a minimizing repeat. If the check also involved
554 matching a wide character, but failed, corruption could cause an
555 erroneous result when trying to check for a repeat of the original
556 character.
557
558 13. Some tidying changes to the testing mechanism:
559
560 (a) The RunTest script now detects the internal link size and whether there
561 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
562 values substituted by "configure". (The RunGrepTest script already did
563 this for UTF-8.) The configure.ac script no longer substitutes the
564 relevant variables.
565
566 (b) The debugging options /B and /D in pcretest show the compiled bytecode
567 with length and offset values. This means that the output is different
568 for different internal link sizes. Test 2 is skipped for link sizes
569 other than 2 because of this, bypassing the problem. Unfortunately,
570 there was also a test in test 3 (the locale tests) that used /B and
571 failed for link sizes other than 2. Rather than cut the whole test out,
572 I have added a new /Z option to pcretest that replaces the length and
573 offset values with spaces. This is now used to make test 3 independent
574 of link size. (Test 2 will be tidied up later.)
575
576 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
577 segmentation fault instead of returning the appropriate error message.
578
579 15. In multiline mode when the newline sequence was set to "any", the pattern
580 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
581 This doesn't seem right; it now treats the CRLF combination as the line
582 ending, and so does not match in that case. It's only a pattern such as ^$
583 that would hit this one: something like ^ABC$ would have failed after \r
584 and then tried again after \r\n.
585
586 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
587 in an attempt to make files that differ only in their line terminators
588 compare equal. This works on Linux.
589
590 17. Under certain error circumstances pcregrep might try to free random memory
591 as it exited. This is now fixed, thanks to valgrind.
592
593 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
594 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
595 was because its rules for how to advance for /g after matching an empty
596 string at the end of a line did not allow for this case. They now check for
597 it specially.
598
599 20. pcretest is supposed to handle patterns and data of any length, by
600 extending its buffers when necessary. It was getting this wrong when the
601 buffer for a data line had to be extended.
602
603 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
604 CRLF as a newline sequence.
605
606 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
607 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
608 I have nevertheless tidied it up.
609
610 23. Added some casts to kill warnings from HP-UX ia64 compiler.
611
612 24. Added a man page for pcre-config.
613
614
615 Version 7.0 19-Dec-06
616 ---------------------
617
618 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
619 moving to gcc 4.1.1.
620
621 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
622 sys/time.h, which is documented as needed for this function. It doesn't
623 seem to matter on Linux, but it showed up on some releases of OS X.
624
625 3. It seems that there are systems where bytes whose values are greater than
626 127 match isprint() in the "C" locale. The "C" locale should be the
627 default when a C program starts up. In most systems, only ASCII printing
628 characters match isprint(). This difference caused the output from pcretest
629 to vary, making some of the tests fail. I have changed pcretest so that:
630
631 (a) When it is outputting text in the compiled version of a pattern, bytes
632 other than 32-126 are always shown as hex escapes.
633
634 (b) When it is outputting text that is a matched part of a subject string,
635 it does the same, unless a different locale has been set for the match
636 (using the /L modifier). In this case, it uses isprint() to decide.
637
638 4. Fixed a major bug that caused incorrect computation of the amount of memory
639 required for a compiled pattern when options that changed within the
640 pattern affected the logic of the preliminary scan that determines the
641 length. The relevant options are -x, and -i in UTF-8 mode. The result was
642 that the computed length was too small. The symptoms of this bug were
643 either the PCRE error "internal error: code overflow" from pcre_compile(),
644 or a glibc crash with a message such as "pcretest: free(): invalid next
645 size (fast)". Examples of patterns that provoked this bug (shown in
646 pcretest format) are:
647
648 /(?-x: )/x
649 /(?x)(?-x: \s*#\s*)/
650 /((?i)[\x{c0}])/8
651 /(?i:[\x{c0}])/8
652
653 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
654 is now done differently.
655
656 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
657 wrapper classes; (b) implement a new function in the C++ scanner that is
658 more efficient than the old way of doing things because it avoids levels of
659 recursion in the regex matching; (c) add a paragraph to the documentation
660 for the FullMatch() function.
661
662 6. The escape sequence \n was being treated as whatever was defined as
663 "newline". Not only was this contrary to the documentation, which states
664 that \n is character 10 (hex 0A), but it also went horribly wrong when
665 "newline" was defined as CRLF. This has been fixed.
666
667 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
668 was being set to -1 for the "end of line" case (supposedly a value that no
669 character can have). Though this value is never used (the check for end of
670 line is "zero bytes in current character"), it caused compiler complaints.
671 I've changed it to 0xffffffff.
672
673 8. In pcre_version.c, the version string was being built by a sequence of
674 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
675 string (as it is for production releases) called a macro with an empty
676 argument. The C standard says the result of this is undefined. The gcc
677 compiler treats it as an empty string (which was what was wanted) but it is
678 reported that Visual C gives an error. The source has been hacked around to
679 avoid this problem.
680
681 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
682 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
683 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
684 of them did).
685
686 10. Originally, pcretest opened its input and output without "b"; then I was
687 told that "b" was needed in some environments, so it was added for release
688 5.0 to both the input and output. (It makes no difference on Unix-like
689 systems.) Later I was told that it is wrong for the input on Windows. I've
690 now abstracted the modes into two macros, to make it easier to fiddle with
691 them, and removed "b" from the input mode under Windows.
692
693 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
694
695 12. Added -help and --help to pcretest as an official way of being reminded
696 of the options.
697
698 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
699 and pcrecpp.cc because they annoy compilers at high warning levels.
700
701 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
702
703 15. Fixed an occurrence of == in configure.ac that should have been = (shell
704 scripts are not C programs :-) and which was not noticed because it works
705 on Linux.
706
707 16. pcretest is supposed to handle any length of pattern and data line (as one
708 line or as a continued sequence of lines) by extending its input buffer if
709 necessary. This feature was broken for very long pattern lines, leading to
710 a string of junk being passed to pcre_compile() if the pattern was longer
711 than about 50K.
712
713 17. I have done a major re-factoring of the way pcre_compile() computes the
714 amount of memory needed for a compiled pattern. Previously, there was code
715 that made a preliminary scan of the pattern in order to do this. That was
716 OK when PCRE was new, but as the facilities have expanded, it has become
717 harder and harder to keep it in step with the real compile phase, and there
718 have been a number of bugs (see for example, 4 above). I have now found a
719 cunning way of running the real compile function in a "fake" mode that
720 enables it to compute how much memory it would need, while actually only
721 ever using a few hundred bytes of working memory and without too many
722 tests of the mode. This should make future maintenance and development
723 easier. A side effect of this work is that the limit of 200 on the nesting
724 depth of parentheses has been removed (though this was never a serious
725 limitation, I suspect). However, there is a downside: pcre_compile() now
726 runs more slowly than before (30% or more, depending on the pattern). I
727 hope this isn't a big issue. There is no effect on runtime performance.
728
729 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
730 newline (only possible for the last line of a file) and it was a
731 pattern that set a locale (followed by /Lsomething), pcretest crashed.
732
733 19. Added additional timing features to pcretest. (1) The -tm option now times
734 matching only, not compiling. (2) Both -t and -tm can be followed, as a
735 separate command line item, by a number that specifies the number of
736 repeats to use when timing. The default is 50000; this gives better
737 precision, but takes uncomfortably long for very large patterns.
738
739 20. Extended pcre_study() to be more clever in cases where a branch of a
740 subpattern has no definite first character. For example, (a*|b*)[cd] would
741 previously give no result from pcre_study(). Now it recognizes that the
742 first character must be a, b, c, or d.
743
744 21. There was an incorrect error "recursive call could loop indefinitely" if
745 a subpattern (or the entire pattern) that was being tested for matching an
746 empty string contained only one non-empty item after a nested subpattern.
747 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
748 incorrectly, because the \d was being skipped in the check.
749
750 22. The pcretest program now has a new pattern option /B and a command line
751 option -b, which is equivalent to adding /B to every pattern. This causes
752 it to show the compiled bytecode, without the additional information that
753 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
754 is the same as /B/I).
755
756 23. A new optimization is now able automatically to treat some sequences such
757 as a*b as a*+b. More specifically, if something simple (such as a character
758 or a simple class like \d) has an unlimited quantifier, and is followed by
759 something that cannot possibly match the quantified thing, the quantifier
760 is automatically "possessified".
761
762 24. A recursive reference to a subpattern whose number was greater than 39
763 went wrong under certain circumstances in UTF-8 mode. This bug could also
764 have affected the operation of pcre_study().
765
766 25. Realized that a little bit of performance could be had by replacing
767 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
768
769 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
770
771 27. Possessive quantifiers such as a++ were previously implemented by turning
772 them into atomic groups such as ($>a+). Now they have their own opcodes,
773 which improves performance. This includes the automatically created ones
774 from 23 above.
775
776 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
777 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
778 the first matched character to be a colon. This applied both to named and
779 numbered groups.
780
781 29. The ucpinternal.h header file was missing its idempotency #ifdef.
782
783 30. I was sent a "project" file called libpcre.a.dev which I understand makes
784 building PCRE on Windows easier, so I have included it in the distribution.
785
786 31. There is now a check in pcretest against a ridiculously large number being
787 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
788 loop, the loop is abandoned.
789
790 32. Forward references to subpatterns in conditions such as (?(2)...) where
791 subpattern 2 is defined later cause pcre_compile() to search forwards in
792 the pattern for the relevant set of parentheses. This search went wrong
793 when there were unescaped parentheses in a character class, parentheses
794 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
795
796 33. "Subroutine" calls and backreferences were previously restricted to
797 referencing subpatterns earlier in the regex. This restriction has now
798 been removed.
799
800 34. Added a number of extra features that are going to be in Perl 5.10. On the
801 whole, these are just syntactic alternatives for features that PCRE had
802 previously implemented using the Python syntax or my own invention. The
803 other formats are all retained for compatibility.
804
805 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
806 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
807 also .NET compatible.
808
809 (b) A recursion or subroutine call to a named group can now be defined as
810 (?&name) as well as (?P>name).
811
812 (c) A backreference to a named group can now be defined as \k<name> or
813 \k'name' as well as (?P=name). The new forms, as well as being in Perl
814 5.10, are also .NET compatible.
815
816 (d) A conditional reference to a named group can now use the syntax
817 (?(<name>) or (?('name') as well as (?(name).
818
819 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
820 groups (named and numbered) that are never evaluated inline, but can be
821 called as "subroutines" from elsewhere. In effect, the DEFINE condition
822 is always false. There may be only one alternative in such a group.
823
824 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
825 as the simple (?(R). The condition is true only if the most recent
826 recursion is that of the given number or name. It does not search out
827 through the entire recursion stack.
828
829 (g) The escape \gN or \g{N} has been added, where N is a positive or
830 negative number, specifying an absolute or relative reference.
831
832 35. Tidied to get rid of some further signed/unsigned compiler warnings and
833 some "unreachable code" warnings.
834
835 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
836 things, this adds five new scripts.
837
838 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
839 There were also incompatibilities regarding the handling of \Q..\E inside
840 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
841 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
842
843 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
844 matches an empty string, and forcibly breaks the loop. There were bugs in
845 this code in non-simple cases. For a pattern such as ^(a()*)* matched
846 against aaaa the result was just "a" rather than "aaaa", for example. Two
847 separate and independent bugs (that affected different cases) have been
848 fixed.
849
850 39. Refactored the code to abolish the use of different opcodes for small
851 capturing bracket numbers. This is a tidy that I avoided doing when I
852 removed the limit on the number of capturing brackets for 3.5 back in 2001.
853 The new approach is not only tidier, it makes it possible to reduce the
854 memory needed to fix the previous bug (38).
855
856 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
857 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
858 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
859 mode.
860
861 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
862 report.
863
864 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
865 copy construction and assignment in the C++ wrapper.
866
867 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
868 couple of bugs that could have given wrong results in the "--newline=crlf"
869 case.
870
871 44. Added a number of casts and did some reorganization of signed/unsigned int
872 variables following suggestions from Dair Grant. Also renamed the variable
873 "this" as "item" because it is a C++ keyword.
874
875 45. Arranged for dftables to add
876
877 #include "pcre_internal.h"
878
879 to pcre_chartables.c because without it, gcc 4.x may remove the array
880 definition from the final binary if PCRE is built into a static library and
881 dead code stripping is activated.
882
883 46. For an unanchored pattern, if a match attempt fails at the start of a
884 newline sequence, and the newline setting is CRLF or ANY, and the next two
885 characters are CRLF, advance by two characters instead of one.
886
887
888 Version 6.7 04-Jul-06
889 ---------------------
890
891 1. In order to handle tests when input lines are enormously long, pcretest has
892 been re-factored so that it automatically extends its buffers when
893 necessary. The code is crude, but this _is_ just a test program. The
894 default size has been increased from 32K to 50K.
895
896 2. The code in pcre_study() was using the value of the re argument before
897 testing it for NULL. (Of course, in any sensible call of the function, it
898 won't be NULL.)
899
900 3. The memmove() emulation function in pcre_internal.h, which is used on
901 systems that lack both memmove() and bcopy() - that is, hardly ever -
902 was missing a "static" storage class specifier.
903
904 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
905 containing an extended class (one that cannot be represented by a bitmap
906 because it contains high-valued characters or Unicode property items, e.g.
907 [\pZ]). Almost always one would set UTF-8 mode when processing such a
908 pattern, but PCRE should not loop if you do not (it no longer does).
909 [Detail: two cases were found: (a) a repeated subpattern containing an
910 extended class; (b) a recursive reference to a subpattern that followed a
911 previous extended class. It wasn't skipping over the extended class
912 correctly when UTF-8 mode was not set.]
913
914 5. A negated single-character class was not being recognized as fixed-length
915 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
916 compile error "lookbehind assertion is not fixed length".
917
918 6. The RunPerlTest auxiliary script was showing an unexpected difference
919 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
920 write a Perl script that can interpret lines of an input file either as
921 byte characters or as UTF-8, which is what "perltest" was being required to
922 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
923 can't do is switch easily at run time between having the "use utf8;" pragma
924 or not. In the end, I fudged it by using the RunPerlTest script to insert
925 "use utf8;" explicitly for the UTF-8 tests.
926
927 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
928 the end of the subject string, contrary to the documentation and to what
929 Perl does. This was true of both matching functions. Now it matches only at
930 the start of the subject and immediately after *internal* newlines.
931
932 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
933 a pointer to an int instead of a pointer to an unsigned long int. This
934 caused problems on 64-bit systems.
935
936 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
937 instance of the 'standard' template library not being so standard".
938
939 10. There was no check on the number of named subpatterns nor the maximum
940 length of a subpattern name. The product of these values is used to compute
941 the size of the memory block for a compiled pattern. By supplying a very
942 long subpattern name and a large number of named subpatterns, the size
943 computation could be caused to overflow. This is now prevented by limiting
944 the length of names to 32 characters, and the number of named subpatterns
945 to 10,000.
946
947 11. Subpatterns that are repeated with specific counts have to be replicated in
948 the compiled pattern. The size of memory for this was computed from the
949 length of the subpattern and the repeat count. The latter is limited to
950 65535, but there was no limit on the former, meaning that integer overflow
951 could in principle occur. The compiled length of a repeated subpattern is
952 now limited to 30,000 bytes in order to prevent this.
953
954 12. Added the optional facility to have named substrings with the same name.
955
956 13. Added the ability to use a named substring as a condition, using the
957 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
958 are numbers (not recommended). Forward references are permitted.
959
960 14. Added forward references in named backreferences (if you see what I mean).
961
962 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
963 pattern could run off the end of the subject. For example, the pattern
964 "(?s)(.{1,5})"8 did this with the subject "ab".
965
966 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
967 PCRE_CASELESS was set when matching characters that were quantified with ?
968 or *.
969
970 17. A character class other than a single negated character that had a minimum
971 but no maximum quantifier - for example [ab]{6,} - was not handled
972 correctly by pce_dfa_exec(). It would match only one character.
973
974 18. A valid (though odd) pattern that looked like a POSIX character
975 class but used an invalid character after [ (for example [[,abc,]]) caused
976 pcre_compile() to give the error "Failed: internal error: code overflow" or
977 in some cases to crash with a glibc free() error. This could even happen if
978 the pattern terminated after [[ but there just happened to be a sequence of
979 letters, a binary zero, and a closing ] in the memory that followed.
980
981 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
982 over the years. Originally (before any Unicode support), just the bottom 8
983 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
984 output from "man perlunicode" includes this:
985
986 The regular expression compiler produces polymorphic opcodes. That
987 is, the pattern adapts to the data and automatically switches to
988 the Unicode character scheme when presented with Unicode data--or
989 instead uses a traditional byte scheme when presented with byte
990 data.
991
992 Sadly, a wide octal escape does not cause a switch, and in a string with
993 no other multibyte characters, these octal escapes are treated as before.
994 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
995 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
996 Unicode string.
997
998 I have not perpetrated such confusion in PCRE. Up till now, it took just
999 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1000 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1001 translate to the appropriate multibyte character.
1002
1003 29. Applied some refactoring to reduce the number of warnings from Microsoft
1004 and Borland compilers. This has included removing the fudge introduced
1005 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1006 a warning about an unused variable.
1007
1008 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1009 characters since release 4.0, because Perl (from release 5.004) does not.
1010 [Or at least, is documented not to: some releases seem to be in conflict
1011 with the documentation.] However, when a pattern was studied with
1012 pcre_study() and all its branches started with \s, PCRE still included VT
1013 as a possible starting character. Of course, this did no harm; it just
1014 caused an unnecessary match attempt.
1015
1016 22. Removed a now-redundant internal flag bit that recorded the fact that case
1017 dependency changed within the pattern. This was once needed for "required
1018 byte" processing, but is no longer used. This recovers a now-scarce options
1019 bit. Also moved the least significant internal flag bit to the most-
1020 significant bit of the word, which was not previously used (hangover from
1021 the days when it was an int rather than a uint) to free up another bit for
1022 the future.
1023
1024 23. Added support for CRLF line endings as well as CR and LF. As well as the
1025 default being selectable at build time, it can now be changed at runtime
1026 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1027 specify that it is scanning data with non-default line endings.
1028
1029 24. Changed the definition of CXXLINK to make it agree with the definition of
1030 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1031
1032 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1033 recursions. This makes a big different to stack usage for some patterns.
1034
1035 26. If a subpattern containing a named recursion or subroutine reference such
1036 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1037 the space required for the compiled pattern went wrong and gave too small a
1038 value. Depending on the environment, this could lead to "Failed: internal
1039 error: code overflow at offset 49" or "glibc detected double free or
1040 corruption" errors.
1041
1042 27. Applied patches from Google (a) to support the new newline modes and (b) to
1043 advance over multibyte UTF-8 characters in GlobalReplace.
1044
1045 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1046 difference for some implementation of PCRE in some Windows version.
1047
1048 29. Added some extra testing facilities to pcretest:
1049
1050 \q<number> in a data line sets the "match limit" value
1051 \Q<number> in a data line sets the "match recursion limt" value
1052 -S <number> sets the stack size, where <number> is in megabytes
1053
1054 The -S option isn't available for Windows.
1055
1056
1057 Version 6.6 06-Feb-06
1058 ---------------------
1059
1060 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1061 in pcreposix.h. I have copied the definition from pcre.h.
1062
1063 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1064 because pcre.h is no longer a built file.
1065
1066 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1067 not normally included in the compiled code.
1068
1069
1070 Version 6.5 01-Feb-06
1071 ---------------------
1072
1073 1. When using the partial match feature with pcre_dfa_exec(), it was not
1074 anchoring the second and subsequent partial matches at the new starting
1075 point. This could lead to incorrect results. For example, with the pattern
1076 /1234/, partially matching against "123" and then "a4" gave a match.
1077
1078 2. Changes to pcregrep:
1079
1080 (a) All non-match returns from pcre_exec() were being treated as failures
1081 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1082 error message is output. Some extra information is given for the
1083 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1084 probably the only errors that are likely to be caused by users (by
1085 specifying a regex that has nested indefinite repeats, for instance).
1086 If there are more than 20 of these errors, pcregrep is abandoned.
1087
1088 (b) A binary zero was treated as data while matching, but terminated the
1089 output line if it was written out. This has been fixed: binary zeroes
1090 are now no different to any other data bytes.
1091
1092 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1093 used to set a locale for matching. The --locale=xxxx long option has
1094 been added (no short equivalent) to specify a locale explicitly on the
1095 pcregrep command, overriding the environment variables.
1096
1097 (d) When -B was used with -n, some line numbers in the output were one less
1098 than they should have been.
1099
1100 (e) Added the -o (--only-matching) option.
1101
1102 (f) If -A or -C was used with -c (count only), some lines of context were
1103 accidentally printed for the final match.
1104
1105 (g) Added the -H (--with-filename) option.
1106
1107 (h) The combination of options -rh failed to suppress file names for files
1108 that were found from directory arguments.
1109
1110 (i) Added the -D (--devices) and -d (--directories) options.
1111
1112 (j) Added the -F (--fixed-strings) option.
1113
1114 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1115
1116 (l) Added the --colo(u)r option.
1117
1118 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1119 is not present by default.
1120
1121 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1122 items such as (?R) or (?1), when the recursion could match a number of
1123 alternatives. If it matched one of the alternatives, but subsequently,
1124 outside the recursion, there was a failure, the code tried to back up into
1125 the recursion. However, because of the way PCRE is implemented, this is not
1126 possible, and the result was an incorrect result from the match.
1127
1128 In order to prevent this happening, the specification of recursion has
1129 been changed so that all such subpatterns are automatically treated as
1130 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1131
1132 4. I had overlooked the fact that, in some locales, there are characters for
1133 which isalpha() is true but neither isupper() nor islower() are true. In
1134 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1135 and ordfeminine) are like this. This affected the treatment of \w and \W
1136 when they appeared in character classes, but not when they appeared outside
1137 a character class. The bit map for "word" characters is now created
1138 separately from the results of isalnum() instead of just taking it from the
1139 upper, lower, and digit maps. (Plus the underscore character, of course.)
1140
1141 5. The above bug also affected the handling of POSIX character classes such as
1142 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1143 permanent tables. Instead, the bit maps for such a class were previously
1144 created as the appropriate unions of the upper, lower, and digit bitmaps.
1145 Now they are created by subtraction from the [[:word:]] class, which has
1146 its own bitmap.
1147
1148 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1149 It is created by subtracting the vertical space characters (\x09, \x0a,
1150 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1151 subtraction was done in the overall bitmap for a character class, meaning
1152 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1153 be recognized. This bug has been fixed.
1154
1155 7. Patches from the folks at Google:
1156
1157 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1158 real life, but is still worth protecting against".
1159
1160 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1161 regular expressions".
1162
1163 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1164 have it.
1165
1166 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1167 "configure" and the latter not, in order to fix a problem somebody had
1168 with compiling the Arg class on HP-UX.
1169
1170 (e) Improve the error-handling of the C++ wrapper a little bit.
1171
1172 (f) New tests for checking recursion limiting.
1173
1174 8. The pcre_memmove() function, which is used only if the environment does not
1175 have a standard memmove() function (and is therefore rarely compiled),
1176 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1177 returning a result (though PCRE never actually uses the result).
1178
1179 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1180 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1181 returned instead of calling malloc() with an overflowing number that would
1182 most likely cause subsequent chaos.
1183
1184 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1185
1186 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1187 with this option is matched, the nmatch and pmatch options of regexec() are
1188 ignored.
1189
1190 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1191 provided in case anyone wants to the the POSIX interface with UTF-8
1192 strings.
1193
1194 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1195 C++ linking (needed for some HP-UX environments).
1196
1197 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1198 (unused parameter) and in the pcre_printint() function (omitted "default"
1199 switch label when the default is to do nothing).
1200
1201 15. Added some code to make it possible, when PCRE is compiled as a C++
1202 library, to replace subject pointers for pcre_exec() with a smart pointer
1203 class, thus making it possible to process discontinuous strings.
1204
1205 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1206 much the same function. They were added by different people who were trying
1207 to make PCRE easy to compile on non-Unix systems. It has been suggested
1208 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1209 for compiling on Windows systems. I have therefore replaced it with
1210 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1211 defaults to "extern" for C or "extern C" for C++, which works fine on
1212 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1213 SCOPE with something explicit in config.h. In addition:
1214
1215 (a) pcreposix.h still had just "extern" instead of either of these macros;
1216 I have replaced it with PCRE_DATA_SCOPE.
1217
1218 (b) Functions such as _pcre_xclass(), which are internal to the library,
1219 but external in the C sense, all had PCRE_EXPORT in their definitions.
1220 This is apparently wrong for the Windows case, so I have removed it.
1221 (It makes no difference on Unix-like systems.)
1222
1223 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1224 of recursive calls to match(). This is different to MATCH_LIMIT because
1225 that limits the total number of calls to match(), not all of which increase
1226 the depth of recursion. Limiting the recursion depth limits the amount of
1227 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1228 when PCRE is compiled, and changed at run time. A patch from Google adds
1229 this functionality to the C++ interface.
1230
1231 18. Changes to the handling of Unicode character properties:
1232
1233 (a) Updated the table to Unicode 4.1.0.
1234
1235 (b) Recognize characters that are not in the table as "Cn" (undefined).
1236
1237 (c) I revised the way the table is implemented to a much improved format
1238 which includes recognition of ranges. It now supports the ranges that
1239 are defined in UnicodeData.txt, and it also amalgamates other
1240 characters into ranges. This has reduced the number of entries in the
1241 table from around 16,000 to around 3,000, thus reducing its size
1242 considerably. I realized I did not need to use a tree structure after
1243 all - a binary chop search is just as efficient. Having reduced the
1244 number of entries, I extended their size from 6 bytes to 8 bytes to
1245 allow for more data.
1246
1247 (d) Added support for Unicode script names via properties such as \p{Han}.
1248
1249 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1250 matching that character.
1251
1252 20. When matching a repeated Unicode property with a minimum greater than zero,
1253 (for example \pL{2,}), PCRE could look past the end of the subject if it
1254 reached it while seeking the minimum number of characters. This could
1255 happen only if some of the characters were more than one byte long, because
1256 there is a check for at least the minimum number of bytes.
1257
1258 21. Refactored the implementation of \p and \P so as to be more general, to
1259 allow for more different types of property in future. This has changed the
1260 compiled form incompatibly. Anybody with saved compiled patterns that use
1261 \p or \P will have to recompile them.
1262
1263 22. Added "Any" and "L&" to the supported property types.
1264
1265 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1266 but give a compile time error if the value is greater than 0xff.
1267
1268 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1269 accidentally not being installed or uninstalled.
1270
1271 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1272 made were to insert the current release number. This seemed silly, because
1273 it made things harder for people building PCRE on systems that don't run
1274 "configure". I have turned pcre.h into a distributed file, no longer built
1275 by "configure", with the version identification directly included. There is
1276 no longer a pcre.h.in file.
1277
1278 However, this change necessitated a change to the pcre-config script as
1279 well. It is built from pcre-config.in, and one of the substitutions was the
1280 release number. I have updated configure.ac so that ./configure now finds
1281 the release number by grepping pcre.h.
1282
1283 26. Added the ability to run the tests under valgrind.
1284
1285
1286 Version 6.4 05-Sep-05
1287 ---------------------
1288
1289 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1290 "--" to be printed when multiple files were scanned, even when none of the
1291 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1292 consider it to be a bug, and have restored the previous behaviour.
1293
1294 2. A couple of code tidies to get rid of compiler warnings.
1295
1296 3. The pcretest program used to cheat by referring to symbols in the library
1297 whose names begin with _pcre_. These are internal symbols that are not
1298 really supposed to be visible externally, and in some environments it is
1299 possible to suppress them. The cheating is now confined to including
1300 certain files from the library's source, which is a bit cleaner.
1301
1302 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1303 file's purpose clearer.
1304
1305 5. Reorganized pcre_ucp_findchar().
1306
1307
1308 Version 6.3 15-Aug-05
1309 ---------------------
1310
1311 1. The file libpcre.pc.in did not have general read permission in the tarball.
1312
1313 2. There were some problems when building without C++ support:
1314
1315 (a) If C++ support was not built, "make install" and "make test" still
1316 tried to test it.
1317
1318 (b) There were problems when the value of CXX was explicitly set. Some
1319 changes have been made to try to fix these, and ...
1320
1321 (c) --disable-cpp can now be used to explicitly disable C++ support.
1322
1323 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1324 backslash in a target when C++ was disabled. This confuses some
1325 versions of "make", apparently. Using an intermediate variable solves
1326 this. (Same for CPP_LOBJ.)
1327
1328 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1329 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1330 necessary on certain architectures.
1331
1332 4. Added a setting of -export-symbols-regex to the link command to remove
1333 those symbols that are exported in the C sense, but actually are local
1334 within the library, and not documented. Their names all begin with
1335 "_pcre_". This is not a perfect job, because (a) we have to except some
1336 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1337 available (and never for static libraries). I have made a note to try to
1338 find a way round (a) in the future.
1339
1340
1341 Version 6.2 01-Aug-05
1342 ---------------------
1343
1344 1. There was no test for integer overflow of quantifier values. A construction
1345 such as {1111111111111111} would give undefined results. What is worse, if
1346 a minimum quantifier for a parenthesized subpattern overflowed and became
1347 negative, the calculation of the memory size went wrong. This could have
1348 led to memory overwriting.
1349
1350 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1351
1352 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1353 operating environments where this matters.
1354
1355 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1356 PCRE options from within the C++ wrapper.
1357
1358 5. Named capturing subpatterns were not being correctly counted when a pattern
1359 was compiled. This caused two problems: (a) If there were more than 100
1360 such subpatterns, the calculation of the memory needed for the whole
1361 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1362 back references of the form \12, where the number was greater than 9, were
1363 not recognized as back references, even though there were sufficient
1364 previous subpatterns.
1365
1366 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1367 versions of gcc, e.g. 2.95.4.
1368
1369
1370 Version 6.1 21-Jun-05
1371 ---------------------
1372
1373 1. There was one reference to the variable "posix" in pcretest.c that was not
1374 surrounded by "#if !defined NOPOSIX".
1375
1376 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1377 the cross-check on the old pcre_info() function, for the benefit of the
1378 cut-down version of PCRE that is currently imported into Exim.
1379
1380 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1381 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1382 patterns that start (?i) but I don't think that matters. The use of (?i) is
1383 just an example; this all applies to the other options as well.
1384
1385 4. Since libtool seems to echo the compile commands it is issuing, the output
1386 from "make" can be reduced a bit by putting "@" in front of each libtool
1387 compile command.
1388
1389 5. Patch from the folks at Google for configure.in to be a bit more thorough
1390 in checking for a suitable C++ installation before trying to compile the
1391 C++ stuff. This should fix a reported problem when a compiler was present,
1392 but no suitable headers.
1393
1394 6. The man pages all had just "PCRE" as their title. I have changed them to
1395 be the relevant file name. I have also arranged that these names are
1396 retained in the file doc/pcre.txt, which is a concatenation in text format
1397 of all the man pages except the little individual ones for each function.
1398
1399 7. The NON-UNIX-USE file had not been updated for the different set of source
1400 files that come with release 6. I also added a few comments about the C++
1401 wrapper.
1402
1403
1404 Version 6.0 07-Jun-05
1405 ---------------------
1406
1407 1. Some minor internal re-organization to help with my DFA experiments.
1408
1409 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1410 didn't matter for the library itself when fully configured, but did matter
1411 when compiling without UCP support, or within Exim, where the ucp files are
1412 not imported.
1413
1414 3. Refactoring of the library code to split up the various functions into
1415 different source modules. The addition of the new DFA matching code (see
1416 below) to a single monolithic source would have made it really too
1417 unwieldy, quite apart from causing all the code to be include in a
1418 statically linked application, when only some functions are used. This is
1419 relevant even without the DFA addition now that patterns can be compiled in
1420 one application and matched in another.
1421
1422 The downside of splitting up is that there have to be some external
1423 functions and data tables that are used internally in different modules of
1424 the library but which are not part of the API. These have all had their
1425 names changed to start with "_pcre_" so that they are unlikely to clash
1426 with other external names.
1427
1428 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1429 a different (DFA) algorithm. Although it is slower than the original
1430 function, it does have some advantages for certain types of matching
1431 problem.
1432
1433 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1434 including restarting after a partial match.
1435
1436 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1437 defined when compiling for Windows was sent to me. I have put it into the
1438 code, though I have no means of testing or verifying it.
1439
1440 7. Added the pcre_refcount() auxiliary function.
1441
1442 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1443 match before or at the first newline in the subject string. In pcretest,
1444 the /f option on a pattern can be used to set this.
1445
1446 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1447 would behave wrongly. This has been present in PCRE since release 4.0.
1448
1449 10. A number of changes to the pcregrep command:
1450
1451 (a) Refactored how -x works; insert ^(...)$ instead of setting
1452 PCRE_ANCHORED and checking the length, in preparation for adding
1453 something similar for -w.
1454
1455 (b) Added the -w (match as a word) option.
1456
1457 (c) Refactored the way lines are read and buffered so as to have more
1458 than one at a time available.
1459
1460 (d) Implemented a pcregrep test script.
1461
1462 (e) Added the -M (multiline match) option. This allows patterns to match
1463 over several lines of the subject. The buffering ensures that at least
1464 8K, or the rest of the document (whichever is the shorter) is available
1465 for matching (and similarly the previous 8K for lookbehind assertions).
1466
1467 (f) Changed the --help output so that it now says
1468
1469 -w, --word-regex(p)
1470
1471 instead of two lines, one with "regex" and the other with "regexp"
1472 because that confused at least one person since the short forms are the
1473 same. (This required a bit of code, as the output is generated
1474 automatically from a table. It wasn't just a text change.)
1475
1476 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1477 option but starts with a hyphen. Could be a pattern or a path name
1478 starting with a hyphen, for instance.
1479
1480 (h) "-" can be given as a file name to represent stdin.
1481
1482 (i) When file names are being printed, "(standard input)" is used for
1483 the standard input, for compatibility with GNU grep. Previously
1484 "<stdin>" was used.
1485
1486 (j) The option --label=xxx can be used to supply a name to be used for
1487 stdin when file names are being printed. There is no short form.
1488
1489 (k) Re-factored the options decoding logic because we are going to add
1490 two more options that take data. Such options can now be given in four
1491 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1492
1493 (l) Added the -A, -B, and -C options for requesting that lines of context
1494 around matches be printed.
1495
1496 (m) Added the -L option to print the names of files that do not contain
1497 any matching lines, that is, the complement of -l.
1498
1499 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1500 continue to scan other files.
1501
1502 (o) The -s option was incorrectly implemented. For compatibility with other
1503 greps, it now suppresses the error message for a non-existent or non-
1504 accessible file (but not the return code). There is a new option called
1505 -q that suppresses the output of matching lines, which was what -s was
1506 previously doing.
1507
1508 (p) Added --include and --exclude options to specify files for inclusion
1509 and exclusion when recursing.
1510
1511 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1512 Hopefully, it now does.
1513
1514 12. Missing cast in pcre_study().
1515
1516 13. Added an "uninstall" target to the makefile.
1517
1518 14. Replaced "extern" in the function prototypes in Makefile.in with
1519 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1520 world, but is set differently for Windows.
1521
1522 15. Added a second compiling function called pcre_compile2(). The only
1523 difference is that it has an extra argument, which is a pointer to an
1524 integer error code. When there is a compile-time failure, this is set
1525 non-zero, in addition to the error test pointer being set to point to an
1526 error message. The new argument may be NULL if no error number is required
1527 (but then you may as well call pcre_compile(), which is now just a
1528 wrapper). This facility is provided because some applications need a
1529 numeric error indication, but it has also enabled me to tidy up the way
1530 compile-time errors are handled in the POSIX wrapper.
1531
1532 16. Added VPATH=.libs to the makefile; this should help when building with one
1533 prefix path and installing with another. (Or so I'm told by someone who
1534 knows more about this stuff than I do.)
1535
1536 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1537 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1538 match everything, including newlines. This is not POSIX-compatible, but
1539 somebody wanted the feature. From pcretest it can be activated by using
1540 both the P and the s flags.
1541
1542 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1543
1544 19. libpcre.pc was being incorrectly installed as executable.
1545
1546 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1547 it now also looks for '\r' so that it will work unmodified on Windows.
1548
1549 21. Added Google's contributed C++ wrapper to the distribution.
1550
1551 22. Added some untidy missing memory free() calls in pcretest, to keep
1552 Electric Fence happy when testing.
1553
1554
1555
1556 Version 5.0 13-Sep-04
1557 ---------------------
1558
1559 1. Internal change: literal characters are no longer packed up into items
1560 containing multiple characters in a single byte-string. Each character
1561 is now matched using a separate opcode. However, there may be more than one
1562 byte in the character in UTF-8 mode.
1563
1564 2. The pcre_callout_block structure has two new fields: pattern_position and
1565 next_item_length. These contain the offset in the pattern to the next match
1566 item, and its length, respectively.
1567
1568 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1569 insertion of callouts before each pattern item. Added the /C option to
1570 pcretest to make use of this.
1571
1572 4. On the advice of a Windows user, the lines
1573
1574 #if defined(_WIN32) || defined(WIN32)
1575 _setmode( _fileno( stdout ), 0x8000 );
1576 #endif /* defined(_WIN32) || defined(WIN32) */
1577
1578 have been added to the source of pcretest. This apparently does useful
1579 magic in relation to line terminators.
1580
1581 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1582 for the benefit of those environments where the "b" makes a difference.
1583
1584 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1585 to know about it. I have put a hack into configure.in that adds in code
1586 to set GCC=yes if CC=icc. This seems to end up at a point in the
1587 generated configure script that is early enough to affect the setting of
1588 compiler options, which is what is needed, but I have no means of testing
1589 whether it really works. (The user who reported this had patched the
1590 generated configure script, which of course I cannot do.)
1591
1592 LATER: After change 22 below (new libtool files), the configure script
1593 seems to know about icc (and also ecc). Therefore, I have commented out
1594 this hack in configure.in.
1595
1596 7. Added support for pkg-config (2 patches were sent in).
1597
1598 8. Negated POSIX character classes that used a combination of internal tables
1599 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1600 [[:^ascii]]. Typically, they would match almost any characters. The other
1601 POSIX classes were not broken in this way.
1602
1603 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1604 to find the match, as PCRE was deluded into thinking that the match had to
1605 start at the start point or following a newline. The same bug applied to
1606 patterns with negative forward assertions or any backward assertions
1607 preceding ".*" at the start, unless the pattern required a fixed first
1608 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1609
1610 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1611 starting at the last subject character, bytes beyond the end of the subject
1612 string were read.
1613
1614 11. Renamed the variable "class" as "classbits" to make life easier for C++
1615 users. (Previously there was a macro definition, but it apparently wasn't
1616 enough.)
1617
1618 12. Added the new field "tables" to the extra data so that tables can be passed
1619 in at exec time, or the internal tables can be re-selected. This allows
1620 a compiled regex to be saved and re-used at a later time by a different
1621 program that might have everything at different addresses.
1622
1623 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1624 -R library as well as a -L library.
1625
1626 14. The debugging options of pcretest (-d on the command line or D on a
1627 pattern) showed incorrect output for anything following an extended class
1628 that contained multibyte characters and which was followed by a quantifier.
1629
1630 15. Added optional support for general category Unicode character properties
1631 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1632 support. It adds about 90K to the size of the library. The meanings of the
1633 inbuilt class escapes such as \d and \s have NOT been changed.
1634
1635 16. Updated pcredemo.c to include calls to free() to release the memory for the
1636 compiled pattern.
1637
1638 17. The generated file chartables.c was being created in the source directory
1639 instead of in the building directory. This caused the build to fail if the
1640 source directory was different from the building directory, and was
1641 read-only.
1642
1643 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1644 file. No doubt somebody will tell me if they don't make sense... Also added
1645 Dan Mooney's comments about building on OpenVMS.
1646
1647 19. Added support for partial matching via the PCRE_PARTIAL option for
1648 pcre_exec() and the \P data escape in pcretest.
1649
1650 20. Extended pcretest with 3 new pattern features:
1651
1652 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1653 write the compiled pattern to the file whose name is "rest-of-line".
1654 This is a straight binary dump of the data, with the saved pointer to
1655 the character tables forced to be NULL. The study data, if any, is
1656 written too. After writing, pcretest reads a new pattern.
1657
1658 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1659 compiled pattern from the given file. There must not be any
1660 occurrences of "<" in the file name (pretty unlikely); if there are,
1661 pcretest will instead treat the initial "<" as a pattern delimiter.
1662 After reading in the pattern, pcretest goes on to read data lines as
1663 usual.
1664
1665 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1666 and 16-bit fields in a compiled pattern, to simulate a pattern that
1667 was compiled on a host of opposite endianness.
1668
1669 21. The pcre-exec() function can now cope with patterns that were compiled on
1670 hosts of opposite endianness, with this restriction:
1671
1672 As for any compiled expression that is saved and used later, the tables
1673 pointer field cannot be preserved; the extra_data field in the arguments
1674 to pcre_exec() should be used to pass in a tables address if a value
1675 other than the default internal tables were used at compile time.
1676
1677 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1678 now diagnosed as an error. Previously, most of the time, a negative number
1679 would have been treated as zero, but if in addition "ovector" was passed as
1680 NULL, a crash could occur.
1681
1682 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1683 new versions from the libtool 1.5 distribution (the last one is a copy of
1684 a file called libtool.m4). This seems to have fixed the need to patch
1685 "configure" to support Darwin 1.3 (which I used to do). However, I still
1686 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1687 workstation).
1688
1689 24. Changed the PCRE licence to be the more standard "BSD" licence.
1690
1691
1692 Version 4.5 01-Dec-03
1693 ---------------------
1694
1695 1. There has been some re-arrangement of the code for the match() function so
1696 that it can be compiled in a version that does not call itself recursively.
1697 Instead, it keeps those local variables that need separate instances for
1698 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1699 needs to "recurse". Keeping track of where control must go is done by means
1700 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1701 hide most of the details from the main code, and operates only if
1702 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1703 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1704 operating.
1705
1706 To make it easier for callers to provide specially tailored get/free
1707 functions for this usage, two new functions, pcre_stack_malloc, and
1708 pcre_stack_free, are used. They are always called in strict stacking order,
1709 and the size of block requested is always the same.
1710
1711 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1712 PCRE has been compiled to use the stack or the heap for recursion. The
1713 -C option of pcretest uses this to show which version is compiled.
1714
1715 A new data escape \S, is added to pcretest; it causes the amounts of store
1716 obtained and freed by both kinds of malloc/free at match time to be added
1717 to the output.
1718
1719 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1720 what's available on my current Linux desktop machine.
1721
1722 3. When matching a UTF-8 string, the test for a valid string at the start has
1723 been extended. If start_offset is not zero, PCRE now checks that it points
1724 to a byte that is the start of a UTF-8 character. If not, it returns
1725 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1726 this is necessary because there may be backward assertions in the pattern.
1727 When matching the same subject several times, it may save resources to use
1728 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1729
1730 4. The code for checking the validity of UTF-8 strings has been tightened so
1731 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1732 containing "overlong sequences".
1733
1734 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1735 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1736 should have been "&", but it just so happened that all the cases this let
1737 through by mistake were picked up later in the function.
1738
1739 6. I had used a variable called "isblank" - this is a C99 function, causing
1740 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1741
1742 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1743 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1744 so the version is shown; (c) stop "make test" echoing "./RunTest".
1745
1746 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1747
1748 9. The prototype for memmove() for systems that don't have it was using
1749 size_t, but the inclusion of the header that defines size_t was later. I've
1750 moved the #includes for the C headers earlier to avoid this.
1751
1752 10. Added some adjustments to the code to make it easier to compiler on certain
1753 special systems:
1754
1755 (a) Some "const" qualifiers were missing.
1756 (b) Added the macro EXPORT before all exported functions; by default this
1757 is defined to be empty.
1758 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1759 that it reads its output file name as an argument instead of writing
1760 to the standard output and assuming this can be redirected.
1761
1762 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1763 class containing characters with values greater than 255, PCRE compilation
1764 went into a loop.
1765
1766 12. A recursive reference to a subpattern that was within another subpattern
1767 that had a minimum quantifier of zero caused PCRE to crash. For example,
1768 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1769 recursion. If the recursively-called subpattern itself had a zero repeat,
1770 that was OK.
1771
1772 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1773 buffer into which it was copied (for escape processing) was still set at
1774 1024, so long lines caused crashes.
1775
1776 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1777 "internal error: code overflow...". This applied to any character class
1778 that was followed by a possessive quantifier.
1779
1780 15. Modified the Makefile to add libpcre.la as a prerequisite for
1781 libpcreposix.la because I was told this is needed for a parallel build to
1782 work.
1783
1784 16. If a pattern that contained .* following optional items at the start was
1785 studied, the wrong optimizing data was generated, leading to matching
1786 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1787 matching string must start with a or b or c. The correct conclusion for
1788 this pattern is that a match can start with any character.
1789
1790
1791 Version 4.4 13-Aug-03
1792 ---------------------
1793
1794 1. In UTF-8 mode, a character class containing characters with values between
1795 127 and 255 was not handled correctly if the compiled pattern was studied.
1796 In fixing this, I have also improved the studying algorithm for such
1797 classes (slightly).
1798
1799 2. Three internal functions had redundant arguments passed to them. Removal
1800 might give a very teeny performance improvement.
1801
1802 3. Documentation bug: the value of the capture_top field in a callout is *one
1803 more than* the number of the hightest numbered captured substring.
1804
1805 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1806 in incorrectly linking with a previously installed version. They now link
1807 explicitly with libpcre.la.
1808
1809 5. configure.in no longer needs to recognize Cygwin specially.
1810
1811 6. A problem in pcre.in for Windows platforms is fixed.
1812
1813 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1814 pcretest, it used to include the size of the study block as part of its
1815 output. Unfortunately, the structure contains a field that has a different
1816 size on different hardware architectures. This meant that the tests that
1817 showed this size failed. As the block is currently always of a fixed size,
1818 this information isn't actually particularly useful in pcretest output, so
1819 I have just removed it.
1820
1821 8. Three pre-processor statements accidentally did not start in column 1.
1822 Sadly, there are *still* compilers around that complain, even though
1823 standard C has not required this for well over a decade. Sigh.
1824
1825 9. In pcretest, the code for checking callouts passed small integers in the
1826 callout_data field, which is a void * field. However, some picky compilers
1827 complained about the casts involved for this on 64-bit systems. Now
1828 pcretest passes the address of the small integer instead, which should get
1829 rid of the warnings.
1830
1831 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1832 both compile and run time, and gives an error if an invalid UTF-8 sequence
1833 is found. There is a option for disabling this check in cases where the
1834 string is known to be correct and/or the maximum performance is wanted.
1835
1836 11. In response to a bug report, I changed one line in Makefile.in from
1837
1838 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1839 to
1840 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1841
1842 to look similar to other lines, but I have no way of telling whether this
1843 is the right thing to do, as I do not use Windows. No doubt I'll get told
1844 if it's wrong...
1845
1846
1847 Version 4.3 21-May-03
1848 ---------------------
1849
1850 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1851 Makefile.
1852
1853 2. Some refactoring to improve the quality of the code:
1854
1855 (i) The utf8_table... variables are now declared "const".
1856
1857 (ii) The code for \cx, which used the "case flipping" table to upper case
1858 lower case letters, now just substracts 32. This is ASCII-specific,
1859 but the whole concept of \cx is ASCII-specific, so it seems
1860 reasonable.
1861
1862 (iii) PCRE was using its character types table to recognize decimal and
1863 hexadecimal digits in the pattern. This is silly, because it handles
1864 only 0-9, a-f, and A-F, but the character types table is locale-
1865 specific, which means strange things might happen. A private
1866 table is now used for this - though it costs 256 bytes, a table is
1867 much faster than multiple explicit tests. Of course, the standard
1868 character types table is still used for matching digits in subject
1869 strings against \d.
1870
1871 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1872 ending in _t are). So I've renamed it as ESC_tee.
1873
1874 3. The first argument for regexec() in the POSIX wrapper should have been
1875 defined as "const".
1876
1877 4. Changed pcretest to use malloc() for its buffers so that they can be
1878 Electric Fenced for debugging.
1879
1880 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1881 to read one or more bytes before the start of the subject string. Often this
1882 had no effect on PCRE's behaviour, but in some circumstances it could
1883 provoke a segmentation fault.
1884
1885 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1886 to try to read one or more bytes before the start of the subject string.
1887
1888 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1889 UTF-8 support could misbehave in various ways if the subject string
1890 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1891 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1892 back over UTF-8 characters.)
1893
1894
1895 Version 4.2 14-Apr-03
1896 ---------------------
1897
1898 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1899
1900 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1901 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1902 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1903 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1904 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1905 and BUILD_EXEEXT
1906 Note: automatic setting of the BUILD variables is not yet working
1907 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1908 compile-time but not at link-time
1909 [LINK]: use for linking executables only
1910 make different versions for Windows and non-Windows
1911 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1912 libraries
1913 [LINK_FOR_BUILD]: new variable
1914 [OBJEXT]: use throughout
1915 [EXEEXT]: use throughout
1916 <winshared>: new target
1917 <wininstall>: new target
1918 <dftables.o>: use native compiler
1919 <dftables>: use native linker
1920 <install>: handle Windows platform correctly
1921 <clean>: ditto
1922 <check>: ditto
1923 copy DLL to top builddir before testing
1924
1925 As part of these changes, -no-undefined was removed again. This was reported
1926 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1927 in any case.
1928
1929 3. Some tidies to get rid of compiler warnings:
1930
1931 . In the match_data structure, match_limit was an unsigned long int, whereas
1932 match_call_count was an int. I've made them both unsigned long ints.
1933
1934 . In pcretest the fact that a const uschar * doesn't automatically cast to
1935 a void * provoked a warning.
1936
1937 . Turning on some more compiler warnings threw up some "shadow" variables
1938 and a few more missing casts.
1939
1940 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1941 option, a class that contained a single character with a value between 128
1942 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1943
1944 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1945 option, a class that contained several characters, but with at least one
1946 whose value was between 128 and 255 caused PCRE to crash.
1947
1948
1949 Version 4.1 12-Mar-03
1950 ---------------------
1951
1952 1. Compiling with gcc -pedantic found a couple of places where casts were
1953 needed, and a string in dftables.c that was longer than standard compilers are
1954 required to support.
1955
1956 2. Compiling with Sun's compiler found a few more places where the code could
1957 be tidied up in order to avoid warnings.
1958
1959 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1960 first of these names is deprecated in the latest Autoconf in favour of the name
1961 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1962 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1963 analogy I have changed it to CFLAGS_FOR_BUILD.
1964
1965 4. Added -no-undefined to the linking command in the Makefile, because this is
1966 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1967 linking step for the pcreposix library.
1968
1969 5. PCRE was failing to diagnose the case of two named groups with the same
1970 name.
1971
1972 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1973 literal character that is needed in the subject for a match, and scans along to
1974 ensure that it is present before embarking on the full matching process. This
1975 saves time in cases of nested unlimited repeats that are never going to match.
1976 Problem: the scan can take a lot of time if the subject is very long (e.g.
1977 megabytes), thus penalizing straightforward matches. It is now done only if the
1978 amount of subject to be scanned is less than 1000 bytes.
1979
1980 7. A lesser problem with the same optimization is that it was recording the
1981 first character of an anchored pattern as "needed", thus provoking a search
1982 right along the subject, even when the first match of the pattern was going to
1983 fail. The "needed" character is now not set for anchored patterns, unless it
1984 follows something in the pattern that is of non-fixed length. Thus, it still
1985 fulfils its original purpose of finding quick non-matches in cases of nested
1986 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1987
1988
1989 Version 4.0 17-Feb-03
1990 ---------------------
1991
1992 1. If a comment in an extended regex that started immediately after a meta-item
1993 extended to the end of string, PCRE compiled incorrect data. This could lead to
1994 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1995
1996 2. Moved to autoconf 2.53 and libtool 1.4.2.
1997
1998 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1999 the special perltest8 script is no longer needed - all the tests can be run
2000 from a single perltest script.
2001
2002 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2003 by \s. It has now been removed in PCRE. This means it isn't recognized as
2004 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2005 class [:space:] *does* include VT, thereby creating a mess.
2006
2007 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2008 space and tab.
2009
2010 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2011 its new features into the main test script, reducing the number of scripts.
2012
2013 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2014 were backward compatible, and made the (?i) apply to the whole pattern, as if
2015 /i were given. Now it behaves more logically, and applies the option setting
2016 only to what follows. PCRE has been changed to follow suit. However, if it
2017 finds options settings right at the start of the pattern, it extracts them into
2018 the global options, as before. Thus, they show up in the info data.
2019
2020 8. Added support for the \Q...\E escape sequence. Characters in between are
2021 treated as literals. This is slightly different from Perl in that $ and @ are
2022 also handled as literals inside the quotes. In Perl, they will cause variable
2023 interpolation. Note the following examples:
2024
2025 Pattern PCRE matches Perl matches
2026
2027 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2028 \Qabc\$xyz\E abc\$xyz abc\$xyz
2029 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2030
2031 For compatibility with Perl, \Q...\E sequences are recognized inside character
2032 classes as well as outside them.
2033
2034 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2035 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2036 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2037 signed/unsigned warnings.
2038
2039 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2040 option for pcretest, so I've replaced it by a simple function that does just
2041 that job.
2042
2043 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2044 "pcregrep -".
2045
2046 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2047 Java package. This provides some syntactic sugar for simple cases of what my
2048 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2049 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2050 item, you can use this simplified notation. Note that only makes sense with
2051 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2052 greediness, whatever the setting of the PCRE_UNGREEDY option.
2053
2054 13. A change of greediness default within a pattern was not taking effect at
2055 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2056 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2057 was abstracted outside.
2058
2059 14. PCRE now supports the \G assertion. It is true when the current matching
2060 position is at the start point of the match. This differs from \A when the
2061 starting offset is non-zero. Used with the /g option of pcretest (or similar
2062 code), it works in the same way as it does for Perl's /g option. If all
2063 alternatives of a regex begin with \G, the expression is anchored to the start
2064 match position, and the "anchored" flag is set in the compiled expression.
2065
2066 15. Some bugs concerning the handling of certain option changes within patterns
2067 have been fixed. These applied to options other than (?ims). For example,
2068 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2069 been the other way round. Some of this was related to change 7 above.
2070
2071 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2072 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2073 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2074 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2075
2076 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2077 mode. Unlike ".", it always matches newline, whatever the setting of
2078 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2079 assertions. Perl allows it, but it doesn't (in general) work because it can't
2080 calculate the length of the lookbehind. At least, that's the case for Perl
2081 5.8.0 - I've been told they are going to document that it doesn't work in
2082 future.
2083
2084 18. Added an error diagnosis for escapes that PCRE does not support: these are
2085 \L, \l, \N, \P, \p, \U, \u, and \X.
2086
2087 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2088 reading past the end of the pattern in cases such as /[abcd/.
2089
2090 20. PCRE was getting more memory than necessary for patterns with classes that
2091 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2092
2093 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2094 compiling PCRE for use with Virtual Pascal.
2095
2096 22. Small fix to the Makefile to make it work properly if the build is done
2097 outside the source tree.
2098
2099 23. Added a new extension: a condition to go with recursion. If a conditional
2100 subpattern starts with (?(R) the "true" branch is used if recursion has
2101 happened, whereas the "false" branch is used only at the top level.
2102
2103 24. When there was a very long string of literal characters (over 255 bytes
2104 without UTF support, over 250 bytes with UTF support), the computation of how
2105 much memory was required could be incorrect, leading to segfaults or other
2106 strange effects.
2107
2108 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2109 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2110 there was a subsequent back reference to those brackets. This meant that, for
2111 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2112 possible to check for precisely this case. All we can do is abandon the
2113 optimization if .* occurs inside capturing brackets when there are any back
2114 references whatsoever. (See below for a better fix that came later.)
2115
2116 26. The handling of the optimization for finding the first character of a
2117 non-anchored pattern, and for finding a character that is required later in the
2118 match were failing in some cases. This didn't break the matching; it just
2119 failed to optimize when it could. The way this is done has been re-implemented.
2120
2121 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2122
2123 28. Added a new feature that provides some of the functionality that Perl
2124 provides with (?{...}). The facility is termed a "callout". The way it is done
2125 in PCRE is for the caller to provide an optional function, by setting
2126 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2127 global variable. By default it is unset, which disables all calling out. To get
2128 the function called, the regex must include (?C) at appropriate points. This
2129 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2130 This provides a means of identifying different callout points. When PCRE
2131 reaches such a point in the regex, if pcre_callout has been set, the external
2132 function is called. It is provided with data in a structure called
2133 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2134 matching continues; if it returns a non-zero value, the match at the current
2135 point fails. However, backtracking will occur if possible. [This was changed
2136 later and other features added - see item 49 below.]
2137
2138 29. pcretest is upgraded to test the callout functionality. It provides a
2139 callout function that displays information. By default, it shows the start of
2140 the match and the current position in the text. There are some new data escapes
2141 to vary what happens:
2142
2143 \C+ in addition, show current contents of captured substrings
2144 \C- do not supply a callout function
2145 \C!n return 1 when callout number n is reached
2146 \C!n!m return 1 when callout number n is reached for the mth time
2147
2148 30. If pcregrep was called with the -l option and just a single file name, it
2149 output "<stdin>" if a match was found, instead of the file name.
2150
2151 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2152 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2153 pcre_exec(). This saves a malloc/free per call. The default value of
2154 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2155 when configuring.
2156
2157 32. The default maximum size of a compiled pattern is 64K. There have been a
2158 few cases of people hitting this limit. The code now uses macros to handle the
2159 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2160 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2161 configuring. Tests 2 and 5 work only with 2-byte links because they output
2162 debugging information about compiled patterns.
2163
2164 33. Internal code re-arrangements:
2165
2166 (a) Moved the debugging function for printing out a compiled regex into
2167 its own source file (printint.c) and used #include to pull it into
2168 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2169 separate copies.
2170
2171 (b) Defined the list of op-code names for debugging as a macro in
2172 internal.h so that it is next to the definition of the opcodes.
2173
2174 (c) Defined a table of op-code lengths for simpler skipping along compiled
2175 code. This is again a macro in internal.h so that it is next to the
2176 definition of the opcodes.
2177
2178 34. Added support for recursive calls to individual subpatterns, along the
2179 lines of Robin Houston's patch (but implemented somewhat differently).
2180
2181 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2182 allow it to read and process whole directories in Win32. This code was
2183 contributed by Lionel Fourquaux; it has not been tested by me.
2184
2185 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2186 used to name a group. Names consist of alphanumerics and underscores, and must
2187 be unique. Back references use the syntax (?P=name) and recursive calls use
2188 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2189 numbers. The function pcre_fullinfo() can be used after compilation to extract
2190 a name/number map. There are three relevant calls:
2191
2192 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2193 PCRE_INFO_NAMECOUNT yields the number of entries
2194 PCRE_INFO_NAMETABLE yields a pointer to the map.
2195
2196 The map is a vector of fixed-size entries. The size of each entry depends on
2197 the length of the longest name used. The first two bytes of each entry are the
2198 group number, most significant byte first. There follows the corresponding
2199 name, zero terminated. The names are in alphabetical order.
2200
2201 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2202 case instead of 255. Making it the same both with and without UTF-8 support
2203 means that the same test output works with both.
2204
2205 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2206 calling malloc() with a zero argument.
2207
2208 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2209 optimization. I've improved things by keeping a bitmap of backreferences with
2210 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2211 fact referenced, the optimization can be applied. It is unlikely that a
2212 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2213 the match to follow \n) will appear inside brackets with a number greater than
2214 31, but if it does, any back reference > 31 suppresses the optimization.
2215
2216 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2217 of disabling numbered capturing parentheses. Any opening parenthesis that is
2218 not followed by ? behaves as if it were followed by ?: but named parentheses
2219 can still be used for capturing (and they will acquire numbers in the usual
2220 way).
2221
2222 41. Redesigned the return codes from the match() function into yes/no/error so
2223 that errors can be passed back from deep inside the nested calls. A malloc
2224 failure while inside a recursive subpattern call now causes the
2225 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2226
2227 42. It is now possible to set a limit on the number of times the match()
2228 function is called in a call to pcre_exec(). This facility makes it possible to
2229 limit the amount of recursion and backtracking, though not in a directly
2230 obvious way, because the match() function is used in a number of different
2231 circumstances. The count starts from zero for each position in the subject
2232 string (for non-anchored patterns). The default limit is, for compatibility, a
2233 large number, namely 10 000 000. You can change this in two ways:
2234
2235 (a) When configuring PCRE before making, you can use --with-match-limit=n
2236 to set a default value for the compiled library.
2237
2238 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2239 a different value is set. See 45 below.
2240
2241 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2242
2243 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2244 of things that can be changed at compile time. The first argument specifies
2245 what is wanted and the second points to where the information is to be placed.
2246 The current list of available information is:
2247
2248 PCRE_CONFIG_UTF8
2249
2250 The output is an integer that is set to one if UTF-8 support is available;
2251 otherwise it is set to zero.
2252
2253 PCRE_CONFIG_NEWLINE
2254
2255 The output is an integer that it set to the value of the code that is used for
2256 newline. It is either LF (10) or CR (13).
2257
2258 PCRE_CONFIG_LINK_SIZE
2259
2260 The output is an integer that contains the number of bytes used for internal
2261 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2262
2263 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2264
2265 The output is an integer that contains the threshold above which the POSIX
2266 interface uses malloc() for output vectors. See item 31 above.
2267
2268 PCRE_CONFIG_MATCH_LIMIT
2269
2270 The output is an unsigned integer that contains the default limit of the number
2271 of match() calls in a pcre_exec() execution. See 42 above.
2272
2273 44. pcretest has been upgraded by the addition of the -C option. This causes it
2274 to extract all the available output from the new pcre_config() function, and to
2275 output it. The program then exits immediately.
2276
2277 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2278 order to support additional features. One way would have been to define
2279 pcre_exec2() (for example) with extra arguments, but this would not have been
2280 extensible, and would also have required all calls to the original function to
2281 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2282 is used for passing in "extra" data from pcre_study().
2283
2284 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2285 contains the following fields:
2286
2287 flags a bitmap indicating which of the following fields are set
2288 study_data opaque data from pcre_study()
2289 match_limit a way of specifying a limit on match() calls for a specific
2290 call to pcre_exec()
2291 callout_data data for callouts (see 49 below)
2292
2293 The flag bits are also defined in pcre.h, and are
2294
2295 PCRE_EXTRA_STUDY_DATA
2296 PCRE_EXTRA_MATCH_LIMIT
2297 PCRE_EXTRA_CALLOUT_DATA
2298
2299 The pcre_study() function now returns one of these new pcre_extra blocks, with
2300 the actual study data pointed to by the study_data field, and the
2301 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2302 before. That is, this change is entirely upwards-compatible and requires no
2303 change to existing code.
2304
2305 If you want to pass in additional data to pcre_exec(), you can either place it
2306 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2307 block.
2308
2309 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2310 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2311 times with different match limits, until it finds the minimum value needed for
2312 pcre_exec() to complete. The value is then output. This can be instructive; for
2313 most simple matches the number is quite small, but for pathological cases it
2314 gets very large very quickly.
2315
2316 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2317 returns the size of the data block pointed to by the study_data field in a
2318 pcre_extra block, that is, the value that was passed as the argument to
2319 pcre_malloc() when PCRE was getting memory in which to place the information
2320 created by pcre_study(). The fourth argument should point to a size_t variable.
2321 pcretest has been extended so that this information is shown after a successful
2322 pcre_study() call when information about the compiled regex is being displayed.
2323
2324 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2325 because what follows is always an absolute path. (Later: it turns out that this
2326 is more than cosmetic for MinGW, because it doesn't like empty path
2327 components.)
2328
2329 49. Some changes have been made to the callout feature (see 28 above):
2330
2331 (i) A callout function now has three choices for what it returns:
2332
2333 0 => success, carry on matching
2334 > 0 => failure at this point, but backtrack if possible
2335 < 0 => serious error, return this value from pcre_exec()
2336
2337 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2338 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2339 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2340 use by callout functions. It will never be used by PCRE itself.
2341
2342 (ii) The pcre_extra structure (see 45 above) has a void * field called
2343 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2344 pcre_callout_block structure has a field of the same name. The contents of
2345 the field passed in the pcre_extra structure are passed to the callout
2346 function in the corresponding field in the callout block. This makes it
2347 easier to use the same callout-containing regex from multiple threads. For
2348 testing, the pcretest program has a new data escape
2349
2350 \C*n pass the number n (may be negative) as callout_data
2351
2352 If the callout function in pcretest receives a non-zero value as
2353 callout_data, it returns that value.
2354
2355 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2356 there were some redundant $(CFLAGS) in commands that are now specified as
2357 $(LINK), which already includes $(CFLAGS).
2358
2359 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2360 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2361 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2362 one-byte characters throughout. Note that case-insensitive matching applies
2363 only to characters whose values are less than 256. PCRE doesn't support the
2364 notion of cases for higher-valued characters.
2365
2366 (i) A character class whose characters are all within 0-255 is handled as
2367 a bit map, and the map is inverted for negative classes. Previously, a
2368 character > 255 always failed to match such a class; however it should
2369 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2370
2371 (ii) A negated character class with a single character < 255 is coded as
2372 "not this character" (OP_NOT). This wasn't working properly when the test
2373 character was multibyte, either singly or repeated.
2374
2375 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2376 mode, for example: \x{100}{2,3}.
2377
2378 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2379 singly or repeated) now correctly test multibyte characters. However,
2380 PCRE doesn't recognize any characters with values greater than 255 as
2381 digits, spaces, or word characters. Such characters always match \D, \S,
2382 and \W, and never match \d, \s, or \w.
2383
2384 (v) Classes may now contain characters and character ranges with values
2385 greater than 255. For example: [ab\x{100}-\x{400}].
2386
2387 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2388 PCRE in UTF-8 mode.
2389
2390 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2391 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2392 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2393 value.)
2394
2395 53. The single man page has become too large. I have therefore split it up into
2396 a number of separate man pages. These also give rise to individual HTML pages;
2397 these are now put in a separate directory, and there is an index.html page that
2398 lists them all. Some hyperlinking between the pages has been installed.
2399
2400 54. Added convenience functions for handling named capturing parentheses.
2401
2402 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2403 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2404 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2405 are faulted.
2406
2407 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2408 calling configure. These values are used when compiling the dftables.c program
2409 which is run to generate the source of the default character tables. They
2410 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2411 you will need to set these values.
2412
2413 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2414
2415
2416 Version 3.9 02-Jan-02
2417 ---------------------
2418
2419 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2420
2421 2. If --disable-static was given, the building process failed when trying to
2422 build pcretest and pcregrep. (For some reason it was using libtool to compile
2423 them, which is not right, as they aren't part of the library.)
2424
2425
2426 Version 3.8 18-Dec-01
2427 ---------------------
2428
2429 1. The experimental UTF-8 code was completely screwed up. It was packing the
2430 bytes in the wrong order. How dumb can you get?
2431
2432
2433 Version 3.7 29-Oct-01
2434 ---------------------
2435
2436 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2437 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2438 this didn't happen under Solaris 8, where I normally test things.
2439
2440 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2441 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2442 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2443 this fix an hour or so after the initial 3.7 release.)
2444
2445
2446 Version 3.6 23-Oct-01
2447 ---------------------
2448
2449 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2450 offsets passed as NULL with zero offset count.
2451
2452 2. The config.guess and config.sub files had not been updated when I moved to
2453 the latest autoconf.
2454
2455
2456 Version 3.5 15-Aug-01
2457 ---------------------
2458
2459 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2460 had been forgotten.
2461
2462 2. By using declared but undefined structures, we can avoid using "void"
2463 definitions in pcre.h while keeping the internal definitions of the structures
2464 private.
2465
2466 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2467 user point of view, this means that both static and shared libraries are built
2468 by default, but this can be individually controlled. More of the work of
2469 handling this static/shared cases is now inside libtool instead of PCRE's make
2470 file.
2471
2472 4. The pcretest utility is now installed along with pcregrep because it is
2473 useful for users (to test regexs) and by doing this, it automatically gets
2474 relinked by libtool. The documentation has been turned into a man page, so
2475 there are now .1, .txt, and .html versions in /doc.
2476
2477 5. Upgrades to pcregrep:
2478 (i) Added long-form option names like gnu grep.
2479 (ii) Added --help to list all options with an explanatory phrase.
2480 (iii) Added -r, --recursive to recurse into sub-directories.
2481 (iv) Added -f, --file to read patterns from a file.
2482
2483 6. pcre_exec() was referring to its "code" argument before testing that
2484 argument for NULL (and giving an error if it was NULL).
2485
2486 7. Upgraded Makefile.in to allow for compiling in a different directory from
2487 the source directory.
2488
2489 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2490 options bits, the pointer it was passed was to an int instead of to an unsigned
2491 long int. This mattered only on 64-bit systems.
2492
2493 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2494 generated) instead of pcre.in, which it its source. Also made the same change
2495 in several of the .c files.
2496
2497 10. A new release of gcc defines printf() as a macro, which broke pcretest
2498 because it had an ifdef in the middle of a string argument for printf(). Fixed
2499 by using separate calls to printf().
2500
2501 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2502 script, to force use of CR or LF instead of \n in the source. On non-Unix
2503 systems, the value can be set in config.h.
2504
2505 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2506 absolute limit. Changed the text of the error message to make this clear, and
2507 likewise updated the man page.
2508
2509 13. The limit of 99 on the number of capturing subpatterns has been removed.
2510 The new limit is 65535, which I hope will not be a "real" limit.
2511
2512
2513 Version 3.4 22-Aug-00
2514 ---------------------
2515
2516 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2517
2518 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2519
2520
2521 Version 3.3 01-Aug-00
2522 ---------------------
2523
2524 1. If an octal character was given, but the value was greater than \377, it
2525 was not getting masked to the least significant bits, as documented. This could
2526 lead to crashes in some systems.
2527
2528 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2529 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2530
2531 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2532 These just pass their arguments on to (pcre_free)(), but they are provided
2533 because some uses of PCRE bind it to non-C systems that can call its functions,
2534 but cannot call free() or pcre_free() directly.
2535
2536 4. Add "make test" as a synonym for "make check". Corrected some comments in
2537 the Makefile.
2538
2539 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2540 Makefile.
2541
2542 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2543 command called pgrep for grepping around the active processes.
2544
2545 7. Added the beginnings of support for UTF-8 character strings.
2546
2547 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2548 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2549 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2550 out for the ar command.)
2551
2552
2553 Version 3.2 12-May-00
2554 ---------------------
2555
2556 This is purely a bug fixing release.
2557
2558 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2559 of ZA. This was just one example of several cases that could provoke this bug,
2560 which was introduced by change 9 of version 2.00. The code for breaking
2561 infinite loops after an iteration that matches an empty string was't working
2562 correctly.
2563
2564 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2565 when matched against abbab (for example). After matching an empty string, it
2566 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2567 caused it to match further down the string than it should.
2568
2569 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2570 was there because it doesn't seem to be needed, and it causes trouble on some
2571 systems, as it is not a Standard C header. It has been removed.
2572
2573 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2574 were reported on the Macintosh. The changes were from
2575
2576 while ((c = *(++ptr)) != 0 && c != '\n');
2577 to
2578 while ((c = *(++ptr)) != 0 && c != '\n') ;
2579
2580 Totally extraordinary, but if that's what it takes...
2581
2582 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2583 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2584 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2585 assumes the way PCRE uses memmove() (always moving upwards).
2586
2587 6. PCRE is being used in one environment where strchr() is not available. There
2588 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2589 faster code anyway.
2590
2591
2592 Version 3.1 09-Feb-00
2593 ---------------------
2594
2595 The only change in this release is the fixing of some bugs in Makefile.in for
2596 the "install" target:
2597
2598 (1) It was failing to install pcreposix.h.
2599
2600 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2601
2602
2603 Version 3.0 01-Feb-00
2604 ---------------------
2605
2606 1. Add support for the /+ modifier to perltest (to output $` like it does in
2607 pcretest).
2608
2609 2. Add support for the /g modifier to perltest.
2610
2611 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2612 matches null strings.
2613
2614 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2615 pattern. Perl treats empty patterns specially - it reuses the most recent
2616 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2617 effect.
2618
2619 5. The POSIX interface was broken in that it was just handing over the POSIX
2620 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2621 required a bigger vector, with some working space on the end. This means that
2622 the POSIX wrapper now has to get and free some memory, and copy the results.
2623
2624 6. Added some simple autoconf support, placing the test data and the
2625 documentation in separate directories, re-organizing some of the
2626 information files, and making it build pcre-config (a GNU standard). Also added
2627 libtool support for building PCRE as a shared library, which is now the
2628 default.
2629
2630 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2631 09 are not valid octal constants. Single digits will be used for minor values
2632 less than 10.
2633
2634 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2635 existing programs that set these in the POSIX interface can use PCRE without
2636 modification.
2637
2638 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2639 return all that pcre_info() returns, plus additional data. The pcre_info()
2640 function is retained for compatibility, but is considered to be obsolete.
2641
2642 10. Added experimental recursion feature (?R) to handle one common case that
2643 Perl 5.6 will be able to do with (?p{...}).
2644
2645 11. Added support for POSIX character classes like [:alpha:], which Perl is
2646 adopting.
2647
2648
2649 Version 2.08 31-Aug-99
2650 ----------------------
2651
2652 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2653 trying to match at the startoffset position, but instead was moving forward to
2654 the next newline as if a previous match had failed.
2655
2656 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2657 and could get into a loop if a null string was matched other than at the start
2658 of the subject.
2659
2660 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2661 be distinguished at compile time, and for completeness also added PCRE_DATE.
2662
2663 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2664 in GnuWin32 environments.
2665
2666
2667 Version 2.07 29-Jul-99
2668 ----------------------
2669
2670 1. The documentation is now supplied in plain text form and HTML as well as in
2671 the form of man page sources.
2672
2673 2. C++ compilers don't like assigning (void *) values to other pointer types.
2674 In particular this affects malloc(). Although there is no problem in Standard
2675 C, I've put in casts to keep C++ compilers happy.
2676
2677 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2678 should be (const char *).
2679
2680 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2681 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2682 However, I haven't made this a standard facility. The documentation doesn't
2683 mention it, and the Makefile doesn't support it.
2684
2685 5. The Makefile now contains an "install" target, with editable destinations at
2686 the top of the file. The pcretest program is not installed.
2687
2688 6. pgrep -V now gives the PCRE version number and date.
2689
2690 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2691 causing the entire string to be ignored, instead of just the last character.
2692
2693 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2694 non-matching string, it can take a very, very long time, even for strings of
2695 quite modest length, because of the nested recursion. PCRE now does better in
2696 some of these cases. It does this by remembering the last required literal
2697 character in the pattern, and pre-searching the subject to ensure it is present
2698 before running the real match. In other words, it applies a heuristic to detect
2699 some types of certain failure quickly, and in the above example, if presented
2700 with a string that has no trailing " it gives "no match" very quickly.
2701
2702 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2703 other alternatives are tried instead.
2704
2705
2706 Version 2.06 09-Jun-99
2707 ----------------------
2708
2709 1. Change pcretest's output for amount of store used to show just the code
2710 space, because the remainder (the data block) varies in size between 32-bit and
2711 64-bit systems.
2712
2713 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2714 start matching at. This allows lookbehinds to work when searching for multiple
2715 occurrences in a string.
2716
2717 3. Added additional options to pcretest for testing multiple occurrences:
2718
2719 /+ outputs the rest of the string that follows a match
2720 /g loops for multiple occurrences, using the new startoffset argument
2721 /G loops for multiple occurrences by passing an incremented pointer
2722
2723 4. PCRE wasn't doing the "first character" optimization for patterns starting
2724 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2725 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2726 the letter 'x'. On long subject strings, this gives a significant speed-up.
2727
2728
2729 Version 2.05 21-Apr-99
2730 ----------------------
2731
2732 1. Changed the type of magic_number from int to long int so that it works
2733 properly on 16-bit systems.
2734
2735 2. Fixed a bug which caused patterns starting with .* not to work correctly
2736 when the subject string contained newline characters. PCRE was assuming
2737 anchoring for such patterns in all cases, which is not correct because .* will
2738 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2739 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2740 must be retried after every newline in the subject.
2741
2742
2743 Version 2.04 18-Feb-99
2744 ----------------------
2745
2746 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2747 computation of the store needed to hold the pattern was incorrect (too large).
2748 If such patterns were nested a few deep, this could multiply and become a real
2749 problem.
2750
2751 2. Added /M option to pcretest to show the memory requirement of a specific
2752 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2753
2754 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2755 compiled in such a way that the backtracking after subsequent failure was
2756 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2757 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2758
2759
2760 Version 2.03 02-Feb-99
2761 ----------------------
2762
2763 1. Fixed typo and small mistake in man page.
2764
2765 2. Added 4th condition (GPL supersedes if conflict) and created separate
2766 LICENCE file containing the conditions.
2767
2768 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2769 Perl, that is the internal \ allows the delimiter to be included in the
2770 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2771 the final delimiter, add \ to the end of the pattern (to test the error).
2772
2773 4. Added the convenience functions for extracting substrings after a successful
2774 match. Updated pcretest to make it able to test these functions.
2775
2776
2777 Version 2.02 14-Jan-99
2778 ----------------------
2779
2780 1. Initialized the working variables associated with each extraction so that
2781 their saving and restoring doesn't refer to uninitialized store.
2782
2783 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2784 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2785 fix the problem.
2786
2787 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2788 calls, and wasn't printing the correct value for compiling calls. Increased the
2789 default value of LOOPREPEAT, and the number of significant figures in the
2790 times.
2791
2792 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2793
2794 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2795 a building problem on Windows NT with a FAT file system.
2796
2797
2798 Version 2.01 21-Oct-98
2799 ----------------------
2800
2801 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2802 to character tables built by pcre_maketables() in the current locale. If NULL
2803 is passed, the default tables are used.
2804
2805
2806 Version 2.00 24-Sep-98
2807 ----------------------
2808
2809 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2810 it any more.
2811
2812 2. Allow quantification of (?>) groups, and make it work correctly.
2813
2814 3. The first character computation wasn't working for (?>) groups.
2815
2816 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2817 end of the subject) and add 5.005's \z, which really does match only at the
2818 very end of the subject.
2819
2820 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2821
2822 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2823 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2824 localized options. All options to pcre_study() were also removed.
2825
2826 7. Add other new features from 5.005:
2827
2828 $(?<= positive lookbehind
2829 $(?<! negative lookbehind
2830 (?imsx-imsx) added the unsetting capability
2831 such a setting is global if at outer level; local otherwise
2832 (?imsx-imsx:) non-capturing groups with option setting
2833 (?(cond)re|re) conditional pattern matching
2834
2835 A backreference to itself in a repeated group matches the previous
2836 captured string.
2837
2838 8. General tidying up of studying (both automatic and via "study")
2839 consequential on the addition of new assertions.
2840
2841 9. As in 5.005, unlimited repeated groups that could match an empty substring
2842 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2843 runtime if any iteration does actually match an empty substring.
2844
2845 10. Include the RunTest script in the distribution.
2846
2847 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2848 discrepancies, some of which were old and were also with respect to 5.004. They
2849 have now been fixed.
2850
2851
2852 Version 1.09 28-Apr-98
2853 ----------------------
2854
2855 1. A negated single character class followed by a quantifier with a minimum
2856 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2857 program crashes, or just wrong answers. This did not apply to negated classes
2858 containing more than one character, or to minima other than one.
2859
2860
2861 Version 1.08 27-Mar-98
2862 ----------------------
2863
2864 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2865
2866 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2867 latter must appear before anything that relies on it in the pattern.
2868
2869
2870 Version 1.07 16-Feb-98
2871 ----------------------
2872
2873 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2874 repeat of a potentially empty string).
2875
2876
2877 Version 1.06 23-Jan-98
2878 ----------------------
2879
2880 1. Added Markus Oberhumer's little patches for C++.
2881
2882 2. Literal strings longer than 255 characters were broken.
2883
2884
2885 Version 1.05 23-Dec-97
2886 ----------------------
2887
2888 1. Negated character classes containing more than one character were failing if
2889 PCRE_CASELESS was set at run time.
2890
2891
2892 Version 1.04 19-Dec-97
2893 ----------------------
2894
2895 1. Corrected the man page, where some "const" qualifiers had been omitted.
2896
2897 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2898 input syntax.
2899
2900 3. Fixed memory leak which occurred when a regex with back references was
2901 matched with an offsets vector that wasn't big enough. The temporary memory
2902 that is used in this case wasn't being freed if the match failed.
2903
2904 4. Tidied pcretest to ensure it frees memory that it gets.
2905
2906 5. Temporary memory was being obtained in the case where the passed offsets
2907 vector was exactly big enough.
2908
2909 6. Corrected definition of offsetof() from change 5 below.
2910
2911 7. I had screwed up change 6 below and broken the rules for the use of
2912 setjmp(). Now fixed.
2913
2914
2915 Version 1.03 18-Dec-97
2916 ----------------------
2917
2918 1. A erroneous regex with a missing opening parenthesis was correctly
2919 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2920 on some systems.
2921
2922 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2923 it was reported that one broken compiler failed on the former because "code" is
2924 also an independent variable.
2925
2926 3. The erroneous regex a[]b caused an array overrun reference.
2927
2928 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2929 fail on data ending with that character. (It was going on too far, and checking
2930 the next character, typically a binary zero.) This was specific to the
2931 optimized code for single-character negative classes.
2932
2933 5. Added a contributed patch from the TIN world which does the following:
2934
2935 + Add an undef for memmove, in case the the system defines a macro for it.
2936
2937 + Add a definition of offsetof(), in case there isn't one. (I don't know
2938 the reason behind this - offsetof() is part of the ANSI standard - but
2939 it does no harm).
2940
2941 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2942 most of the places where whitespace preceded '#'. I have given up and
2943 allowed the remaining 2 cases to be at the margin.
2944
2945 + Rename some variables in pcre to eliminate shadowing. This seems very
2946 pedantic, but does no harm, of course.
2947
2948 6. Moved the call to setjmp() into its own function, to get rid of warnings
2949 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2950
2951 7. Constructs such as \d{8,} were compiling into the equivalent of
2952 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2953 outcome, but in this particular case used more store than had been allocated,
2954 which caused the bug to be discovered because it threw up an internal error.
2955
2956 8. The debugging code in both pcre and pcretest for outputting the compiled
2957 form of a regex was going wrong in the case of back references followed by
2958 curly-bracketed repeats.
2959
2960
2961 Version 1.02 12-Dec-97
2962 ----------------------
2963
2964 1. Typos in pcre.3 and comments in the source fixed.
2965
2966 2. Applied a contributed patch to get rid of places where it used to remove
2967 'const' from variables, and fixed some signed/unsigned and uninitialized
2968 variable warnings.
2969
2970 3. Added the "runtest" target to Makefile.
2971
2972 4. Set default compiler flag to -O2 rather than just -O.
2973
2974
2975 Version 1.01 19-Nov-97
2976 ----------------------
2977
2978 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2979 like /([ab]*)*/, that is, for classes with more than one character in them.
2980
2981 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2982 as /((?>a*))*/ (a PCRE_EXTRA facility).
2983
2984
2985 Version 1.00 18-Nov-97
2986 ----------------------
2987
2988 1. Added compile-time macros to support systems such as SunOS4 which don't have
2989 memmove() or strerror() but have other things that can be used instead.
2990
2991 2. Arranged that "make clean" removes the executables.
2992
2993
2994 Version 0.99 27-Oct-97
2995 ----------------------
2996
2997 1. Fixed bug in code for optimizing classes with only one character. It was
2998 initializing a 32-byte map regardless, which could cause it to run off the end
2999 of the memory it had got.
3000
3001 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3002
3003
3004 Version 0.98 22-Oct-97
3005 ----------------------
3006
3007 1. Fixed bug in code for handling temporary memory usage when there are more
3008 back references than supplied space in the ovector. This could cause segfaults.
3009
3010
3011 Version 0.97 21-Oct-97
3012 ----------------------
3013
3014 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3015
3016 2. Optimized negated single characters not to use a bit map.
3017
3018 3. Brought error texts together as macro definitions; clarified some of them;
3019 fixed one that was wrong - it said "range out of order" when it meant "invalid
3020 escape sequence".
3021
3022 4. Changed some char * arguments to const char *.
3023
3024 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3025
3026 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3027 pcretest.
3028
3029
3030 Version 0.96 16-Oct-97
3031 ----------------------
3032
3033 1. Added a simple "pgrep" utility to the distribution.
3034
3035 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3036 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3037 where "ddd" means "one or more decimal digits".
3038
3039 3. Fixed serious bug. If a pattern had a back reference, but the call to
3040 pcre_exec() didn't supply a large enough ovector to record the related
3041 identifying subpattern, the match always failed. PCRE now remembers the number
3042 of the largest back reference, and gets some temporary memory in which to save
3043 the offsets during matching if necessary, in order to ensure that
3044 backreferences always work.
3045
3046 4. Increased the compatibility with Perl in a number of ways:
3047
3048 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3049 to request this handling. The option can be set at compile or exec time.
3050
3051 (b) $ matches before a terminating newline by default; an option
3052 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3053 mode). The option can be set at compile or exec time.
3054
3055 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3056 the same as Perl's. If the decimal number it represents is less than 10
3057 or there aren't that many previous left capturing parentheses, an octal
3058 escape is read. Inside a character class, it's always an octal escape,
3059 even if it is a single digit.
3060
3061 (d) An escaped but undefined alphabetic character is taken as a literal,
3062 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3063 escapes.
3064
3065 (e) {0} is now permitted. (The previous item is removed from the compiled
3066 pattern).
3067
3068 5. Changed all the names of code files so that the basic parts are no longer
3069 than 10 characters, and abolished the teeny "globals.c" file.
3070
3071 6. Changed the handling of character classes; they are now done with a 32-byte
3072 bit map always.
3073
3074 7. Added the -d and /D options to pcretest to make it possible to look at the
3075 internals of compilation without having to recompile pcre.
3076
3077
3078 Version 0.95 23-Sep-97
3079 ----------------------
3080
3081 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3082 \x20 at the start of a run of normal characters. These were being treated as
3083 real characters, instead of the source characters being re-checked.
3084
3085
3086 Version 0.94 18-Sep-97
3087 ----------------------
3088
3089 1. The functions are now thread-safe, with the caveat that the global variables
3090 containing pointers to malloc() and free() or alternative functions are the
3091 same for all threads.
3092
3093 2. Get pcre_study() to generate a bitmap of initial characters for non-
3094 anchored patterns when this is possible, and use it if passed to pcre_exec().
3095
3096
3097 Version 0.93 15-Sep-97
3098 ----------------------
3099
3100 1. /(b)|(:+)/ was computing an incorrect first character.
3101
3102 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3103 but not actually doing anything yet.
3104
3105 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3106 as Perl does (e.g. [-az] or [az-]).
3107
3108 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3109 all possible positions.
3110
3111 5. Split up into different modules to avoid including unneeded functions in a
3112 compiled binary. However, compile and exec are still in one module. The "study"
3113 function is split off.
3114
3115 6. The character tables are now in a separate module whose source is generated
3116 by an auxiliary program - but can then be edited by hand if required. There are
3117 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3118 toupper() in the code.
3119
3120 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3121 make them global. Abolish the function for setting them, as the caller can now
3122 set them directly.
3123
3124
3125 Version 0.92 11-Sep-97
3126 ----------------------
3127
3128 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3129 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3130
3131 2. Caseless matching was not working in character classes if the characters in
3132 the pattern were in upper case.
3133
3134 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3135
3136 4. Make PCRE_ANCHORED public and accept as a compile option.
3137
3138 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3139 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3140 pass them.
3141
3142 6. Give an error if bad option bits passed at compile or run time.
3143
3144 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3145 pcretest to cause it to pass that flag.
3146
3147 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3148 options, and the first character, if set.
3149
3150 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3151
3152
3153 Version 0.91 10-Sep-97
3154 ----------------------
3155
3156 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3157 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3158
3159 2. PCRE was looping on encountering an indefinitely repeated back reference to
3160 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3161 Perl does - treats the match as successful.
3162
3163 ****

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