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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 307 - (show annotations) (download)
Sun Jan 20 20:36:45 2008 UTC (6 years, 11 months ago) by ph10
File size: 149597 byte(s)
Add Ron Aaron's Makefile patch (experimentally).

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 3. Applied Christian Ehrlicher's patch to update the CMake build files to
15 bring them up to date and include new features. This patch includes:
16
17 - Fixed PH's badly added libz and libbz2 support.
18 - Fixed a problem with static linking.
19 - Added pcredemo.
20 - Fixed dftables problem and added an option.
21 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
22 HAVE_LONG_LONG.
23 - Added readline support for pcretest.
24 - Added an listing of the option settings after cmake has run.
25
26 4. A user submitted a patch to Makefile that makes it easy to created a dll
27 under mingw. I added stuff to Makefile.am that cause it to include this
28 special target, without affecting anything else.
29
30
31 Version 7.5 10-Jan-08
32 ---------------------
33
34 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
35 values in parens when parsing an RE using the C++ wrapper."
36
37 2. Negative specials like \S did not work in character classes in UTF-8 mode.
38 Characters greater than 255 were excluded from the class instead of being
39 included.
40
41 3. The same bug as (2) above applied to negated POSIX classes such as
42 [:^space:].
43
44 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
45 defined or documented. It seems to have been a typo for PCRE_STATIC, so
46 I have changed it.
47
48 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
49 first named subpattern) and a construct such as (?&a) would reference the
50 first named subpattern whose name started with "a" (in other words, the
51 length check was missing). Both these problems are fixed. "Subpattern name
52 expected" is now given for (?&) (a zero-length name), and this patch also
53 makes it give the same error for \k'' (previously it complained that that
54 was a reference to a non-existent subpattern).
55
56 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
57 this is right because (?- can be followed by option settings as well as by
58 digits. I have, however, made the messages clearer.
59
60 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
61 than the number used in the conditional) now cause a compile-time error.
62 This is actually not compatible with Perl, which accepts such patterns, but
63 treats the conditional as always being FALSE (as PCRE used to), but it
64 seems to me that giving a diagnostic is better.
65
66 8. Change "alphameric" to the more common word "alphanumeric" in comments
67 and messages.
68
69 9. Fix two occurrences of "backslash" in comments that should have been
70 "backspace".
71
72 10. Remove two redundant lines of code that can never be obeyed (their function
73 was moved elsewhere).
74
75 11. The program that makes PCRE's Unicode character property table had a bug
76 which caused it to generate incorrect table entries for sequences of
77 characters that have the same character type, but are in different scripts.
78 It amalgamated them into a single range, with the script of the first of
79 them. In other words, some characters were in the wrong script. There were
80 thirteen such cases, affecting characters in the following ranges:
81
82 U+002b0 - U+002c1
83 U+0060c - U+0060d
84 U+0061e - U+00612
85 U+0064b - U+0065e
86 U+0074d - U+0076d
87 U+01800 - U+01805
88 U+01d00 - U+01d77
89 U+01d9b - U+01dbf
90 U+0200b - U+0200f
91 U+030fc - U+030fe
92 U+03260 - U+0327f
93 U+0fb46 - U+0fbb1
94 U+10450 - U+1049d
95
96 12. The -o option (show only the matching part of a line) for pcregrep was not
97 compatible with GNU grep in that, if there was more than one match in a
98 line, it showed only the first of them. It now behaves in the same way as
99 GNU grep.
100
101 13. If the -o and -v options were combined for pcregrep, it printed a blank
102 line for every non-matching line. GNU grep prints nothing, and pcregrep now
103 does the same. The return code can be used to tell if there were any
104 non-matching lines.
105
106 14. Added --file-offsets and --line-offsets to pcregrep.
107
108 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
109 infinitely looping recursion. The bug was that positive lookaheads were not
110 being skipped when checking for a possible empty match (negative lookaheads
111 and both kinds of lookbehind were skipped).
112
113 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
114 inclusion of <windows.h> to before rather than after the definition of
115 INVALID_FILE_ATTRIBUTES (patch from David Byron).
116
117 17. Specifying a possessive quantifier with a specific limit for a Unicode
118 character property caused pcre_compile() to compile bad code, which led at
119 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
120 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
121 caused the error; without that there was no problem.
122
123 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
124
125 19. Added --enable-pcretest-libreadline.
126
127 20. In pcrecpp.cc, the variable 'count' was incremented twice in
128 RE::GlobalReplace(). As a result, the number of replacements returned was
129 double what it should be. I removed one of the increments, but Craig sent a
130 later patch that removed the other one (the right fix) and added unit tests
131 that check the return values (which was not done before).
132
133 21. Several CMake things:
134
135 (1) Arranged that, when cmake is used on Unix, the libraries end up with
136 the names libpcre and libpcreposix, not just pcre and pcreposix.
137
138 (2) The above change means that pcretest and pcregrep are now correctly
139 linked with the newly-built libraries, not previously installed ones.
140
141 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
142
143 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
144 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
145 UTF-8 newline character). The key issue is that the pattern starts .*;
146 this means that the match must be either at the beginning, or after a
147 newline. The bug was in the code for advancing after a failed match and
148 checking that the new position followed a newline. It was not taking
149 account of UTF-8 characters correctly.
150
151 23. PCRE was behaving differently from Perl in the way it recognized POSIX
152 character classes. PCRE was not treating the sequence [:...:] as a
153 character class unless the ... were all letters. Perl, however, seems to
154 allow any characters between [: and :], though of course it rejects as
155 unknown any "names" that contain non-letters, because all the known class
156 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
157 for example, whereas PCRE did not - it did not recognize a POSIX character
158 class. This seemed a bit dangerous, so the code has been changed to be
159 closer to Perl. The behaviour is not identical to Perl, because PCRE will
160 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
161 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
162 Perl does, and where it didn't before.
163
164 24. Rewrite so as to remove the single use of %n from pcregrep because in some
165 Windows environments %n is disabled by default.
166
167
168 Version 7.4 21-Sep-07
169 ---------------------
170
171 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
172 means that a class such as [\s] counted as "explicit reference to CR or
173 LF". That isn't really right - the whole point of the change was to try to
174 help when there was an actual mention of one of the two characters. So now
175 the change happens only if \r or \n (or a literal CR or LF) character is
176 encountered.
177
178 2. The 32-bit options word was also used for 6 internal flags, but the numbers
179 of both had grown to the point where there were only 3 bits left.
180 Fortunately, there was spare space in the data structure, and so I have
181 moved the internal flags into a new 16-bit field to free up more option
182 bits.
183
184 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
185 but did not set the internal JCHANGED flag - either of these is enough to
186 control the way the "get" function works - but the PCRE_INFO_JCHANGED
187 facility is supposed to tell if (?J) was ever used, so now (?J) at the
188 start sets both bits.
189
190 4. Added options (at build time, compile time, exec time) to change \R from
191 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
192
193 5. doc/pcresyntax.html was missing from the distribution.
194
195 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
196 compatibility, even though it is no longer used.
197
198 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
199 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
200 windows.h file is present (where different names are used). [This was
201 reversed later after testing - see 16 below.]
202
203 8. Changed all #include <config.h> to #include "config.h". There were also
204 some further <pcre.h> cases that I changed to "pcre.h".
205
206 9. When pcregrep was used with the --colour option, it missed the line ending
207 sequence off the lines that it output.
208
209 10. It was pointed out to me that arrays of string pointers cause lots of
210 relocations when a shared library is dynamically loaded. A technique of
211 using a single long string with a table of offsets can drastically reduce
212 these. I have refactored PCRE in four places to do this. The result is
213 dramatic:
214
215 Originally: 290
216 After changing UCP table: 187
217 After changing error message table: 43
218 After changing table of "verbs" 36
219 After changing table of Posix names 22
220
221 Thanks to the folks working on Gregex for glib for this insight.
222
223 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
224 unicode-properties was also set.
225
226 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
227
228 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
229 checked only for CRLF.
230
231 14. Added casts to pcretest.c to avoid compiler warnings.
232
233 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
234
235 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
236 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
237 entirely. This removes changes made in 7 above.
238
239 17. The CMake files have been updated, and there is now more information about
240 building with CMake in the NON-UNIX-USE document.
241
242
243 Version 7.3 28-Aug-07
244 ---------------------
245
246 1. In the rejigging of the build system that eventually resulted in 7.1, the
247 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
248 brackets there is not right, since it causes compilers to look for an
249 installed pcre.h, not the version that is in the source that is being
250 compiled (which of course may be different). I have changed it back to:
251
252 #include "pcre.h"
253
254 I have a vague recollection that the change was concerned with compiling in
255 different directories, but in the new build system, that is taken care of
256 by the VPATH setting the Makefile.
257
258 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
259 when the subject happened to end in the byte 0x85 (e.g. if the last
260 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
261 characters but of course it shouldn't be taken as a newline when it is part
262 of another character. The bug was that, for an unlimited repeat of . in
263 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
264 characters when looking for a newline.
265
266 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
267
268 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
269 in debug output.
270
271 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
272 long printing in the pcrecpp unittest when running under MinGW.
273
274 6. ESC_K was left out of the EBCDIC table.
275
276 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
277 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
278 limit also applies to "virtual nesting" when a pattern is recursive, and in
279 this case 1000 isn't so big. I have been able to remove this limit at the
280 expense of backing off one optimization in certain circumstances. Normally,
281 when pcre_exec() would call its internal match() function recursively and
282 immediately return the result unconditionally, it uses a "tail recursion"
283 feature to save stack. However, when a subpattern that can match an empty
284 string has an unlimited repetition quantifier, it no longer makes this
285 optimization. That gives it a stack frame in which to save the data for
286 checking that an empty string has been matched. Previously this was taken
287 from the 1000-entry workspace that had been reserved. So now there is no
288 explicit limit, but more stack is used.
289
290 8. Applied Daniel's patches to solve problems with the import/export magic
291 syntax that is required for Windows, and which was going wrong for the
292 pcreposix and pcrecpp parts of the library. These were overlooked when this
293 problem was solved for the main library.
294
295 9. There were some crude static tests to avoid integer overflow when computing
296 the size of patterns that contain repeated groups with explicit upper
297 limits. As the maximum quantifier is 65535, the maximum group length was
298 set at 30,000 so that the product of these two numbers did not overflow a
299 32-bit integer. However, it turns out that people want to use groups that
300 are longer than 30,000 bytes (though not repeat them that many times).
301 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
302 made it possible to implement the integer overflow checks in a much more
303 dynamic way, which I have now done. The artificial limitation on group
304 length has been removed - we now have only the limit on the total length of
305 the compiled pattern, which depends on the LINK_SIZE setting.
306
307 10. Fixed a bug in the documentation for get/copy named substring when
308 duplicate names are permitted. If none of the named substrings are set, the
309 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
310 empty string.
311
312 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
313 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
314 because the ] is interpreted as the first data character and the
315 terminating ] is not found. PCRE has been made compatible with Perl in this
316 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
317 cause memory overwriting.
318
319 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
320 string has been matched (to stop an infinite loop). It was not recognizing
321 a conditional subpattern that could match an empty string if that
322 subpattern was within another subpattern. For example, it looped when
323 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
324 condition was not nested. This bug has been fixed.
325
326 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
327 past the start of the subject in the presence of bytes with the top bit
328 set, for example "\x8aBCD".
329
330 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
331 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
332
333 14. Optimized (?!) to (*FAIL).
334
335 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
336 This restricts code points to be within the range 0 to 0x10FFFF, excluding
337 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
338 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
339 does: it's just the validity check that is more restrictive.
340
341 16. Inserted checks for integer overflows during escape sequence (backslash)
342 processing, and also fixed erroneous offset values for syntax errors during
343 backslash processing.
344
345 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
346 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
347
348 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
349 caused an overrun.
350
351 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
352 something other than just ASCII characters) inside a group that had an
353 unlimited repeat caused a loop at compile time (while checking to see
354 whether the group could match an empty string).
355
356 20. Debugging a pattern containing \p or \P could cause a crash. For example,
357 [\P{Any}] did so. (Error in the code for printing property names.)
358
359 21. An orphan \E inside a character class could cause a crash.
360
361 22. A repeated capturing bracket such as (A)? could cause a wild memory
362 reference during compilation.
363
364 23. There are several functions in pcre_compile() that scan along a compiled
365 expression for various reasons (e.g. to see if it's fixed length for look
366 behind). There were bugs in these functions when a repeated \p or \P was
367 present in the pattern. These operators have additional parameters compared
368 with \d, etc, and these were not being taken into account when moving along
369 the compiled data. Specifically:
370
371 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
372 length.
373
374 (b) An item such as \pL+ within a repeated group could cause crashes or
375 loops.
376
377 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
378 "reference to non-existent subpattern" error.
379
380 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
381
382 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
383 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
384
385 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
386
387 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
388 character were causing crashes (broken optimization).
389
390 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
391 \p or \P) caused a compile-time loop.
392
393 28. More problems have arisen in unanchored patterns when CRLF is a valid line
394 break. For example, the unstudied pattern [\r\n]A does not match the string
395 "\r\nA" because change 7.0/46 below moves the current point on by two
396 characters after failing to match at the start. However, the pattern \nA
397 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
398 the same is true. There doesn't seem any very clean way out of this, but
399 what I have chosen to do makes the common cases work: PCRE now takes note
400 of whether there can be an explicit match for \r or \n anywhere in the
401 pattern, and if so, 7.0/46 no longer applies. As part of this change,
402 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
403 pattern has explicit CR or LF references.
404
405 29. Added (*CR) etc for changing newline setting at start of pattern.
406
407
408 Version 7.2 19-Jun-07
409 ---------------------
410
411 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
412 which is apparently normally available under Windows.
413
414 2. Re-jig the pcregrep tests with different newline settings in an attempt
415 to make them independent of the local environment's newline setting.
416
417 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
418
419 4. Some of the "internals" tests were previously cut out when the link size
420 was not 2, because the output contained actual offsets. The recent new
421 "Z" feature of pcretest means that these can be cut out, making the tests
422 usable with all link sizes.
423
424 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
425 stack recursion. This gives a massive performance boost under BSD, but just
426 a small improvement under Linux. However, it saves one field in the frame
427 in all cases.
428
429 6. Added more features from the forthcoming Perl 5.10:
430
431 (a) (?-n) (where n is a string of digits) is a relative subroutine or
432 recursion call. It refers to the nth most recently opened parentheses.
433
434 (b) (?+n) is also a relative subroutine call; it refers to the nth next
435 to be opened parentheses.
436
437 (c) Conditions that refer to capturing parentheses can be specified
438 relatively, for example, (?(-2)... or (?(+3)...
439
440 (d) \K resets the start of the current match so that everything before
441 is not part of it.
442
443 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
444
445 (f) \g{name} is another synonym - part of Perl 5.10's unification of
446 reference syntax.
447
448 (g) (?| introduces a group in which the numbering of parentheses in each
449 alternative starts with the same number.
450
451 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
452
453 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
454 PCRE_INFO_JCHANGED.
455
456 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
457 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
458 for detecting groups that can match an empty string.
459
460 9. A pattern with a very large number of alternatives (more than several
461 hundred) was running out of internal workspace during the pre-compile
462 phase, where pcre_compile() figures out how much memory will be needed. A
463 bit of new cunning has reduced the workspace needed for groups with
464 alternatives. The 1000-alternative test pattern now uses 12 bytes of
465 workspace instead of running out of the 4096 that are available.
466
467 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
468
469 11. Applied patch from Google to remove an optimization that didn't quite work.
470 The report of the bug said:
471
472 pcrecpp::RE("a*").FullMatch("aaa") matches, while
473 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
474 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
475
476 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
477 it matched the wrong number of bytes.
478
479
480 Version 7.1 24-Apr-07
481 ---------------------
482
483 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
484 that is more "standard", making use of automake and other Autotools. There
485 is some re-arrangement of the files and adjustment of comments consequent
486 on this.
487
488 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
489 for recursive directory scanning broke on some systems because the files
490 are not scanned in any specific order and on different systems the order
491 was different. A call to "sort" has been inserted into RunGrepTest for the
492 approprate test as a short-term fix. In the longer term there may be an
493 alternative.
494
495 3. I had an email from Eric Raymond about problems translating some of PCRE's
496 man pages to HTML (despite the fact that I distribute HTML pages, some
497 people do their own conversions for various reasons). The problems
498 concerned the use of low-level troff macros .br and .in. I have therefore
499 removed all such uses from the man pages (some were redundant, some could
500 be replaced by .nf/.fi pairs). The 132html script that I use to generate
501 HTML has been updated to handle .nf/.fi and to complain if it encounters
502 .br or .in.
503
504 4. Updated comments in configure.ac that get placed in config.h.in and also
505 arranged for config.h to be included in the distribution, with the name
506 config.h.generic, for the benefit of those who have to compile without
507 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
508
509 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
510 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
511 makevp.bat for the current PCRE, using the additional files
512 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
513
514 6. A Windows user reported a minor discrepancy with test 2, which turned out
515 to be caused by a trailing space on an input line that had got lost in his
516 copy. The trailing space was an accident, so I've just removed it.
517
518 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
519 that is needed.
520
521 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
522 as "const" (a) because they are and (b) because it helps the PHP
523 maintainers who have recently made a script to detect big data structures
524 in the php code that should be moved to the .rodata section. I remembered
525 to update Builducptable as well, so it won't revert if ucptable.h is ever
526 re-created.
527
528 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
529 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
530 order to be able to cut out the UTF-8 tables in the latter when UTF-8
531 support is not required. This saves 1.5-2K of code, which is important in
532 some applications.
533
534 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
535 so as not to refer to the tables, even though these functions will never be
536 called when UTF-8 support is disabled. Otherwise there are problems with a
537 shared library.
538
539 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
540
541 (a) It was defining its arguments as char * instead of void *.
542
543 (b) It was assuming that all moves were upwards in memory; this was true
544 a long time ago when I wrote it, but is no longer the case.
545
546 The emulated memove() is provided for those environments that have neither
547 memmove() nor bcopy(). I didn't think anyone used it these days, but that
548 is clearly not the case, as these two bugs were recently reported.
549
550 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
551 and Detrail to create the HTML documentation, the .txt form of the man
552 pages, and it removes trailing spaces from listed files. It also creates
553 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
554 case, it wraps all the #defines with #ifndefs. This script should be run
555 before "make dist".
556
557 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
558 with Unicode property support.
559
560 (a) For a maximizing quantifier, if the two different cases of the
561 character were of different lengths in their UTF-8 codings (there are
562 some cases like this - I found 11), and the matching function had to
563 back up over a mixture of the two cases, it incorrectly assumed they
564 were both the same length.
565
566 (b) When PCRE was configured to use the heap rather than the stack for
567 recursion during matching, it was not correctly preserving the data for
568 the other case of a UTF-8 character when checking ahead for a match
569 while processing a minimizing repeat. If the check also involved
570 matching a wide character, but failed, corruption could cause an
571 erroneous result when trying to check for a repeat of the original
572 character.
573
574 13. Some tidying changes to the testing mechanism:
575
576 (a) The RunTest script now detects the internal link size and whether there
577 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
578 values substituted by "configure". (The RunGrepTest script already did
579 this for UTF-8.) The configure.ac script no longer substitutes the
580 relevant variables.
581
582 (b) The debugging options /B and /D in pcretest show the compiled bytecode
583 with length and offset values. This means that the output is different
584 for different internal link sizes. Test 2 is skipped for link sizes
585 other than 2 because of this, bypassing the problem. Unfortunately,
586 there was also a test in test 3 (the locale tests) that used /B and
587 failed for link sizes other than 2. Rather than cut the whole test out,
588 I have added a new /Z option to pcretest that replaces the length and
589 offset values with spaces. This is now used to make test 3 independent
590 of link size. (Test 2 will be tidied up later.)
591
592 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
593 segmentation fault instead of returning the appropriate error message.
594
595 15. In multiline mode when the newline sequence was set to "any", the pattern
596 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
597 This doesn't seem right; it now treats the CRLF combination as the line
598 ending, and so does not match in that case. It's only a pattern such as ^$
599 that would hit this one: something like ^ABC$ would have failed after \r
600 and then tried again after \r\n.
601
602 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
603 in an attempt to make files that differ only in their line terminators
604 compare equal. This works on Linux.
605
606 17. Under certain error circumstances pcregrep might try to free random memory
607 as it exited. This is now fixed, thanks to valgrind.
608
609 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
610 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
611 was because its rules for how to advance for /g after matching an empty
612 string at the end of a line did not allow for this case. They now check for
613 it specially.
614
615 20. pcretest is supposed to handle patterns and data of any length, by
616 extending its buffers when necessary. It was getting this wrong when the
617 buffer for a data line had to be extended.
618
619 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
620 CRLF as a newline sequence.
621
622 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
623 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
624 I have nevertheless tidied it up.
625
626 23. Added some casts to kill warnings from HP-UX ia64 compiler.
627
628 24. Added a man page for pcre-config.
629
630
631 Version 7.0 19-Dec-06
632 ---------------------
633
634 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
635 moving to gcc 4.1.1.
636
637 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
638 sys/time.h, which is documented as needed for this function. It doesn't
639 seem to matter on Linux, but it showed up on some releases of OS X.
640
641 3. It seems that there are systems where bytes whose values are greater than
642 127 match isprint() in the "C" locale. The "C" locale should be the
643 default when a C program starts up. In most systems, only ASCII printing
644 characters match isprint(). This difference caused the output from pcretest
645 to vary, making some of the tests fail. I have changed pcretest so that:
646
647 (a) When it is outputting text in the compiled version of a pattern, bytes
648 other than 32-126 are always shown as hex escapes.
649
650 (b) When it is outputting text that is a matched part of a subject string,
651 it does the same, unless a different locale has been set for the match
652 (using the /L modifier). In this case, it uses isprint() to decide.
653
654 4. Fixed a major bug that caused incorrect computation of the amount of memory
655 required for a compiled pattern when options that changed within the
656 pattern affected the logic of the preliminary scan that determines the
657 length. The relevant options are -x, and -i in UTF-8 mode. The result was
658 that the computed length was too small. The symptoms of this bug were
659 either the PCRE error "internal error: code overflow" from pcre_compile(),
660 or a glibc crash with a message such as "pcretest: free(): invalid next
661 size (fast)". Examples of patterns that provoked this bug (shown in
662 pcretest format) are:
663
664 /(?-x: )/x
665 /(?x)(?-x: \s*#\s*)/
666 /((?i)[\x{c0}])/8
667 /(?i:[\x{c0}])/8
668
669 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
670 is now done differently.
671
672 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
673 wrapper classes; (b) implement a new function in the C++ scanner that is
674 more efficient than the old way of doing things because it avoids levels of
675 recursion in the regex matching; (c) add a paragraph to the documentation
676 for the FullMatch() function.
677
678 6. The escape sequence \n was being treated as whatever was defined as
679 "newline". Not only was this contrary to the documentation, which states
680 that \n is character 10 (hex 0A), but it also went horribly wrong when
681 "newline" was defined as CRLF. This has been fixed.
682
683 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
684 was being set to -1 for the "end of line" case (supposedly a value that no
685 character can have). Though this value is never used (the check for end of
686 line is "zero bytes in current character"), it caused compiler complaints.
687 I've changed it to 0xffffffff.
688
689 8. In pcre_version.c, the version string was being built by a sequence of
690 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
691 string (as it is for production releases) called a macro with an empty
692 argument. The C standard says the result of this is undefined. The gcc
693 compiler treats it as an empty string (which was what was wanted) but it is
694 reported that Visual C gives an error. The source has been hacked around to
695 avoid this problem.
696
697 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
698 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
699 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
700 of them did).
701
702 10. Originally, pcretest opened its input and output without "b"; then I was
703 told that "b" was needed in some environments, so it was added for release
704 5.0 to both the input and output. (It makes no difference on Unix-like
705 systems.) Later I was told that it is wrong for the input on Windows. I've
706 now abstracted the modes into two macros, to make it easier to fiddle with
707 them, and removed "b" from the input mode under Windows.
708
709 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
710
711 12. Added -help and --help to pcretest as an official way of being reminded
712 of the options.
713
714 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
715 and pcrecpp.cc because they annoy compilers at high warning levels.
716
717 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
718
719 15. Fixed an occurrence of == in configure.ac that should have been = (shell
720 scripts are not C programs :-) and which was not noticed because it works
721 on Linux.
722
723 16. pcretest is supposed to handle any length of pattern and data line (as one
724 line or as a continued sequence of lines) by extending its input buffer if
725 necessary. This feature was broken for very long pattern lines, leading to
726 a string of junk being passed to pcre_compile() if the pattern was longer
727 than about 50K.
728
729 17. I have done a major re-factoring of the way pcre_compile() computes the
730 amount of memory needed for a compiled pattern. Previously, there was code
731 that made a preliminary scan of the pattern in order to do this. That was
732 OK when PCRE was new, but as the facilities have expanded, it has become
733 harder and harder to keep it in step with the real compile phase, and there
734 have been a number of bugs (see for example, 4 above). I have now found a
735 cunning way of running the real compile function in a "fake" mode that
736 enables it to compute how much memory it would need, while actually only
737 ever using a few hundred bytes of working memory and without too many
738 tests of the mode. This should make future maintenance and development
739 easier. A side effect of this work is that the limit of 200 on the nesting
740 depth of parentheses has been removed (though this was never a serious
741 limitation, I suspect). However, there is a downside: pcre_compile() now
742 runs more slowly than before (30% or more, depending on the pattern). I
743 hope this isn't a big issue. There is no effect on runtime performance.
744
745 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
746 newline (only possible for the last line of a file) and it was a
747 pattern that set a locale (followed by /Lsomething), pcretest crashed.
748
749 19. Added additional timing features to pcretest. (1) The -tm option now times
750 matching only, not compiling. (2) Both -t and -tm can be followed, as a
751 separate command line item, by a number that specifies the number of
752 repeats to use when timing. The default is 50000; this gives better
753 precision, but takes uncomfortably long for very large patterns.
754
755 20. Extended pcre_study() to be more clever in cases where a branch of a
756 subpattern has no definite first character. For example, (a*|b*)[cd] would
757 previously give no result from pcre_study(). Now it recognizes that the
758 first character must be a, b, c, or d.
759
760 21. There was an incorrect error "recursive call could loop indefinitely" if
761 a subpattern (or the entire pattern) that was being tested for matching an
762 empty string contained only one non-empty item after a nested subpattern.
763 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
764 incorrectly, because the \d was being skipped in the check.
765
766 22. The pcretest program now has a new pattern option /B and a command line
767 option -b, which is equivalent to adding /B to every pattern. This causes
768 it to show the compiled bytecode, without the additional information that
769 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
770 is the same as /B/I).
771
772 23. A new optimization is now able automatically to treat some sequences such
773 as a*b as a*+b. More specifically, if something simple (such as a character
774 or a simple class like \d) has an unlimited quantifier, and is followed by
775 something that cannot possibly match the quantified thing, the quantifier
776 is automatically "possessified".
777
778 24. A recursive reference to a subpattern whose number was greater than 39
779 went wrong under certain circumstances in UTF-8 mode. This bug could also
780 have affected the operation of pcre_study().
781
782 25. Realized that a little bit of performance could be had by replacing
783 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
784
785 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
786
787 27. Possessive quantifiers such as a++ were previously implemented by turning
788 them into atomic groups such as ($>a+). Now they have their own opcodes,
789 which improves performance. This includes the automatically created ones
790 from 23 above.
791
792 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
793 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
794 the first matched character to be a colon. This applied both to named and
795 numbered groups.
796
797 29. The ucpinternal.h header file was missing its idempotency #ifdef.
798
799 30. I was sent a "project" file called libpcre.a.dev which I understand makes
800 building PCRE on Windows easier, so I have included it in the distribution.
801
802 31. There is now a check in pcretest against a ridiculously large number being
803 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
804 loop, the loop is abandoned.
805
806 32. Forward references to subpatterns in conditions such as (?(2)...) where
807 subpattern 2 is defined later cause pcre_compile() to search forwards in
808 the pattern for the relevant set of parentheses. This search went wrong
809 when there were unescaped parentheses in a character class, parentheses
810 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
811
812 33. "Subroutine" calls and backreferences were previously restricted to
813 referencing subpatterns earlier in the regex. This restriction has now
814 been removed.
815
816 34. Added a number of extra features that are going to be in Perl 5.10. On the
817 whole, these are just syntactic alternatives for features that PCRE had
818 previously implemented using the Python syntax or my own invention. The
819 other formats are all retained for compatibility.
820
821 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
822 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
823 also .NET compatible.
824
825 (b) A recursion or subroutine call to a named group can now be defined as
826 (?&name) as well as (?P>name).
827
828 (c) A backreference to a named group can now be defined as \k<name> or
829 \k'name' as well as (?P=name). The new forms, as well as being in Perl
830 5.10, are also .NET compatible.
831
832 (d) A conditional reference to a named group can now use the syntax
833 (?(<name>) or (?('name') as well as (?(name).
834
835 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
836 groups (named and numbered) that are never evaluated inline, but can be
837 called as "subroutines" from elsewhere. In effect, the DEFINE condition
838 is always false. There may be only one alternative in such a group.
839
840 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
841 as the simple (?(R). The condition is true only if the most recent
842 recursion is that of the given number or name. It does not search out
843 through the entire recursion stack.
844
845 (g) The escape \gN or \g{N} has been added, where N is a positive or
846 negative number, specifying an absolute or relative reference.
847
848 35. Tidied to get rid of some further signed/unsigned compiler warnings and
849 some "unreachable code" warnings.
850
851 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
852 things, this adds five new scripts.
853
854 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
855 There were also incompatibilities regarding the handling of \Q..\E inside
856 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
857 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
858
859 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
860 matches an empty string, and forcibly breaks the loop. There were bugs in
861 this code in non-simple cases. For a pattern such as ^(a()*)* matched
862 against aaaa the result was just "a" rather than "aaaa", for example. Two
863 separate and independent bugs (that affected different cases) have been
864 fixed.
865
866 39. Refactored the code to abolish the use of different opcodes for small
867 capturing bracket numbers. This is a tidy that I avoided doing when I
868 removed the limit on the number of capturing brackets for 3.5 back in 2001.
869 The new approach is not only tidier, it makes it possible to reduce the
870 memory needed to fix the previous bug (38).
871
872 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
873 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
874 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
875 mode.
876
877 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
878 report.
879
880 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
881 copy construction and assignment in the C++ wrapper.
882
883 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
884 couple of bugs that could have given wrong results in the "--newline=crlf"
885 case.
886
887 44. Added a number of casts and did some reorganization of signed/unsigned int
888 variables following suggestions from Dair Grant. Also renamed the variable
889 "this" as "item" because it is a C++ keyword.
890
891 45. Arranged for dftables to add
892
893 #include "pcre_internal.h"
894
895 to pcre_chartables.c because without it, gcc 4.x may remove the array
896 definition from the final binary if PCRE is built into a static library and
897 dead code stripping is activated.
898
899 46. For an unanchored pattern, if a match attempt fails at the start of a
900 newline sequence, and the newline setting is CRLF or ANY, and the next two
901 characters are CRLF, advance by two characters instead of one.
902
903
904 Version 6.7 04-Jul-06
905 ---------------------
906
907 1. In order to handle tests when input lines are enormously long, pcretest has
908 been re-factored so that it automatically extends its buffers when
909 necessary. The code is crude, but this _is_ just a test program. The
910 default size has been increased from 32K to 50K.
911
912 2. The code in pcre_study() was using the value of the re argument before
913 testing it for NULL. (Of course, in any sensible call of the function, it
914 won't be NULL.)
915
916 3. The memmove() emulation function in pcre_internal.h, which is used on
917 systems that lack both memmove() and bcopy() - that is, hardly ever -
918 was missing a "static" storage class specifier.
919
920 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
921 containing an extended class (one that cannot be represented by a bitmap
922 because it contains high-valued characters or Unicode property items, e.g.
923 [\pZ]). Almost always one would set UTF-8 mode when processing such a
924 pattern, but PCRE should not loop if you do not (it no longer does).
925 [Detail: two cases were found: (a) a repeated subpattern containing an
926 extended class; (b) a recursive reference to a subpattern that followed a
927 previous extended class. It wasn't skipping over the extended class
928 correctly when UTF-8 mode was not set.]
929
930 5. A negated single-character class was not being recognized as fixed-length
931 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
932 compile error "lookbehind assertion is not fixed length".
933
934 6. The RunPerlTest auxiliary script was showing an unexpected difference
935 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
936 write a Perl script that can interpret lines of an input file either as
937 byte characters or as UTF-8, which is what "perltest" was being required to
938 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
939 can't do is switch easily at run time between having the "use utf8;" pragma
940 or not. In the end, I fudged it by using the RunPerlTest script to insert
941 "use utf8;" explicitly for the UTF-8 tests.
942
943 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
944 the end of the subject string, contrary to the documentation and to what
945 Perl does. This was true of both matching functions. Now it matches only at
946 the start of the subject and immediately after *internal* newlines.
947
948 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
949 a pointer to an int instead of a pointer to an unsigned long int. This
950 caused problems on 64-bit systems.
951
952 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
953 instance of the 'standard' template library not being so standard".
954
955 10. There was no check on the number of named subpatterns nor the maximum
956 length of a subpattern name. The product of these values is used to compute
957 the size of the memory block for a compiled pattern. By supplying a very
958 long subpattern name and a large number of named subpatterns, the size
959 computation could be caused to overflow. This is now prevented by limiting
960 the length of names to 32 characters, and the number of named subpatterns
961 to 10,000.
962
963 11. Subpatterns that are repeated with specific counts have to be replicated in
964 the compiled pattern. The size of memory for this was computed from the
965 length of the subpattern and the repeat count. The latter is limited to
966 65535, but there was no limit on the former, meaning that integer overflow
967 could in principle occur. The compiled length of a repeated subpattern is
968 now limited to 30,000 bytes in order to prevent this.
969
970 12. Added the optional facility to have named substrings with the same name.
971
972 13. Added the ability to use a named substring as a condition, using the
973 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
974 are numbers (not recommended). Forward references are permitted.
975
976 14. Added forward references in named backreferences (if you see what I mean).
977
978 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
979 pattern could run off the end of the subject. For example, the pattern
980 "(?s)(.{1,5})"8 did this with the subject "ab".
981
982 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
983 PCRE_CASELESS was set when matching characters that were quantified with ?
984 or *.
985
986 17. A character class other than a single negated character that had a minimum
987 but no maximum quantifier - for example [ab]{6,} - was not handled
988 correctly by pce_dfa_exec(). It would match only one character.
989
990 18. A valid (though odd) pattern that looked like a POSIX character
991 class but used an invalid character after [ (for example [[,abc,]]) caused
992 pcre_compile() to give the error "Failed: internal error: code overflow" or
993 in some cases to crash with a glibc free() error. This could even happen if
994 the pattern terminated after [[ but there just happened to be a sequence of
995 letters, a binary zero, and a closing ] in the memory that followed.
996
997 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
998 over the years. Originally (before any Unicode support), just the bottom 8
999 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1000 output from "man perlunicode" includes this:
1001
1002 The regular expression compiler produces polymorphic opcodes. That
1003 is, the pattern adapts to the data and automatically switches to
1004 the Unicode character scheme when presented with Unicode data--or
1005 instead uses a traditional byte scheme when presented with byte
1006 data.
1007
1008 Sadly, a wide octal escape does not cause a switch, and in a string with
1009 no other multibyte characters, these octal escapes are treated as before.
1010 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1011 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1012 Unicode string.
1013
1014 I have not perpetrated such confusion in PCRE. Up till now, it took just
1015 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1016 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1017 translate to the appropriate multibyte character.
1018
1019 29. Applied some refactoring to reduce the number of warnings from Microsoft
1020 and Borland compilers. This has included removing the fudge introduced
1021 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1022 a warning about an unused variable.
1023
1024 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1025 characters since release 4.0, because Perl (from release 5.004) does not.
1026 [Or at least, is documented not to: some releases seem to be in conflict
1027 with the documentation.] However, when a pattern was studied with
1028 pcre_study() and all its branches started with \s, PCRE still included VT
1029 as a possible starting character. Of course, this did no harm; it just
1030 caused an unnecessary match attempt.
1031
1032 22. Removed a now-redundant internal flag bit that recorded the fact that case
1033 dependency changed within the pattern. This was once needed for "required
1034 byte" processing, but is no longer used. This recovers a now-scarce options
1035 bit. Also moved the least significant internal flag bit to the most-
1036 significant bit of the word, which was not previously used (hangover from
1037 the days when it was an int rather than a uint) to free up another bit for
1038 the future.
1039
1040 23. Added support for CRLF line endings as well as CR and LF. As well as the
1041 default being selectable at build time, it can now be changed at runtime
1042 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1043 specify that it is scanning data with non-default line endings.
1044
1045 24. Changed the definition of CXXLINK to make it agree with the definition of
1046 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1047
1048 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1049 recursions. This makes a big different to stack usage for some patterns.
1050
1051 26. If a subpattern containing a named recursion or subroutine reference such
1052 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1053 the space required for the compiled pattern went wrong and gave too small a
1054 value. Depending on the environment, this could lead to "Failed: internal
1055 error: code overflow at offset 49" or "glibc detected double free or
1056 corruption" errors.
1057
1058 27. Applied patches from Google (a) to support the new newline modes and (b) to
1059 advance over multibyte UTF-8 characters in GlobalReplace.
1060
1061 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1062 difference for some implementation of PCRE in some Windows version.
1063
1064 29. Added some extra testing facilities to pcretest:
1065
1066 \q<number> in a data line sets the "match limit" value
1067 \Q<number> in a data line sets the "match recursion limt" value
1068 -S <number> sets the stack size, where <number> is in megabytes
1069
1070 The -S option isn't available for Windows.
1071
1072
1073 Version 6.6 06-Feb-06
1074 ---------------------
1075
1076 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1077 in pcreposix.h. I have copied the definition from pcre.h.
1078
1079 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1080 because pcre.h is no longer a built file.
1081
1082 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1083 not normally included in the compiled code.
1084
1085
1086 Version 6.5 01-Feb-06
1087 ---------------------
1088
1089 1. When using the partial match feature with pcre_dfa_exec(), it was not
1090 anchoring the second and subsequent partial matches at the new starting
1091 point. This could lead to incorrect results. For example, with the pattern
1092 /1234/, partially matching against "123" and then "a4" gave a match.
1093
1094 2. Changes to pcregrep:
1095
1096 (a) All non-match returns from pcre_exec() were being treated as failures
1097 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1098 error message is output. Some extra information is given for the
1099 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1100 probably the only errors that are likely to be caused by users (by
1101 specifying a regex that has nested indefinite repeats, for instance).
1102 If there are more than 20 of these errors, pcregrep is abandoned.
1103
1104 (b) A binary zero was treated as data while matching, but terminated the
1105 output line if it was written out. This has been fixed: binary zeroes
1106 are now no different to any other data bytes.
1107
1108 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1109 used to set a locale for matching. The --locale=xxxx long option has
1110 been added (no short equivalent) to specify a locale explicitly on the
1111 pcregrep command, overriding the environment variables.
1112
1113 (d) When -B was used with -n, some line numbers in the output were one less
1114 than they should have been.
1115
1116 (e) Added the -o (--only-matching) option.
1117
1118 (f) If -A or -C was used with -c (count only), some lines of context were
1119 accidentally printed for the final match.
1120
1121 (g) Added the -H (--with-filename) option.
1122
1123 (h) The combination of options -rh failed to suppress file names for files
1124 that were found from directory arguments.
1125
1126 (i) Added the -D (--devices) and -d (--directories) options.
1127
1128 (j) Added the -F (--fixed-strings) option.
1129
1130 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1131
1132 (l) Added the --colo(u)r option.
1133
1134 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1135 is not present by default.
1136
1137 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1138 items such as (?R) or (?1), when the recursion could match a number of
1139 alternatives. If it matched one of the alternatives, but subsequently,
1140 outside the recursion, there was a failure, the code tried to back up into
1141 the recursion. However, because of the way PCRE is implemented, this is not
1142 possible, and the result was an incorrect result from the match.
1143
1144 In order to prevent this happening, the specification of recursion has
1145 been changed so that all such subpatterns are automatically treated as
1146 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1147
1148 4. I had overlooked the fact that, in some locales, there are characters for
1149 which isalpha() is true but neither isupper() nor islower() are true. In
1150 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1151 and ordfeminine) are like this. This affected the treatment of \w and \W
1152 when they appeared in character classes, but not when they appeared outside
1153 a character class. The bit map for "word" characters is now created
1154 separately from the results of isalnum() instead of just taking it from the
1155 upper, lower, and digit maps. (Plus the underscore character, of course.)
1156
1157 5. The above bug also affected the handling of POSIX character classes such as
1158 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1159 permanent tables. Instead, the bit maps for such a class were previously
1160 created as the appropriate unions of the upper, lower, and digit bitmaps.
1161 Now they are created by subtraction from the [[:word:]] class, which has
1162 its own bitmap.
1163
1164 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1165 It is created by subtracting the vertical space characters (\x09, \x0a,
1166 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1167 subtraction was done in the overall bitmap for a character class, meaning
1168 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1169 be recognized. This bug has been fixed.
1170
1171 7. Patches from the folks at Google:
1172
1173 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1174 real life, but is still worth protecting against".
1175
1176 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1177 regular expressions".
1178
1179 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1180 have it.
1181
1182 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1183 "configure" and the latter not, in order to fix a problem somebody had
1184 with compiling the Arg class on HP-UX.
1185
1186 (e) Improve the error-handling of the C++ wrapper a little bit.
1187
1188 (f) New tests for checking recursion limiting.
1189
1190 8. The pcre_memmove() function, which is used only if the environment does not
1191 have a standard memmove() function (and is therefore rarely compiled),
1192 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1193 returning a result (though PCRE never actually uses the result).
1194
1195 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1196 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1197 returned instead of calling malloc() with an overflowing number that would
1198 most likely cause subsequent chaos.
1199
1200 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1201
1202 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1203 with this option is matched, the nmatch and pmatch options of regexec() are
1204 ignored.
1205
1206 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1207 provided in case anyone wants to the the POSIX interface with UTF-8
1208 strings.
1209
1210 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1211 C++ linking (needed for some HP-UX environments).
1212
1213 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1214 (unused parameter) and in the pcre_printint() function (omitted "default"
1215 switch label when the default is to do nothing).
1216
1217 15. Added some code to make it possible, when PCRE is compiled as a C++
1218 library, to replace subject pointers for pcre_exec() with a smart pointer
1219 class, thus making it possible to process discontinuous strings.
1220
1221 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1222 much the same function. They were added by different people who were trying
1223 to make PCRE easy to compile on non-Unix systems. It has been suggested
1224 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1225 for compiling on Windows systems. I have therefore replaced it with
1226 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1227 defaults to "extern" for C or "extern C" for C++, which works fine on
1228 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1229 SCOPE with something explicit in config.h. In addition:
1230
1231 (a) pcreposix.h still had just "extern" instead of either of these macros;
1232 I have replaced it with PCRE_DATA_SCOPE.
1233
1234 (b) Functions such as _pcre_xclass(), which are internal to the library,
1235 but external in the C sense, all had PCRE_EXPORT in their definitions.
1236 This is apparently wrong for the Windows case, so I have removed it.
1237 (It makes no difference on Unix-like systems.)
1238
1239 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1240 of recursive calls to match(). This is different to MATCH_LIMIT because
1241 that limits the total number of calls to match(), not all of which increase
1242 the depth of recursion. Limiting the recursion depth limits the amount of
1243 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1244 when PCRE is compiled, and changed at run time. A patch from Google adds
1245 this functionality to the C++ interface.
1246
1247 18. Changes to the handling of Unicode character properties:
1248
1249 (a) Updated the table to Unicode 4.1.0.
1250
1251 (b) Recognize characters that are not in the table as "Cn" (undefined).
1252
1253 (c) I revised the way the table is implemented to a much improved format
1254 which includes recognition of ranges. It now supports the ranges that
1255 are defined in UnicodeData.txt, and it also amalgamates other
1256 characters into ranges. This has reduced the number of entries in the
1257 table from around 16,000 to around 3,000, thus reducing its size
1258 considerably. I realized I did not need to use a tree structure after
1259 all - a binary chop search is just as efficient. Having reduced the
1260 number of entries, I extended their size from 6 bytes to 8 bytes to
1261 allow for more data.
1262
1263 (d) Added support for Unicode script names via properties such as \p{Han}.
1264
1265 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1266 matching that character.
1267
1268 20. When matching a repeated Unicode property with a minimum greater than zero,
1269 (for example \pL{2,}), PCRE could look past the end of the subject if it
1270 reached it while seeking the minimum number of characters. This could
1271 happen only if some of the characters were more than one byte long, because
1272 there is a check for at least the minimum number of bytes.
1273
1274 21. Refactored the implementation of \p and \P so as to be more general, to
1275 allow for more different types of property in future. This has changed the
1276 compiled form incompatibly. Anybody with saved compiled patterns that use
1277 \p or \P will have to recompile them.
1278
1279 22. Added "Any" and "L&" to the supported property types.
1280
1281 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1282 but give a compile time error if the value is greater than 0xff.
1283
1284 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1285 accidentally not being installed or uninstalled.
1286
1287 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1288 made were to insert the current release number. This seemed silly, because
1289 it made things harder for people building PCRE on systems that don't run
1290 "configure". I have turned pcre.h into a distributed file, no longer built
1291 by "configure", with the version identification directly included. There is
1292 no longer a pcre.h.in file.
1293
1294 However, this change necessitated a change to the pcre-config script as
1295 well. It is built from pcre-config.in, and one of the substitutions was the
1296 release number. I have updated configure.ac so that ./configure now finds
1297 the release number by grepping pcre.h.
1298
1299 26. Added the ability to run the tests under valgrind.
1300
1301
1302 Version 6.4 05-Sep-05
1303 ---------------------
1304
1305 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1306 "--" to be printed when multiple files were scanned, even when none of the
1307 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1308 consider it to be a bug, and have restored the previous behaviour.
1309
1310 2. A couple of code tidies to get rid of compiler warnings.
1311
1312 3. The pcretest program used to cheat by referring to symbols in the library
1313 whose names begin with _pcre_. These are internal symbols that are not
1314 really supposed to be visible externally, and in some environments it is
1315 possible to suppress them. The cheating is now confined to including
1316 certain files from the library's source, which is a bit cleaner.
1317
1318 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1319 file's purpose clearer.
1320
1321 5. Reorganized pcre_ucp_findchar().
1322
1323
1324 Version 6.3 15-Aug-05
1325 ---------------------
1326
1327 1. The file libpcre.pc.in did not have general read permission in the tarball.
1328
1329 2. There were some problems when building without C++ support:
1330
1331 (a) If C++ support was not built, "make install" and "make test" still
1332 tried to test it.
1333
1334 (b) There were problems when the value of CXX was explicitly set. Some
1335 changes have been made to try to fix these, and ...
1336
1337 (c) --disable-cpp can now be used to explicitly disable C++ support.
1338
1339 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1340 backslash in a target when C++ was disabled. This confuses some
1341 versions of "make", apparently. Using an intermediate variable solves
1342 this. (Same for CPP_LOBJ.)
1343
1344 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1345 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1346 necessary on certain architectures.
1347
1348 4. Added a setting of -export-symbols-regex to the link command to remove
1349 those symbols that are exported in the C sense, but actually are local
1350 within the library, and not documented. Their names all begin with
1351 "_pcre_". This is not a perfect job, because (a) we have to except some
1352 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1353 available (and never for static libraries). I have made a note to try to
1354 find a way round (a) in the future.
1355
1356
1357 Version 6.2 01-Aug-05
1358 ---------------------
1359
1360 1. There was no test for integer overflow of quantifier values. A construction
1361 such as {1111111111111111} would give undefined results. What is worse, if
1362 a minimum quantifier for a parenthesized subpattern overflowed and became
1363 negative, the calculation of the memory size went wrong. This could have
1364 led to memory overwriting.
1365
1366 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1367
1368 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1369 operating environments where this matters.
1370
1371 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1372 PCRE options from within the C++ wrapper.
1373
1374 5. Named capturing subpatterns were not being correctly counted when a pattern
1375 was compiled. This caused two problems: (a) If there were more than 100
1376 such subpatterns, the calculation of the memory needed for the whole
1377 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1378 back references of the form \12, where the number was greater than 9, were
1379 not recognized as back references, even though there were sufficient
1380 previous subpatterns.
1381
1382 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1383 versions of gcc, e.g. 2.95.4.
1384
1385
1386 Version 6.1 21-Jun-05
1387 ---------------------
1388
1389 1. There was one reference to the variable "posix" in pcretest.c that was not
1390 surrounded by "#if !defined NOPOSIX".
1391
1392 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1393 the cross-check on the old pcre_info() function, for the benefit of the
1394 cut-down version of PCRE that is currently imported into Exim.
1395
1396 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1397 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1398 patterns that start (?i) but I don't think that matters. The use of (?i) is
1399 just an example; this all applies to the other options as well.
1400
1401 4. Since libtool seems to echo the compile commands it is issuing, the output
1402 from "make" can be reduced a bit by putting "@" in front of each libtool
1403 compile command.
1404
1405 5. Patch from the folks at Google for configure.in to be a bit more thorough
1406 in checking for a suitable C++ installation before trying to compile the
1407 C++ stuff. This should fix a reported problem when a compiler was present,
1408 but no suitable headers.
1409
1410 6. The man pages all had just "PCRE" as their title. I have changed them to
1411 be the relevant file name. I have also arranged that these names are
1412 retained in the file doc/pcre.txt, which is a concatenation in text format
1413 of all the man pages except the little individual ones for each function.
1414
1415 7. The NON-UNIX-USE file had not been updated for the different set of source
1416 files that come with release 6. I also added a few comments about the C++
1417 wrapper.
1418
1419
1420 Version 6.0 07-Jun-05
1421 ---------------------
1422
1423 1. Some minor internal re-organization to help with my DFA experiments.
1424
1425 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1426 didn't matter for the library itself when fully configured, but did matter
1427 when compiling without UCP support, or within Exim, where the ucp files are
1428 not imported.
1429
1430 3. Refactoring of the library code to split up the various functions into
1431 different source modules. The addition of the new DFA matching code (see
1432 below) to a single monolithic source would have made it really too
1433 unwieldy, quite apart from causing all the code to be include in a
1434 statically linked application, when only some functions are used. This is
1435 relevant even without the DFA addition now that patterns can be compiled in
1436 one application and matched in another.
1437
1438 The downside of splitting up is that there have to be some external
1439 functions and data tables that are used internally in different modules of
1440 the library but which are not part of the API. These have all had their
1441 names changed to start with "_pcre_" so that they are unlikely to clash
1442 with other external names.
1443
1444 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1445 a different (DFA) algorithm. Although it is slower than the original
1446 function, it does have some advantages for certain types of matching
1447 problem.
1448
1449 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1450 including restarting after a partial match.
1451
1452 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1453 defined when compiling for Windows was sent to me. I have put it into the
1454 code, though I have no means of testing or verifying it.
1455
1456 7. Added the pcre_refcount() auxiliary function.
1457
1458 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1459 match before or at the first newline in the subject string. In pcretest,
1460 the /f option on a pattern can be used to set this.
1461
1462 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1463 would behave wrongly. This has been present in PCRE since release 4.0.
1464
1465 10. A number of changes to the pcregrep command:
1466
1467 (a) Refactored how -x works; insert ^(...)$ instead of setting
1468 PCRE_ANCHORED and checking the length, in preparation for adding
1469 something similar for -w.
1470
1471 (b) Added the -w (match as a word) option.
1472
1473 (c) Refactored the way lines are read and buffered so as to have more
1474 than one at a time available.
1475
1476 (d) Implemented a pcregrep test script.
1477
1478 (e) Added the -M (multiline match) option. This allows patterns to match
1479 over several lines of the subject. The buffering ensures that at least
1480 8K, or the rest of the document (whichever is the shorter) is available
1481 for matching (and similarly the previous 8K for lookbehind assertions).
1482
1483 (f) Changed the --help output so that it now says
1484
1485 -w, --word-regex(p)
1486
1487 instead of two lines, one with "regex" and the other with "regexp"
1488 because that confused at least one person since the short forms are the
1489 same. (This required a bit of code, as the output is generated
1490 automatically from a table. It wasn't just a text change.)
1491
1492 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1493 option but starts with a hyphen. Could be a pattern or a path name
1494 starting with a hyphen, for instance.
1495
1496 (h) "-" can be given as a file name to represent stdin.
1497
1498 (i) When file names are being printed, "(standard input)" is used for
1499 the standard input, for compatibility with GNU grep. Previously
1500 "<stdin>" was used.
1501
1502 (j) The option --label=xxx can be used to supply a name to be used for
1503 stdin when file names are being printed. There is no short form.
1504
1505 (k) Re-factored the options decoding logic because we are going to add
1506 two more options that take data. Such options can now be given in four
1507 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1508
1509 (l) Added the -A, -B, and -C options for requesting that lines of context
1510 around matches be printed.
1511
1512 (m) Added the -L option to print the names of files that do not contain
1513 any matching lines, that is, the complement of -l.
1514
1515 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1516 continue to scan other files.
1517
1518 (o) The -s option was incorrectly implemented. For compatibility with other
1519 greps, it now suppresses the error message for a non-existent or non-
1520 accessible file (but not the return code). There is a new option called
1521 -q that suppresses the output of matching lines, which was what -s was
1522 previously doing.
1523
1524 (p) Added --include and --exclude options to specify files for inclusion
1525 and exclusion when recursing.
1526
1527 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1528 Hopefully, it now does.
1529
1530 12. Missing cast in pcre_study().
1531
1532 13. Added an "uninstall" target to the makefile.
1533
1534 14. Replaced "extern" in the function prototypes in Makefile.in with
1535 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1536 world, but is set differently for Windows.
1537
1538 15. Added a second compiling function called pcre_compile2(). The only
1539 difference is that it has an extra argument, which is a pointer to an
1540 integer error code. When there is a compile-time failure, this is set
1541 non-zero, in addition to the error test pointer being set to point to an
1542 error message. The new argument may be NULL if no error number is required
1543 (but then you may as well call pcre_compile(), which is now just a
1544 wrapper). This facility is provided because some applications need a
1545 numeric error indication, but it has also enabled me to tidy up the way
1546 compile-time errors are handled in the POSIX wrapper.
1547
1548 16. Added VPATH=.libs to the makefile; this should help when building with one
1549 prefix path and installing with another. (Or so I'm told by someone who
1550 knows more about this stuff than I do.)
1551
1552 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1553 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1554 match everything, including newlines. This is not POSIX-compatible, but
1555 somebody wanted the feature. From pcretest it can be activated by using
1556 both the P and the s flags.
1557
1558 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1559
1560 19. libpcre.pc was being incorrectly installed as executable.
1561
1562 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1563 it now also looks for '\r' so that it will work unmodified on Windows.
1564
1565 21. Added Google's contributed C++ wrapper to the distribution.
1566
1567 22. Added some untidy missing memory free() calls in pcretest, to keep
1568 Electric Fence happy when testing.
1569
1570
1571
1572 Version 5.0 13-Sep-04
1573 ---------------------
1574
1575 1. Internal change: literal characters are no longer packed up into items
1576 containing multiple characters in a single byte-string. Each character
1577 is now matched using a separate opcode. However, there may be more than one
1578 byte in the character in UTF-8 mode.
1579
1580 2. The pcre_callout_block structure has two new fields: pattern_position and
1581 next_item_length. These contain the offset in the pattern to the next match
1582 item, and its length, respectively.
1583
1584 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1585 insertion of callouts before each pattern item. Added the /C option to
1586 pcretest to make use of this.
1587
1588 4. On the advice of a Windows user, the lines
1589
1590 #if defined(_WIN32) || defined(WIN32)
1591 _setmode( _fileno( stdout ), 0x8000 );
1592 #endif /* defined(_WIN32) || defined(WIN32) */
1593
1594 have been added to the source of pcretest. This apparently does useful
1595 magic in relation to line terminators.
1596
1597 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1598 for the benefit of those environments where the "b" makes a difference.
1599
1600 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1601 to know about it. I have put a hack into configure.in that adds in code
1602 to set GCC=yes if CC=icc. This seems to end up at a point in the
1603 generated configure script that is early enough to affect the setting of
1604 compiler options, which is what is needed, but I have no means of testing
1605 whether it really works. (The user who reported this had patched the
1606 generated configure script, which of course I cannot do.)
1607
1608 LATER: After change 22 below (new libtool files), the configure script
1609 seems to know about icc (and also ecc). Therefore, I have commented out
1610 this hack in configure.in.
1611
1612 7. Added support for pkg-config (2 patches were sent in).
1613
1614 8. Negated POSIX character classes that used a combination of internal tables
1615 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1616 [[:^ascii]]. Typically, they would match almost any characters. The other
1617 POSIX classes were not broken in this way.
1618
1619 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1620 to find the match, as PCRE was deluded into thinking that the match had to
1621 start at the start point or following a newline. The same bug applied to
1622 patterns with negative forward assertions or any backward assertions
1623 preceding ".*" at the start, unless the pattern required a fixed first
1624 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1625
1626 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1627 starting at the last subject character, bytes beyond the end of the subject
1628 string were read.
1629
1630 11. Renamed the variable "class" as "classbits" to make life easier for C++
1631 users. (Previously there was a macro definition, but it apparently wasn't
1632 enough.)
1633
1634 12. Added the new field "tables" to the extra data so that tables can be passed
1635 in at exec time, or the internal tables can be re-selected. This allows
1636 a compiled regex to be saved and re-used at a later time by a different
1637 program that might have everything at different addresses.
1638
1639 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1640 -R library as well as a -L library.
1641
1642 14. The debugging options of pcretest (-d on the command line or D on a
1643 pattern) showed incorrect output for anything following an extended class
1644 that contained multibyte characters and which was followed by a quantifier.
1645
1646 15. Added optional support for general category Unicode character properties
1647 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1648 support. It adds about 90K to the size of the library. The meanings of the
1649 inbuilt class escapes such as \d and \s have NOT been changed.
1650
1651 16. Updated pcredemo.c to include calls to free() to release the memory for the
1652 compiled pattern.
1653
1654 17. The generated file chartables.c was being created in the source directory
1655 instead of in the building directory. This caused the build to fail if the
1656 source directory was different from the building directory, and was
1657 read-only.
1658
1659 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1660 file. No doubt somebody will tell me if they don't make sense... Also added
1661 Dan Mooney's comments about building on OpenVMS.
1662
1663 19. Added support for partial matching via the PCRE_PARTIAL option for
1664 pcre_exec() and the \P data escape in pcretest.
1665
1666 20. Extended pcretest with 3 new pattern features:
1667
1668 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1669 write the compiled pattern to the file whose name is "rest-of-line".
1670 This is a straight binary dump of the data, with the saved pointer to
1671 the character tables forced to be NULL. The study data, if any, is
1672 written too. After writing, pcretest reads a new pattern.
1673
1674 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1675 compiled pattern from the given file. There must not be any
1676 occurrences of "<" in the file name (pretty unlikely); if there are,
1677 pcretest will instead treat the initial "<" as a pattern delimiter.
1678 After reading in the pattern, pcretest goes on to read data lines as
1679 usual.
1680
1681 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1682 and 16-bit fields in a compiled pattern, to simulate a pattern that
1683 was compiled on a host of opposite endianness.
1684
1685 21. The pcre-exec() function can now cope with patterns that were compiled on
1686 hosts of opposite endianness, with this restriction:
1687
1688 As for any compiled expression that is saved and used later, the tables
1689 pointer field cannot be preserved; the extra_data field in the arguments
1690 to pcre_exec() should be used to pass in a tables address if a value
1691 other than the default internal tables were used at compile time.
1692
1693 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1694 now diagnosed as an error. Previously, most of the time, a negative number
1695 would have been treated as zero, but if in addition "ovector" was passed as
1696 NULL, a crash could occur.
1697
1698 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1699 new versions from the libtool 1.5 distribution (the last one is a copy of
1700 a file called libtool.m4). This seems to have fixed the need to patch
1701 "configure" to support Darwin 1.3 (which I used to do). However, I still
1702 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1703 workstation).
1704
1705 24. Changed the PCRE licence to be the more standard "BSD" licence.
1706
1707
1708 Version 4.5 01-Dec-03
1709 ---------------------
1710
1711 1. There has been some re-arrangement of the code for the match() function so
1712 that it can be compiled in a version that does not call itself recursively.
1713 Instead, it keeps those local variables that need separate instances for
1714 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1715 needs to "recurse". Keeping track of where control must go is done by means
1716 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1717 hide most of the details from the main code, and operates only if
1718 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1719 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1720 operating.
1721
1722 To make it easier for callers to provide specially tailored get/free
1723 functions for this usage, two new functions, pcre_stack_malloc, and
1724 pcre_stack_free, are used. They are always called in strict stacking order,
1725 and the size of block requested is always the same.
1726
1727 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1728 PCRE has been compiled to use the stack or the heap for recursion. The
1729 -C option of pcretest uses this to show which version is compiled.
1730
1731 A new data escape \S, is added to pcretest; it causes the amounts of store
1732 obtained and freed by both kinds of malloc/free at match time to be added
1733 to the output.
1734
1735 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1736 what's available on my current Linux desktop machine.
1737
1738 3. When matching a UTF-8 string, the test for a valid string at the start has
1739 been extended. If start_offset is not zero, PCRE now checks that it points
1740 to a byte that is the start of a UTF-8 character. If not, it returns
1741 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1742 this is necessary because there may be backward assertions in the pattern.
1743 When matching the same subject several times, it may save resources to use
1744 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1745
1746 4. The code for checking the validity of UTF-8 strings has been tightened so
1747 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1748 containing "overlong sequences".
1749
1750 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1751 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1752 should have been "&", but it just so happened that all the cases this let
1753 through by mistake were picked up later in the function.
1754
1755 6. I had used a variable called "isblank" - this is a C99 function, causing
1756 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1757
1758 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1759 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1760 so the version is shown; (c) stop "make test" echoing "./RunTest".
1761
1762 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1763
1764 9. The prototype for memmove() for systems that don't have it was using
1765 size_t, but the inclusion of the header that defines size_t was later. I've
1766 moved the #includes for the C headers earlier to avoid this.
1767
1768 10. Added some adjustments to the code to make it easier to compiler on certain
1769 special systems:
1770
1771 (a) Some "const" qualifiers were missing.
1772 (b) Added the macro EXPORT before all exported functions; by default this
1773 is defined to be empty.
1774 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1775 that it reads its output file name as an argument instead of writing
1776 to the standard output and assuming this can be redirected.
1777
1778 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1779 class containing characters with values greater than 255, PCRE compilation
1780 went into a loop.
1781
1782 12. A recursive reference to a subpattern that was within another subpattern
1783 that had a minimum quantifier of zero caused PCRE to crash. For example,
1784 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1785 recursion. If the recursively-called subpattern itself had a zero repeat,
1786 that was OK.
1787
1788 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1789 buffer into which it was copied (for escape processing) was still set at
1790 1024, so long lines caused crashes.
1791
1792 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1793 "internal error: code overflow...". This applied to any character class
1794 that was followed by a possessive quantifier.
1795
1796 15. Modified the Makefile to add libpcre.la as a prerequisite for
1797 libpcreposix.la because I was told this is needed for a parallel build to
1798 work.
1799
1800 16. If a pattern that contained .* following optional items at the start was
1801 studied, the wrong optimizing data was generated, leading to matching
1802 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1803 matching string must start with a or b or c. The correct conclusion for
1804 this pattern is that a match can start with any character.
1805
1806
1807 Version 4.4 13-Aug-03
1808 ---------------------
1809
1810 1. In UTF-8 mode, a character class containing characters with values between
1811 127 and 255 was not handled correctly if the compiled pattern was studied.
1812 In fixing this, I have also improved the studying algorithm for such
1813 classes (slightly).
1814
1815 2. Three internal functions had redundant arguments passed to them. Removal
1816 might give a very teeny performance improvement.
1817
1818 3. Documentation bug: the value of the capture_top field in a callout is *one
1819 more than* the number of the hightest numbered captured substring.
1820
1821 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1822 in incorrectly linking with a previously installed version. They now link
1823 explicitly with libpcre.la.
1824
1825 5. configure.in no longer needs to recognize Cygwin specially.
1826
1827 6. A problem in pcre.in for Windows platforms is fixed.
1828
1829 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1830 pcretest, it used to include the size of the study block as part of its
1831 output. Unfortunately, the structure contains a field that has a different
1832 size on different hardware architectures. This meant that the tests that
1833 showed this size failed. As the block is currently always of a fixed size,
1834 this information isn't actually particularly useful in pcretest output, so
1835 I have just removed it.
1836
1837 8. Three pre-processor statements accidentally did not start in column 1.
1838 Sadly, there are *still* compilers around that complain, even though
1839 standard C has not required this for well over a decade. Sigh.
1840
1841 9. In pcretest, the code for checking callouts passed small integers in the
1842 callout_data field, which is a void * field. However, some picky compilers
1843 complained about the casts involved for this on 64-bit systems. Now
1844 pcretest passes the address of the small integer instead, which should get
1845 rid of the warnings.
1846
1847 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1848 both compile and run time, and gives an error if an invalid UTF-8 sequence
1849 is found. There is a option for disabling this check in cases where the
1850 string is known to be correct and/or the maximum performance is wanted.
1851
1852 11. In response to a bug report, I changed one line in Makefile.in from
1853
1854 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1855 to
1856 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1857
1858 to look similar to other lines, but I have no way of telling whether this
1859 is the right thing to do, as I do not use Windows. No doubt I'll get told
1860 if it's wrong...
1861
1862
1863 Version 4.3 21-May-03
1864 ---------------------
1865
1866 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1867 Makefile.
1868
1869 2. Some refactoring to improve the quality of the code:
1870
1871 (i) The utf8_table... variables are now declared "const".
1872
1873 (ii) The code for \cx, which used the "case flipping" table to upper case
1874 lower case letters, now just substracts 32. This is ASCII-specific,
1875 but the whole concept of \cx is ASCII-specific, so it seems
1876 reasonable.
1877
1878 (iii) PCRE was using its character types table to recognize decimal and
1879 hexadecimal digits in the pattern. This is silly, because it handles
1880 only 0-9, a-f, and A-F, but the character types table is locale-
1881 specific, which means strange things might happen. A private
1882 table is now used for this - though it costs 256 bytes, a table is
1883 much faster than multiple explicit tests. Of course, the standard
1884 character types table is still used for matching digits in subject
1885 strings against \d.
1886
1887 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1888 ending in _t are). So I've renamed it as ESC_tee.
1889
1890 3. The first argument for regexec() in the POSIX wrapper should have been
1891 defined as "const".
1892
1893 4. Changed pcretest to use malloc() for its buffers so that they can be
1894 Electric Fenced for debugging.
1895
1896 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1897 to read one or more bytes before the start of the subject string. Often this
1898 had no effect on PCRE's behaviour, but in some circumstances it could
1899 provoke a segmentation fault.
1900
1901 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1902 to try to read one or more bytes before the start of the subject string.
1903
1904 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1905 UTF-8 support could misbehave in various ways if the subject string
1906 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1907 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1908 back over UTF-8 characters.)
1909
1910
1911 Version 4.2 14-Apr-03
1912 ---------------------
1913
1914 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1915
1916 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1917 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1918 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1919 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1920 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1921 and BUILD_EXEEXT
1922 Note: automatic setting of the BUILD variables is not yet working
1923 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1924 compile-time but not at link-time
1925 [LINK]: use for linking executables only
1926 make different versions for Windows and non-Windows
1927 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1928 libraries
1929 [LINK_FOR_BUILD]: new variable
1930 [OBJEXT]: use throughout
1931 [EXEEXT]: use throughout
1932 <winshared>: new target
1933 <wininstall>: new target
1934 <dftables.o>: use native compiler
1935 <dftables>: use native linker
1936 <install>: handle Windows platform correctly
1937 <clean>: ditto
1938 <check>: ditto
1939 copy DLL to top builddir before testing
1940
1941 As part of these changes, -no-undefined was removed again. This was reported
1942 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1943 in any case.
1944
1945 3. Some tidies to get rid of compiler warnings:
1946
1947 . In the match_data structure, match_limit was an unsigned long int, whereas
1948 match_call_count was an int. I've made them both unsigned long ints.
1949
1950 . In pcretest the fact that a const uschar * doesn't automatically cast to
1951 a void * provoked a warning.
1952
1953 . Turning on some more compiler warnings threw up some "shadow" variables
1954 and a few more missing casts.
1955
1956 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1957 option, a class that contained a single character with a value between 128
1958 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1959
1960 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1961 option, a class that contained several characters, but with at least one
1962 whose value was between 128 and 255 caused PCRE to crash.
1963
1964
1965 Version 4.1 12-Mar-03
1966 ---------------------
1967
1968 1. Compiling with gcc -pedantic found a couple of places where casts were
1969 needed, and a string in dftables.c that was longer than standard compilers are
1970 required to support.
1971
1972 2. Compiling with Sun's compiler found a few more places where the code could
1973 be tidied up in order to avoid warnings.
1974
1975 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1976 first of these names is deprecated in the latest Autoconf in favour of the name
1977 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1978 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1979 analogy I have changed it to CFLAGS_FOR_BUILD.
1980
1981 4. Added -no-undefined to the linking command in the Makefile, because this is
1982 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1983 linking step for the pcreposix library.
1984
1985 5. PCRE was failing to diagnose the case of two named groups with the same
1986 name.
1987
1988 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1989 literal character that is needed in the subject for a match, and scans along to
1990 ensure that it is present before embarking on the full matching process. This
1991 saves time in cases of nested unlimited repeats that are never going to match.
1992 Problem: the scan can take a lot of time if the subject is very long (e.g.
1993 megabytes), thus penalizing straightforward matches. It is now done only if the
1994 amount of subject to be scanned is less than 1000 bytes.
1995
1996 7. A lesser problem with the same optimization is that it was recording the
1997 first character of an anchored pattern as "needed", thus provoking a search
1998 right along the subject, even when the first match of the pattern was going to
1999 fail. The "needed" character is now not set for anchored patterns, unless it
2000 follows something in the pattern that is of non-fixed length. Thus, it still
2001 fulfils its original purpose of finding quick non-matches in cases of nested
2002 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2003
2004
2005 Version 4.0 17-Feb-03
2006 ---------------------
2007
2008 1. If a comment in an extended regex that started immediately after a meta-item
2009 extended to the end of string, PCRE compiled incorrect data. This could lead to
2010 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2011
2012 2. Moved to autoconf 2.53 and libtool 1.4.2.
2013
2014 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2015 the special perltest8 script is no longer needed - all the tests can be run
2016 from a single perltest script.
2017
2018 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2019 by \s. It has now been removed in PCRE. This means it isn't recognized as
2020 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2021 class [:space:] *does* include VT, thereby creating a mess.
2022
2023 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2024 space and tab.
2025
2026 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2027 its new features into the main test script, reducing the number of scripts.
2028
2029 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2030 were backward compatible, and made the (?i) apply to the whole pattern, as if
2031 /i were given. Now it behaves more logically, and applies the option setting
2032 only to what follows. PCRE has been changed to follow suit. However, if it
2033 finds options settings right at the start of the pattern, it extracts them into
2034 the global options, as before. Thus, they show up in the info data.
2035
2036 8. Added support for the \Q...\E escape sequence. Characters in between are
2037 treated as literals. This is slightly different from Perl in that $ and @ are
2038 also handled as literals inside the quotes. In Perl, they will cause variable
2039 interpolation. Note the following examples:
2040
2041 Pattern PCRE matches Perl matches
2042
2043 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2044 \Qabc\$xyz\E abc\$xyz abc\$xyz
2045 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2046
2047 For compatibility with Perl, \Q...\E sequences are recognized inside character
2048 classes as well as outside them.
2049
2050 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2051 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2052 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2053 signed/unsigned warnings.
2054
2055 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2056 option for pcretest, so I've replaced it by a simple function that does just
2057 that job.
2058
2059 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2060 "pcregrep -".
2061
2062 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2063 Java package. This provides some syntactic sugar for simple cases of what my
2064 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2065 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2066 item, you can use this simplified notation. Note that only makes sense with
2067 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2068 greediness, whatever the setting of the PCRE_UNGREEDY option.
2069
2070 13. A change of greediness default within a pattern was not taking effect at
2071 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2072 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2073 was abstracted outside.
2074
2075 14. PCRE now supports the \G assertion. It is true when the current matching
2076 position is at the start point of the match. This differs from \A when the
2077 starting offset is non-zero. Used with the /g option of pcretest (or similar
2078 code), it works in the same way as it does for Perl's /g option. If all
2079 alternatives of a regex begin with \G, the expression is anchored to the start
2080 match position, and the "anchored" flag is set in the compiled expression.
2081
2082 15. Some bugs concerning the handling of certain option changes within patterns
2083 have been fixed. These applied to options other than (?ims). For example,
2084 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2085 been the other way round. Some of this was related to change 7 above.
2086
2087 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2088 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2089 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2090 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2091
2092 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2093 mode. Unlike ".", it always matches newline, whatever the setting of
2094 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2095 assertions. Perl allows it, but it doesn't (in general) work because it can't
2096 calculate the length of the lookbehind. At least, that's the case for Perl
2097 5.8.0 - I've been told they are going to document that it doesn't work in
2098 future.
2099
2100 18. Added an error diagnosis for escapes that PCRE does not support: these are
2101 \L, \l, \N, \P, \p, \U, \u, and \X.
2102
2103 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2104 reading past the end of the pattern in cases such as /[abcd/.
2105
2106 20. PCRE was getting more memory than necessary for patterns with classes that
2107 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2108
2109 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2110 compiling PCRE for use with Virtual Pascal.
2111
2112 22. Small fix to the Makefile to make it work properly if the build is done
2113 outside the source tree.
2114
2115 23. Added a new extension: a condition to go with recursion. If a conditional
2116 subpattern starts with (?(R) the "true" branch is used if recursion has
2117 happened, whereas the "false" branch is used only at the top level.
2118
2119 24. When there was a very long string of literal characters (over 255 bytes
2120 without UTF support, over 250 bytes with UTF support), the computation of how
2121 much memory was required could be incorrect, leading to segfaults or other
2122 strange effects.
2123
2124 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2125 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2126 there was a subsequent back reference to those brackets. This meant that, for
2127 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2128 possible to check for precisely this case. All we can do is abandon the
2129 optimization if .* occurs inside capturing brackets when there are any back
2130 references whatsoever. (See below for a better fix that came later.)
2131
2132 26. The handling of the optimization for finding the first character of a
2133 non-anchored pattern, and for finding a character that is required later in the
2134 match were failing in some cases. This didn't break the matching; it just
2135 failed to optimize when it could. The way this is done has been re-implemented.
2136
2137 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2138
2139 28. Added a new feature that provides some of the functionality that Perl
2140 provides with (?{...}). The facility is termed a "callout". The way it is done
2141 in PCRE is for the caller to provide an optional function, by setting
2142 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2143 global variable. By default it is unset, which disables all calling out. To get
2144 the function called, the regex must include (?C) at appropriate points. This
2145 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2146 This provides a means of identifying different callout points. When PCRE
2147 reaches such a point in the regex, if pcre_callout has been set, the external
2148 function is called. It is provided with data in a structure called
2149 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2150 matching continues; if it returns a non-zero value, the match at the current
2151 point fails. However, backtracking will occur if possible. [This was changed
2152 later and other features added - see item 49 below.]
2153
2154 29. pcretest is upgraded to test the callout functionality. It provides a
2155 callout function that displays information. By default, it shows the start of
2156 the match and the current position in the text. There are some new data escapes
2157 to vary what happens:
2158
2159 \C+ in addition, show current contents of captured substrings
2160 \C- do not supply a callout function
2161 \C!n return 1 when callout number n is reached
2162 \C!n!m return 1 when callout number n is reached for the mth time
2163
2164 30. If pcregrep was called with the -l option and just a single file name, it
2165 output "<stdin>" if a match was found, instead of the file name.
2166
2167 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2168 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2169 pcre_exec(). This saves a malloc/free per call. The default value of
2170 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2171 when configuring.
2172
2173 32. The default maximum size of a compiled pattern is 64K. There have been a
2174 few cases of people hitting this limit. The code now uses macros to handle the
2175 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2176 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2177 configuring. Tests 2 and 5 work only with 2-byte links because they output
2178 debugging information about compiled patterns.
2179
2180 33. Internal code re-arrangements:
2181
2182 (a) Moved the debugging function for printing out a compiled regex into
2183 its own source file (printint.c) and used #include to pull it into
2184 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2185 separate copies.
2186
2187 (b) Defined the list of op-code names for debugging as a macro in
2188 internal.h so that it is next to the definition of the opcodes.
2189
2190 (c) Defined a table of op-code lengths for simpler skipping along compiled
2191 code. This is again a macro in internal.h so that it is next to the
2192 definition of the opcodes.
2193
2194 34. Added support for recursive calls to individual subpatterns, along the
2195 lines of Robin Houston's patch (but implemented somewhat differently).
2196
2197 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2198 allow it to read and process whole directories in Win32. This code was
2199 contributed by Lionel Fourquaux; it has not been tested by me.
2200
2201 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2202 used to name a group. Names consist of alphanumerics and underscores, and must
2203 be unique. Back references use the syntax (?P=name) and recursive calls use
2204 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2205 numbers. The function pcre_fullinfo() can be used after compilation to extract
2206 a name/number map. There are three relevant calls:
2207
2208 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2209 PCRE_INFO_NAMECOUNT yields the number of entries
2210 PCRE_INFO_NAMETABLE yields a pointer to the map.
2211
2212 The map is a vector of fixed-size entries. The size of each entry depends on
2213 the length of the longest name used. The first two bytes of each entry are the
2214 group number, most significant byte first. There follows the corresponding
2215 name, zero terminated. The names are in alphabetical order.
2216
2217 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2218 case instead of 255. Making it the same both with and without UTF-8 support
2219 means that the same test output works with both.
2220
2221 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2222 calling malloc() with a zero argument.
2223
2224 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2225 optimization. I've improved things by keeping a bitmap of backreferences with
2226 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2227 fact referenced, the optimization can be applied. It is unlikely that a
2228 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2229 the match to follow \n) will appear inside brackets with a number greater than
2230 31, but if it does, any back reference > 31 suppresses the optimization.
2231
2232 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2233 of disabling numbered capturing parentheses. Any opening parenthesis that is
2234 not followed by ? behaves as if it were followed by ?: but named parentheses
2235 can still be used for capturing (and they will acquire numbers in the usual
2236 way).
2237
2238 41. Redesigned the return codes from the match() function into yes/no/error so
2239 that errors can be passed back from deep inside the nested calls. A malloc
2240 failure while inside a recursive subpattern call now causes the
2241 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2242
2243 42. It is now possible to set a limit on the number of times the match()
2244 function is called in a call to pcre_exec(). This facility makes it possible to
2245 limit the amount of recursion and backtracking, though not in a directly
2246 obvious way, because the match() function is used in a number of different
2247 circumstances. The count starts from zero for each position in the subject
2248 string (for non-anchored patterns). The default limit is, for compatibility, a
2249 large number, namely 10 000 000. You can change this in two ways:
2250
2251 (a) When configuring PCRE before making, you can use --with-match-limit=n
2252 to set a default value for the compiled library.
2253
2254 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2255 a different value is set. See 45 below.
2256
2257 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2258
2259 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2260 of things that can be changed at compile time. The first argument specifies
2261 what is wanted and the second points to where the information is to be placed.
2262 The current list of available information is:
2263
2264 PCRE_CONFIG_UTF8
2265
2266 The output is an integer that is set to one if UTF-8 support is available;
2267 otherwise it is set to zero.
2268
2269 PCRE_CONFIG_NEWLINE
2270
2271 The output is an integer that it set to the value of the code that is used for
2272 newline. It is either LF (10) or CR (13).
2273
2274 PCRE_CONFIG_LINK_SIZE
2275
2276 The output is an integer that contains the number of bytes used for internal
2277 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2278
2279 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2280
2281 The output is an integer that contains the threshold above which the POSIX
2282 interface uses malloc() for output vectors. See item 31 above.
2283
2284 PCRE_CONFIG_MATCH_LIMIT
2285
2286 The output is an unsigned integer that contains the default limit of the number
2287 of match() calls in a pcre_exec() execution. See 42 above.
2288
2289 44. pcretest has been upgraded by the addition of the -C option. This causes it
2290 to extract all the available output from the new pcre_config() function, and to
2291 output it. The program then exits immediately.
2292
2293 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2294 order to support additional features. One way would have been to define
2295 pcre_exec2() (for example) with extra arguments, but this would not have been
2296 extensible, and would also have required all calls to the original function to
2297 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2298 is used for passing in "extra" data from pcre_study().
2299
2300 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2301 contains the following fields:
2302
2303 flags a bitmap indicating which of the following fields are set
2304 study_data opaque data from pcre_study()
2305 match_limit a way of specifying a limit on match() calls for a specific
2306 call to pcre_exec()
2307 callout_data data for callouts (see 49 below)
2308
2309 The flag bits are also defined in pcre.h, and are
2310
2311 PCRE_EXTRA_STUDY_DATA
2312 PCRE_EXTRA_MATCH_LIMIT
2313 PCRE_EXTRA_CALLOUT_DATA
2314
2315 The pcre_study() function now returns one of these new pcre_extra blocks, with
2316 the actual study data pointed to by the study_data field, and the
2317 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2318 before. That is, this change is entirely upwards-compatible and requires no
2319 change to existing code.
2320
2321 If you want to pass in additional data to pcre_exec(), you can either place it
2322 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2323 block.
2324
2325 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2326 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2327 times with different match limits, until it finds the minimum value needed for
2328 pcre_exec() to complete. The value is then output. This can be instructive; for
2329 most simple matches the number is quite small, but for pathological cases it
2330 gets very large very quickly.
2331
2332 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2333 returns the size of the data block pointed to by the study_data field in a
2334 pcre_extra block, that is, the value that was passed as the argument to
2335 pcre_malloc() when PCRE was getting memory in which to place the information
2336 created by pcre_study(). The fourth argument should point to a size_t variable.
2337 pcretest has been extended so that this information is shown after a successful
2338 pcre_study() call when information about the compiled regex is being displayed.
2339
2340 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2341 because what follows is always an absolute path. (Later: it turns out that this
2342 is more than cosmetic for MinGW, because it doesn't like empty path
2343 components.)
2344
2345 49. Some changes have been made to the callout feature (see 28 above):
2346
2347 (i) A callout function now has three choices for what it returns:
2348
2349 0 => success, carry on matching
2350 > 0 => failure at this point, but backtrack if possible
2351 < 0 => serious error, return this value from pcre_exec()
2352
2353 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2354 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2355 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2356 use by callout functions. It will never be used by PCRE itself.
2357
2358 (ii) The pcre_extra structure (see 45 above) has a void * field called
2359 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2360 pcre_callout_block structure has a field of the same name. The contents of
2361 the field passed in the pcre_extra structure are passed to the callout
2362 function in the corresponding field in the callout block. This makes it
2363 easier to use the same callout-containing regex from multiple threads. For
2364 testing, the pcretest program has a new data escape
2365
2366 \C*n pass the number n (may be negative) as callout_data
2367
2368 If the callout function in pcretest receives a non-zero value as
2369 callout_data, it returns that value.
2370
2371 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2372 there were some redundant $(CFLAGS) in commands that are now specified as
2373 $(LINK), which already includes $(CFLAGS).
2374
2375 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2376 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2377 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2378 one-byte characters throughout. Note that case-insensitive matching applies
2379 only to characters whose values are less than 256. PCRE doesn't support the
2380 notion of cases for higher-valued characters.
2381
2382 (i) A character class whose characters are all within 0-255 is handled as
2383 a bit map, and the map is inverted for negative classes. Previously, a
2384 character > 255 always failed to match such a class; however it should
2385 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2386
2387 (ii) A negated character class with a single character < 255 is coded as
2388 "not this character" (OP_NOT). This wasn't working properly when the test
2389 character was multibyte, either singly or repeated.
2390
2391 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2392 mode, for example: \x{100}{2,3}.
2393
2394 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2395 singly or repeated) now correctly test multibyte characters. However,
2396 PCRE doesn't recognize any characters with values greater than 255 as
2397 digits, spaces, or word characters. Such characters always match \D, \S,
2398 and \W, and never match \d, \s, or \w.
2399
2400 (v) Classes may now contain characters and character ranges with values
2401 greater than 255. For example: [ab\x{100}-\x{400}].
2402
2403 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2404 PCRE in UTF-8 mode.
2405
2406 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2407 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2408 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2409 value.)
2410
2411 53. The single man page has become too large. I have therefore split it up into
2412 a number of separate man pages. These also give rise to individual HTML pages;
2413 these are now put in a separate directory, and there is an index.html page that
2414 lists them all. Some hyperlinking between the pages has been installed.
2415
2416 54. Added convenience functions for handling named capturing parentheses.
2417
2418 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2419 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2420 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2421 are faulted.
2422
2423 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2424 calling configure. These values are used when compiling the dftables.c program
2425 which is run to generate the source of the default character tables. They
2426 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2427 you will need to set these values.
2428
2429 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2430
2431
2432 Version 3.9 02-Jan-02
2433 ---------------------
2434
2435 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2436
2437 2. If --disable-static was given, the building process failed when trying to
2438 build pcretest and pcregrep. (For some reason it was using libtool to compile
2439 them, which is not right, as they aren't part of the library.)
2440
2441
2442 Version 3.8 18-Dec-01
2443 ---------------------
2444
2445 1. The experimental UTF-8 code was completely screwed up. It was packing the
2446 bytes in the wrong order. How dumb can you get?
2447
2448
2449 Version 3.7 29-Oct-01
2450 ---------------------
2451
2452 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2453 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2454 this didn't happen under Solaris 8, where I normally test things.
2455
2456 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2457 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2458 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2459 this fix an hour or so after the initial 3.7 release.)
2460
2461
2462 Version 3.6 23-Oct-01
2463 ---------------------
2464
2465 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2466 offsets passed as NULL with zero offset count.
2467
2468 2. The config.guess and config.sub files had not been updated when I moved to
2469 the latest autoconf.
2470
2471
2472 Version 3.5 15-Aug-01
2473 ---------------------
2474
2475 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2476 had been forgotten.
2477
2478 2. By using declared but undefined structures, we can avoid using "void"
2479 definitions in pcre.h while keeping the internal definitions of the structures
2480 private.
2481
2482 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2483 user point of view, this means that both static and shared libraries are built
2484 by default, but this can be individually controlled. More of the work of
2485 handling this static/shared cases is now inside libtool instead of PCRE's make
2486 file.
2487
2488 4. The pcretest utility is now installed along with pcregrep because it is
2489 useful for users (to test regexs) and by doing this, it automatically gets
2490 relinked by libtool. The documentation has been turned into a man page, so
2491 there are now .1, .txt, and .html versions in /doc.
2492
2493 5. Upgrades to pcregrep:
2494 (i) Added long-form option names like gnu grep.
2495 (ii) Added --help to list all options with an explanatory phrase.
2496 (iii) Added -r, --recursive to recurse into sub-directories.
2497 (iv) Added -f, --file to read patterns from a file.
2498
2499 6. pcre_exec() was referring to its "code" argument before testing that
2500 argument for NULL (and giving an error if it was NULL).
2501
2502 7. Upgraded Makefile.in to allow for compiling in a different directory from
2503 the source directory.
2504
2505 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2506 options bits, the pointer it was passed was to an int instead of to an unsigned
2507 long int. This mattered only on 64-bit systems.
2508
2509 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2510 generated) instead of pcre.in, which it its source. Also made the same change
2511 in several of the .c files.
2512
2513 10. A new release of gcc defines printf() as a macro, which broke pcretest
2514 because it had an ifdef in the middle of a string argument for printf(). Fixed
2515 by using separate calls to printf().
2516
2517 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2518 script, to force use of CR or LF instead of \n in the source. On non-Unix
2519 systems, the value can be set in config.h.
2520
2521 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2522 absolute limit. Changed the text of the error message to make this clear, and
2523 likewise updated the man page.
2524
2525 13. The limit of 99 on the number of capturing subpatterns has been removed.
2526 The new limit is 65535, which I hope will not be a "real" limit.
2527
2528
2529 Version 3.4 22-Aug-00
2530 ---------------------
2531
2532 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2533
2534 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2535
2536
2537 Version 3.3 01-Aug-00
2538 ---------------------
2539
2540 1. If an octal character was given, but the value was greater than \377, it
2541 was not getting masked to the least significant bits, as documented. This could
2542 lead to crashes in some systems.
2543
2544 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2545 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2546
2547 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2548 These just pass their arguments on to (pcre_free)(), but they are provided
2549 because some uses of PCRE bind it to non-C systems that can call its functions,
2550 but cannot call free() or pcre_free() directly.
2551
2552 4. Add "make test" as a synonym for "make check". Corrected some comments in
2553 the Makefile.
2554
2555 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2556 Makefile.
2557
2558 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2559 command called pgrep for grepping around the active processes.
2560
2561 7. Added the beginnings of support for UTF-8 character strings.
2562
2563 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2564 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2565 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2566 out for the ar command.)
2567
2568
2569 Version 3.2 12-May-00
2570 ---------------------
2571
2572 This is purely a bug fixing release.
2573
2574 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2575 of ZA. This was just one example of several cases that could provoke this bug,
2576 which was introduced by change 9 of version 2.00. The code for breaking
2577 infinite loops after an iteration that matches an empty string was't working
2578 correctly.
2579
2580 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2581 when matched against abbab (for example). After matching an empty string, it
2582 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2583 caused it to match further down the string than it should.
2584
2585 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2586 was there because it doesn't seem to be needed, and it causes trouble on some
2587 systems, as it is not a Standard C header. It has been removed.
2588
2589 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2590 were reported on the Macintosh. The changes were from
2591
2592 while ((c = *(++ptr)) != 0 && c != '\n');
2593 to
2594 while ((c = *(++ptr)) != 0 && c != '\n') ;
2595
2596 Totally extraordinary, but if that's what it takes...
2597
2598 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2599 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2600 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2601 assumes the way PCRE uses memmove() (always moving upwards).
2602
2603 6. PCRE is being used in one environment where strchr() is not available. There
2604 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2605 faster code anyway.
2606
2607
2608 Version 3.1 09-Feb-00
2609 ---------------------
2610
2611 The only change in this release is the fixing of some bugs in Makefile.in for
2612 the "install" target:
2613
2614 (1) It was failing to install pcreposix.h.
2615
2616 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2617
2618
2619 Version 3.0 01-Feb-00
2620 ---------------------
2621
2622 1. Add support for the /+ modifier to perltest (to output $` like it does in
2623 pcretest).
2624
2625 2. Add support for the /g modifier to perltest.
2626
2627 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2628 matches null strings.
2629
2630 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2631 pattern. Perl treats empty patterns specially - it reuses the most recent
2632 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2633 effect.
2634
2635 5. The POSIX interface was broken in that it was just handing over the POSIX
2636 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2637 required a bigger vector, with some working space on the end. This means that
2638 the POSIX wrapper now has to get and free some memory, and copy the results.
2639
2640 6. Added some simple autoconf support, placing the test data and the
2641 documentation in separate directories, re-organizing some of the
2642 information files, and making it build pcre-config (a GNU standard). Also added
2643 libtool support for building PCRE as a shared library, which is now the
2644 default.
2645
2646 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2647 09 are not valid octal constants. Single digits will be used for minor values
2648 less than 10.
2649
2650 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2651 existing programs that set these in the POSIX interface can use PCRE without
2652 modification.
2653
2654 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2655 return all that pcre_info() returns, plus additional data. The pcre_info()
2656 function is retained for compatibility, but is considered to be obsolete.
2657
2658 10. Added experimental recursion feature (?R) to handle one common case that
2659 Perl 5.6 will be able to do with (?p{...}).
2660
2661 11. Added support for POSIX character classes like [:alpha:], which Perl is
2662 adopting.
2663
2664
2665 Version 2.08 31-Aug-99
2666 ----------------------
2667
2668 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2669 trying to match at the startoffset position, but instead was moving forward to
2670 the next newline as if a previous match had failed.
2671
2672 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2673 and could get into a loop if a null string was matched other than at the start
2674 of the subject.
2675
2676 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2677 be distinguished at compile time, and for completeness also added PCRE_DATE.
2678
2679 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2680 in GnuWin32 environments.
2681
2682
2683 Version 2.07 29-Jul-99
2684 ----------------------
2685
2686 1. The documentation is now supplied in plain text form and HTML as well as in
2687 the form of man page sources.
2688
2689 2. C++ compilers don't like assigning (void *) values to other pointer types.
2690 In particular this affects malloc(). Although there is no problem in Standard
2691 C, I've put in casts to keep C++ compilers happy.
2692
2693 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2694 should be (const char *).
2695
2696 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2697 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2698 However, I haven't made this a standard facility. The documentation doesn't
2699 mention it, and the Makefile doesn't support it.
2700
2701 5. The Makefile now contains an "install" target, with editable destinations at
2702 the top of the file. The pcretest program is not installed.
2703
2704 6. pgrep -V now gives the PCRE version number and date.
2705
2706 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2707 causing the entire string to be ignored, instead of just the last character.
2708
2709 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2710 non-matching string, it can take a very, very long time, even for strings of
2711 quite modest length, because of the nested recursion. PCRE now does better in
2712 some of these cases. It does this by remembering the last required literal
2713 character in the pattern, and pre-searching the subject to ensure it is present
2714 before running the real match. In other words, it applies a heuristic to detect
2715 some types of certain failure quickly, and in the above example, if presented
2716 with a string that has no trailing " it gives "no match" very quickly.
2717
2718 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2719 other alternatives are tried instead.
2720
2721
2722 Version 2.06 09-Jun-99
2723 ----------------------
2724
2725 1. Change pcretest's output for amount of store used to show just the code
2726 space, because the remainder (the data block) varies in size between 32-bit and
2727 64-bit systems.
2728
2729 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2730 start matching at. This allows lookbehinds to work when searching for multiple
2731 occurrences in a string.
2732
2733 3. Added additional options to pcretest for testing multiple occurrences:
2734
2735 /+ outputs the rest of the string that follows a match
2736 /g loops for multiple occurrences, using the new startoffset argument
2737 /G loops for multiple occurrences by passing an incremented pointer
2738
2739 4. PCRE wasn't doing the "first character" optimization for patterns starting
2740 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2741 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2742 the letter 'x'. On long subject strings, this gives a significant speed-up.
2743
2744
2745 Version 2.05 21-Apr-99
2746 ----------------------
2747
2748 1. Changed the type of magic_number from int to long int so that it works
2749 properly on 16-bit systems.
2750
2751 2. Fixed a bug which caused patterns starting with .* not to work correctly
2752 when the subject string contained newline characters. PCRE was assuming
2753 anchoring for such patterns in all cases, which is not correct because .* will
2754 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2755 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2756 must be retried after every newline in the subject.
2757
2758
2759 Version 2.04 18-Feb-99
2760 ----------------------
2761
2762 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2763 computation of the store needed to hold the pattern was incorrect (too large).
2764 If such patterns were nested a few deep, this could multiply and become a real
2765 problem.
2766
2767 2. Added /M option to pcretest to show the memory requirement of a specific
2768 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2769
2770 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2771 compiled in such a way that the backtracking after subsequent failure was
2772 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2773 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2774
2775
2776 Version 2.03 02-Feb-99
2777 ----------------------
2778
2779 1. Fixed typo and small mistake in man page.
2780
2781 2. Added 4th condition (GPL supersedes if conflict) and created separate
2782 LICENCE file containing the conditions.
2783
2784 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2785 Perl, that is the internal \ allows the delimiter to be included in the
2786 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2787 the final delimiter, add \ to the end of the pattern (to test the error).
2788
2789 4. Added the convenience functions for extracting substrings after a successful
2790 match. Updated pcretest to make it able to test these functions.
2791
2792
2793 Version 2.02 14-Jan-99
2794 ----------------------
2795
2796 1. Initialized the working variables associated with each extraction so that
2797 their saving and restoring doesn't refer to uninitialized store.
2798
2799 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2800 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2801 fix the problem.
2802
2803 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2804 calls, and wasn't printing the correct value for compiling calls. Increased the
2805 default value of LOOPREPEAT, and the number of significant figures in the
2806 times.
2807
2808 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2809
2810 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2811 a building problem on Windows NT with a FAT file system.
2812
2813
2814 Version 2.01 21-Oct-98
2815 ----------------------
2816
2817 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2818 to character tables built by pcre_maketables() in the current locale. If NULL
2819 is passed, the default tables are used.
2820
2821
2822 Version 2.00 24-Sep-98
2823 ----------------------
2824
2825 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2826 it any more.
2827
2828 2. Allow quantification of (?>) groups, and make it work correctly.
2829
2830 3. The first character computation wasn't working for (?>) groups.
2831
2832 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2833 end of the subject) and add 5.005's \z, which really does match only at the
2834 very end of the subject.
2835
2836 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2837
2838 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2839 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2840 localized options. All options to pcre_study() were also removed.
2841
2842 7. Add other new features from 5.005:
2843
2844 $(?<= positive lookbehind
2845 $(?<! negative lookbehind
2846 (?imsx-imsx) added the unsetting capability
2847 such a setting is global if at outer level; local otherwise
2848 (?imsx-imsx:) non-capturing groups with option setting
2849 (?(cond)re|re) conditional pattern matching
2850
2851 A backreference to itself in a repeated group matches the previous
2852 captured string.
2853
2854 8. General tidying up of studying (both automatic and via "study")
2855 consequential on the addition of new assertions.
2856
2857 9. As in 5.005, unlimited repeated groups that could match an empty substring
2858 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2859 runtime if any iteration does actually match an empty substring.
2860
2861 10. Include the RunTest script in the distribution.
2862
2863 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2864 discrepancies, some of which were old and were also with respect to 5.004. They
2865 have now been fixed.
2866
2867
2868 Version 1.09 28-Apr-98
2869 ----------------------
2870
2871 1. A negated single character class followed by a quantifier with a minimum
2872 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2873 program crashes, or just wrong answers. This did not apply to negated classes
2874 containing more than one character, or to minima other than one.
2875
2876
2877 Version 1.08 27-Mar-98
2878 ----------------------
2879
2880 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2881
2882 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2883 latter must appear before anything that relies on it in the pattern.
2884
2885
2886 Version 1.07 16-Feb-98
2887 ----------------------
2888
2889 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2890 repeat of a potentially empty string).
2891
2892
2893 Version 1.06 23-Jan-98
2894 ----------------------
2895
2896 1. Added Markus Oberhumer's little patches for C++.
2897
2898 2. Literal strings longer than 255 characters were broken.
2899
2900
2901 Version 1.05 23-Dec-97
2902 ----------------------
2903
2904 1. Negated character classes containing more than one character were failing if
2905 PCRE_CASELESS was set at run time.
2906
2907
2908 Version 1.04 19-Dec-97
2909 ----------------------
2910
2911 1. Corrected the man page, where some "const" qualifiers had been omitted.
2912
2913 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2914 input syntax.
2915
2916 3. Fixed memory leak which occurred when a regex with back references was
2917 matched with an offsets vector that wasn't big enough. The temporary memory
2918 that is used in this case wasn't being freed if the match failed.
2919
2920 4. Tidied pcretest to ensure it frees memory that it gets.
2921
2922 5. Temporary memory was being obtained in the case where the passed offsets
2923 vector was exactly big enough.
2924
2925 6. Corrected definition of offsetof() from change 5 below.
2926
2927 7. I had screwed up change 6 below and broken the rules for the use of
2928 setjmp(). Now fixed.
2929
2930
2931 Version 1.03 18-Dec-97
2932 ----------------------
2933
2934 1. A erroneous regex with a missing opening parenthesis was correctly
2935 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2936 on some systems.
2937
2938 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2939 it was reported that one broken compiler failed on the former because "code" is
2940 also an independent variable.
2941
2942 3. The erroneous regex a[]b caused an array overrun reference.
2943
2944 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2945 fail on data ending with that character. (It was going on too far, and checking
2946 the next character, typically a binary zero.) This was specific to the
2947 optimized code for single-character negative classes.
2948
2949 5. Added a contributed patch from the TIN world which does the following:
2950
2951 + Add an undef for memmove, in case the the system defines a macro for it.
2952
2953 + Add a definition of offsetof(), in case there isn't one. (I don't know
2954 the reason behind this - offsetof() is part of the ANSI standard - but
2955 it does no harm).
2956
2957 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2958 most of the places where whitespace preceded '#'. I have given up and
2959 allowed the remaining 2 cases to be at the margin.
2960
2961 + Rename some variables in pcre to eliminate shadowing. This seems very
2962 pedantic, but does no harm, of course.
2963
2964 6. Moved the call to setjmp() into its own function, to get rid of warnings
2965 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2966
2967 7. Constructs such as \d{8,} were compiling into the equivalent of
2968 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2969 outcome, but in this particular case used more store than had been allocated,
2970 which caused the bug to be discovered because it threw up an internal error.
2971
2972 8. The debugging code in both pcre and pcretest for outputting the compiled
2973 form of a regex was going wrong in the case of back references followed by
2974 curly-bracketed repeats.
2975
2976
2977 Version 1.02 12-Dec-97
2978 ----------------------
2979
2980 1. Typos in pcre.3 and comments in the source fixed.
2981
2982 2. Applied a contributed patch to get rid of places where it used to remove
2983 'const' from variables, and fixed some signed/unsigned and uninitialized
2984 variable warnings.
2985
2986 3. Added the "runtest" target to Makefile.
2987
2988 4. Set default compiler flag to -O2 rather than just -O.
2989
2990
2991 Version 1.01 19-Nov-97
2992 ----------------------
2993
2994 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2995 like /([ab]*)*/, that is, for classes with more than one character in them.
2996
2997 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2998 as /((?>a*))*/ (a PCRE_EXTRA facility).
2999
3000
3001 Version 1.00 18-Nov-97
3002 ----------------------
3003
3004 1. Added compile-time macros to support systems such as SunOS4 which don't have
3005 memmove() or strerror() but have other things that can be used instead.
3006
3007 2. Arranged that "make clean" removes the executables.
3008
3009
3010 Version 0.99 27-Oct-97
3011 ----------------------
3012
3013 1. Fixed bug in code for optimizing classes with only one character. It was
3014 initializing a 32-byte map regardless, which could cause it to run off the end
3015 of the memory it had got.
3016
3017 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3018
3019
3020 Version 0.98 22-Oct-97
3021 ----------------------
3022
3023 1. Fixed bug in code for handling temporary memory usage when there are more
3024 back references than supplied space in the ovector. This could cause segfaults.
3025
3026
3027 Version 0.97 21-Oct-97
3028 ----------------------
3029
3030 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3031
3032 2. Optimized negated single characters not to use a bit map.
3033
3034 3. Brought error texts together as macro definitions; clarified some of them;
3035 fixed one that was wrong - it said "range out of order" when it meant "invalid
3036 escape sequence".
3037
3038 4. Changed some char * arguments to const char *.
3039
3040 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3041
3042 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3043 pcretest.
3044
3045
3046 Version 0.96 16-Oct-97
3047 ----------------------
3048
3049 1. Added a simple "pgrep" utility to the distribution.
3050
3051 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3052 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3053 where "ddd" means "one or more decimal digits".
3054
3055 3. Fixed serious bug. If a pattern had a back reference, but the call to
3056 pcre_exec() didn't supply a large enough ovector to record the related
3057 identifying subpattern, the match always failed. PCRE now remembers the number
3058 of the largest back reference, and gets some temporary memory in which to save
3059 the offsets during matching if necessary, in order to ensure that
3060 backreferences always work.
3061
3062 4. Increased the compatibility with Perl in a number of ways:
3063
3064 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3065 to request this handling. The option can be set at compile or exec time.
3066
3067 (b) $ matches before a terminating newline by default; an option
3068 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3069 mode). The option can be set at compile or exec time.
3070
3071 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3072 the same as Perl's. If the decimal number it represents is less than 10
3073 or there aren't that many previous left capturing parentheses, an octal
3074 escape is read. Inside a character class, it's always an octal escape,
3075 even if it is a single digit.
3076
3077 (d) An escaped but undefined alphabetic character is taken as a literal,
3078 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3079 escapes.
3080
3081 (e) {0} is now permitted. (The previous item is removed from the compiled
3082 pattern).
3083
3084 5. Changed all the names of code files so that the basic parts are no longer
3085 than 10 characters, and abolished the teeny "globals.c" file.
3086
3087 6. Changed the handling of character classes; they are now done with a 32-byte
3088 bit map always.
3089
3090 7. Added the -d and /D options to pcretest to make it possible to look at the
3091 internals of compilation without having to recompile pcre.
3092
3093
3094 Version 0.95 23-Sep-97
3095 ----------------------
3096
3097 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3098 \x20 at the start of a run of normal characters. These were being treated as
3099 real characters, instead of the source characters being re-checked.
3100
3101
3102 Version 0.94 18-Sep-97
3103 ----------------------
3104
3105 1. The functions are now thread-safe, with the caveat that the global variables
3106 containing pointers to malloc() and free() or alternative functions are the
3107 same for all threads.
3108
3109 2. Get pcre_study() to generate a bitmap of initial characters for non-
3110 anchored patterns when this is possible, and use it if passed to pcre_exec().
3111
3112
3113 Version 0.93 15-Sep-97
3114 ----------------------
3115
3116 1. /(b)|(:+)/ was computing an incorrect first character.
3117
3118 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3119 but not actually doing anything yet.
3120
3121 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3122 as Perl does (e.g. [-az] or [az-]).
3123
3124 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3125 all possible positions.
3126
3127 5. Split up into different modules to avoid including unneeded functions in a
3128 compiled binary. However, compile and exec are still in one module. The "study"
3129 function is split off.
3130
3131 6. The character tables are now in a separate module whose source is generated
3132 by an auxiliary program - but can then be edited by hand if required. There are
3133 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3134 toupper() in the code.
3135
3136 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3137 make them global. Abolish the function for setting them, as the caller can now
3138 set them directly.
3139
3140
3141 Version 0.92 11-Sep-97
3142 ----------------------
3143
3144 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3145 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3146
3147 2. Caseless matching was not working in character classes if the characters in
3148 the pattern were in upper case.
3149
3150 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3151
3152 4. Make PCRE_ANCHORED public and accept as a compile option.
3153
3154 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3155 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3156 pass them.
3157
3158 6. Give an error if bad option bits passed at compile or run time.
3159
3160 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3161 pcretest to cause it to pass that flag.
3162
3163 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3164 options, and the first character, if set.
3165
3166 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3167
3168
3169 Version 0.91 10-Sep-97
3170 ----------------------
3171
3172 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3173 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3174
3175 2. PCRE was looping on encountering an indefinitely repeated back reference to
3176 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3177 Perl does - treats the match as successful.
3178
3179 ****

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