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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 303 - (show annotations) (download)
Sun Jan 20 19:39:43 2008 UTC (6 years, 8 months ago) by ph10
File size: 148975 byte(s)
Apply Christian Ehrlicher's patch to update the CMake files.

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

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