/[pcre]/code/tags/pcre-7.8/ChangeLog
ViewVC logotype

Contents of /code/tags/pcre-7.8/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 323 - (show annotations) (download)
Wed Mar 5 17:23:42 2008 UTC (6 years, 4 months ago) by ph10
Original Path: code/trunk/ChangeLog
File size: 151294 byte(s)
Remove a line of dead code, identified by coverity and reported by Nuno Lopes.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.7 05-Mar-08
5 ---------------------
6
7 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
8 a string to a long long, pretend we don't even have a long long." This is
9 done by checking for the strtoq, strtoll, and _strtoi64 functions.
10
11 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
12 pre-7.6 versions, which defined a global no_arg variable instead of putting
13 it in the RE class.
14
15 3. Remove a line of dead code, identified by coverity and reported by Nuno
16 Lopes.
17
18
19 Version 7.6 28-Jan-08
20 ---------------------
21
22 1. A character class containing a very large number of characters with
23 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
24 overflow.
25
26 2. Patch to cut out the "long long" test in pcrecpp_unittest when
27 HAVE_LONG_LONG is not defined.
28
29 3. Applied Christian Ehrlicher's patch to update the CMake build files to
30 bring them up to date and include new features. This patch includes:
31
32 - Fixed PH's badly added libz and libbz2 support.
33 - Fixed a problem with static linking.
34 - Added pcredemo. [But later removed - see 7 below.]
35 - Fixed dftables problem and added an option.
36 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
37 HAVE_LONG_LONG.
38 - Added readline support for pcretest.
39 - Added an listing of the option settings after cmake has run.
40
41 4. A user submitted a patch to Makefile that makes it easy to create
42 "pcre.dll" under mingw when using Configure/Make. I added stuff to
43 Makefile.am that cause it to include this special target, without
44 affecting anything else. Note that the same mingw target plus all
45 the other distribution libraries and programs are now supported
46 when configuring with CMake (see 6 below) instead of with
47 Configure/Make.
48
49 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
50 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
51 exported in the Windows port". It has not yet been confirmed that the patch
52 solves the problem, but it does no harm.
53
54 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
55 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
56 with CMake, and also correct the comment about stack recursion.
57
58 7. Remove the automatic building of pcredemo from the ./configure system and
59 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
60 of a program that users should build themselves after PCRE is installed, so
61 building it automatically is not really right. What is more, it gave
62 trouble in some build environments.
63
64 8. Further tidies to CMakeLists.txt from Sheri and Christian.
65
66
67 Version 7.5 10-Jan-08
68 ---------------------
69
70 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
71 values in parens when parsing an RE using the C++ wrapper."
72
73 2. Negative specials like \S did not work in character classes in UTF-8 mode.
74 Characters greater than 255 were excluded from the class instead of being
75 included.
76
77 3. The same bug as (2) above applied to negated POSIX classes such as
78 [:^space:].
79
80 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
81 defined or documented. It seems to have been a typo for PCRE_STATIC, so
82 I have changed it.
83
84 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
85 first named subpattern) and a construct such as (?&a) would reference the
86 first named subpattern whose name started with "a" (in other words, the
87 length check was missing). Both these problems are fixed. "Subpattern name
88 expected" is now given for (?&) (a zero-length name), and this patch also
89 makes it give the same error for \k'' (previously it complained that that
90 was a reference to a non-existent subpattern).
91
92 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
93 this is right because (?- can be followed by option settings as well as by
94 digits. I have, however, made the messages clearer.
95
96 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
97 than the number used in the conditional) now cause a compile-time error.
98 This is actually not compatible with Perl, which accepts such patterns, but
99 treats the conditional as always being FALSE (as PCRE used to), but it
100 seems to me that giving a diagnostic is better.
101
102 8. Change "alphameric" to the more common word "alphanumeric" in comments
103 and messages.
104
105 9. Fix two occurrences of "backslash" in comments that should have been
106 "backspace".
107
108 10. Remove two redundant lines of code that can never be obeyed (their function
109 was moved elsewhere).
110
111 11. The program that makes PCRE's Unicode character property table had a bug
112 which caused it to generate incorrect table entries for sequences of
113 characters that have the same character type, but are in different scripts.
114 It amalgamated them into a single range, with the script of the first of
115 them. In other words, some characters were in the wrong script. There were
116 thirteen such cases, affecting characters in the following ranges:
117
118 U+002b0 - U+002c1
119 U+0060c - U+0060d
120 U+0061e - U+00612
121 U+0064b - U+0065e
122 U+0074d - U+0076d
123 U+01800 - U+01805
124 U+01d00 - U+01d77
125 U+01d9b - U+01dbf
126 U+0200b - U+0200f
127 U+030fc - U+030fe
128 U+03260 - U+0327f
129 U+0fb46 - U+0fbb1
130 U+10450 - U+1049d
131
132 12. The -o option (show only the matching part of a line) for pcregrep was not
133 compatible with GNU grep in that, if there was more than one match in a
134 line, it showed only the first of them. It now behaves in the same way as
135 GNU grep.
136
137 13. If the -o and -v options were combined for pcregrep, it printed a blank
138 line for every non-matching line. GNU grep prints nothing, and pcregrep now
139 does the same. The return code can be used to tell if there were any
140 non-matching lines.
141
142 14. Added --file-offsets and --line-offsets to pcregrep.
143
144 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
145 infinitely looping recursion. The bug was that positive lookaheads were not
146 being skipped when checking for a possible empty match (negative lookaheads
147 and both kinds of lookbehind were skipped).
148
149 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
150 inclusion of <windows.h> to before rather than after the definition of
151 INVALID_FILE_ATTRIBUTES (patch from David Byron).
152
153 17. Specifying a possessive quantifier with a specific limit for a Unicode
154 character property caused pcre_compile() to compile bad code, which led at
155 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
156 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
157 caused the error; without that there was no problem.
158
159 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
160
161 19. Added --enable-pcretest-libreadline.
162
163 20. In pcrecpp.cc, the variable 'count' was incremented twice in
164 RE::GlobalReplace(). As a result, the number of replacements returned was
165 double what it should be. I removed one of the increments, but Craig sent a
166 later patch that removed the other one (the right fix) and added unit tests
167 that check the return values (which was not done before).
168
169 21. Several CMake things:
170
171 (1) Arranged that, when cmake is used on Unix, the libraries end up with
172 the names libpcre and libpcreposix, not just pcre and pcreposix.
173
174 (2) The above change means that pcretest and pcregrep are now correctly
175 linked with the newly-built libraries, not previously installed ones.
176
177 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
178
179 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
180 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
181 UTF-8 newline character). The key issue is that the pattern starts .*;
182 this means that the match must be either at the beginning, or after a
183 newline. The bug was in the code for advancing after a failed match and
184 checking that the new position followed a newline. It was not taking
185 account of UTF-8 characters correctly.
186
187 23. PCRE was behaving differently from Perl in the way it recognized POSIX
188 character classes. PCRE was not treating the sequence [:...:] as a
189 character class unless the ... were all letters. Perl, however, seems to
190 allow any characters between [: and :], though of course it rejects as
191 unknown any "names" that contain non-letters, because all the known class
192 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
193 for example, whereas PCRE did not - it did not recognize a POSIX character
194 class. This seemed a bit dangerous, so the code has been changed to be
195 closer to Perl. The behaviour is not identical to Perl, because PCRE will
196 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
197 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
198 Perl does, and where it didn't before.
199
200 24. Rewrite so as to remove the single use of %n from pcregrep because in some
201 Windows environments %n is disabled by default.
202
203
204 Version 7.4 21-Sep-07
205 ---------------------
206
207 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
208 means that a class such as [\s] counted as "explicit reference to CR or
209 LF". That isn't really right - the whole point of the change was to try to
210 help when there was an actual mention of one of the two characters. So now
211 the change happens only if \r or \n (or a literal CR or LF) character is
212 encountered.
213
214 2. The 32-bit options word was also used for 6 internal flags, but the numbers
215 of both had grown to the point where there were only 3 bits left.
216 Fortunately, there was spare space in the data structure, and so I have
217 moved the internal flags into a new 16-bit field to free up more option
218 bits.
219
220 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
221 but did not set the internal JCHANGED flag - either of these is enough to
222 control the way the "get" function works - but the PCRE_INFO_JCHANGED
223 facility is supposed to tell if (?J) was ever used, so now (?J) at the
224 start sets both bits.
225
226 4. Added options (at build time, compile time, exec time) to change \R from
227 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
228
229 5. doc/pcresyntax.html was missing from the distribution.
230
231 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
232 compatibility, even though it is no longer used.
233
234 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
235 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
236 windows.h file is present (where different names are used). [This was
237 reversed later after testing - see 16 below.]
238
239 8. Changed all #include <config.h> to #include "config.h". There were also
240 some further <pcre.h> cases that I changed to "pcre.h".
241
242 9. When pcregrep was used with the --colour option, it missed the line ending
243 sequence off the lines that it output.
244
245 10. It was pointed out to me that arrays of string pointers cause lots of
246 relocations when a shared library is dynamically loaded. A technique of
247 using a single long string with a table of offsets can drastically reduce
248 these. I have refactored PCRE in four places to do this. The result is
249 dramatic:
250
251 Originally: 290
252 After changing UCP table: 187
253 After changing error message table: 43
254 After changing table of "verbs" 36
255 After changing table of Posix names 22
256
257 Thanks to the folks working on Gregex for glib for this insight.
258
259 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
260 unicode-properties was also set.
261
262 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
263
264 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
265 checked only for CRLF.
266
267 14. Added casts to pcretest.c to avoid compiler warnings.
268
269 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
270
271 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
272 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
273 entirely. This removes changes made in 7 above.
274
275 17. The CMake files have been updated, and there is now more information about
276 building with CMake in the NON-UNIX-USE document.
277
278
279 Version 7.3 28-Aug-07
280 ---------------------
281
282 1. In the rejigging of the build system that eventually resulted in 7.1, the
283 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
284 brackets there is not right, since it causes compilers to look for an
285 installed pcre.h, not the version that is in the source that is being
286 compiled (which of course may be different). I have changed it back to:
287
288 #include "pcre.h"
289
290 I have a vague recollection that the change was concerned with compiling in
291 different directories, but in the new build system, that is taken care of
292 by the VPATH setting the Makefile.
293
294 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
295 when the subject happened to end in the byte 0x85 (e.g. if the last
296 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
297 characters but of course it shouldn't be taken as a newline when it is part
298 of another character. The bug was that, for an unlimited repeat of . in
299 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
300 characters when looking for a newline.
301
302 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
303
304 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
305 in debug output.
306
307 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
308 long printing in the pcrecpp unittest when running under MinGW.
309
310 6. ESC_K was left out of the EBCDIC table.
311
312 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
313 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
314 limit also applies to "virtual nesting" when a pattern is recursive, and in
315 this case 1000 isn't so big. I have been able to remove this limit at the
316 expense of backing off one optimization in certain circumstances. Normally,
317 when pcre_exec() would call its internal match() function recursively and
318 immediately return the result unconditionally, it uses a "tail recursion"
319 feature to save stack. However, when a subpattern that can match an empty
320 string has an unlimited repetition quantifier, it no longer makes this
321 optimization. That gives it a stack frame in which to save the data for
322 checking that an empty string has been matched. Previously this was taken
323 from the 1000-entry workspace that had been reserved. So now there is no
324 explicit limit, but more stack is used.
325
326 8. Applied Daniel's patches to solve problems with the import/export magic
327 syntax that is required for Windows, and which was going wrong for the
328 pcreposix and pcrecpp parts of the library. These were overlooked when this
329 problem was solved for the main library.
330
331 9. There were some crude static tests to avoid integer overflow when computing
332 the size of patterns that contain repeated groups with explicit upper
333 limits. As the maximum quantifier is 65535, the maximum group length was
334 set at 30,000 so that the product of these two numbers did not overflow a
335 32-bit integer. However, it turns out that people want to use groups that
336 are longer than 30,000 bytes (though not repeat them that many times).
337 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
338 made it possible to implement the integer overflow checks in a much more
339 dynamic way, which I have now done. The artificial limitation on group
340 length has been removed - we now have only the limit on the total length of
341 the compiled pattern, which depends on the LINK_SIZE setting.
342
343 10. Fixed a bug in the documentation for get/copy named substring when
344 duplicate names are permitted. If none of the named substrings are set, the
345 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
346 empty string.
347
348 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
349 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
350 because the ] is interpreted as the first data character and the
351 terminating ] is not found. PCRE has been made compatible with Perl in this
352 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
353 cause memory overwriting.
354
355 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
356 string has been matched (to stop an infinite loop). It was not recognizing
357 a conditional subpattern that could match an empty string if that
358 subpattern was within another subpattern. For example, it looped when
359 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
360 condition was not nested. This bug has been fixed.
361
362 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
363 past the start of the subject in the presence of bytes with the top bit
364 set, for example "\x8aBCD".
365
366 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
367 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
368
369 14. Optimized (?!) to (*FAIL).
370
371 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
372 This restricts code points to be within the range 0 to 0x10FFFF, excluding
373 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
374 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
375 does: it's just the validity check that is more restrictive.
376
377 16. Inserted checks for integer overflows during escape sequence (backslash)
378 processing, and also fixed erroneous offset values for syntax errors during
379 backslash processing.
380
381 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
382 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
383
384 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
385 caused an overrun.
386
387 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
388 something other than just ASCII characters) inside a group that had an
389 unlimited repeat caused a loop at compile time (while checking to see
390 whether the group could match an empty string).
391
392 20. Debugging a pattern containing \p or \P could cause a crash. For example,
393 [\P{Any}] did so. (Error in the code for printing property names.)
394
395 21. An orphan \E inside a character class could cause a crash.
396
397 22. A repeated capturing bracket such as (A)? could cause a wild memory
398 reference during compilation.
399
400 23. There are several functions in pcre_compile() that scan along a compiled
401 expression for various reasons (e.g. to see if it's fixed length for look
402 behind). There were bugs in these functions when a repeated \p or \P was
403 present in the pattern. These operators have additional parameters compared
404 with \d, etc, and these were not being taken into account when moving along
405 the compiled data. Specifically:
406
407 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
408 length.
409
410 (b) An item such as \pL+ within a repeated group could cause crashes or
411 loops.
412
413 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
414 "reference to non-existent subpattern" error.
415
416 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
417
418 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
419 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
420
421 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
422
423 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
424 character were causing crashes (broken optimization).
425
426 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
427 \p or \P) caused a compile-time loop.
428
429 28. More problems have arisen in unanchored patterns when CRLF is a valid line
430 break. For example, the unstudied pattern [\r\n]A does not match the string
431 "\r\nA" because change 7.0/46 below moves the current point on by two
432 characters after failing to match at the start. However, the pattern \nA
433 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
434 the same is true. There doesn't seem any very clean way out of this, but
435 what I have chosen to do makes the common cases work: PCRE now takes note
436 of whether there can be an explicit match for \r or \n anywhere in the
437 pattern, and if so, 7.0/46 no longer applies. As part of this change,
438 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
439 pattern has explicit CR or LF references.
440
441 29. Added (*CR) etc for changing newline setting at start of pattern.
442
443
444 Version 7.2 19-Jun-07
445 ---------------------
446
447 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
448 which is apparently normally available under Windows.
449
450 2. Re-jig the pcregrep tests with different newline settings in an attempt
451 to make them independent of the local environment's newline setting.
452
453 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
454
455 4. Some of the "internals" tests were previously cut out when the link size
456 was not 2, because the output contained actual offsets. The recent new
457 "Z" feature of pcretest means that these can be cut out, making the tests
458 usable with all link sizes.
459
460 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
461 stack recursion. This gives a massive performance boost under BSD, but just
462 a small improvement under Linux. However, it saves one field in the frame
463 in all cases.
464
465 6. Added more features from the forthcoming Perl 5.10:
466
467 (a) (?-n) (where n is a string of digits) is a relative subroutine or
468 recursion call. It refers to the nth most recently opened parentheses.
469
470 (b) (?+n) is also a relative subroutine call; it refers to the nth next
471 to be opened parentheses.
472
473 (c) Conditions that refer to capturing parentheses can be specified
474 relatively, for example, (?(-2)... or (?(+3)...
475
476 (d) \K resets the start of the current match so that everything before
477 is not part of it.
478
479 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
480
481 (f) \g{name} is another synonym - part of Perl 5.10's unification of
482 reference syntax.
483
484 (g) (?| introduces a group in which the numbering of parentheses in each
485 alternative starts with the same number.
486
487 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
488
489 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
490 PCRE_INFO_JCHANGED.
491
492 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
493 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
494 for detecting groups that can match an empty string.
495
496 9. A pattern with a very large number of alternatives (more than several
497 hundred) was running out of internal workspace during the pre-compile
498 phase, where pcre_compile() figures out how much memory will be needed. A
499 bit of new cunning has reduced the workspace needed for groups with
500 alternatives. The 1000-alternative test pattern now uses 12 bytes of
501 workspace instead of running out of the 4096 that are available.
502
503 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
504
505 11. Applied patch from Google to remove an optimization that didn't quite work.
506 The report of the bug said:
507
508 pcrecpp::RE("a*").FullMatch("aaa") matches, while
509 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
510 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
511
512 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
513 it matched the wrong number of bytes.
514
515
516 Version 7.1 24-Apr-07
517 ---------------------
518
519 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
520 that is more "standard", making use of automake and other Autotools. There
521 is some re-arrangement of the files and adjustment of comments consequent
522 on this.
523
524 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
525 for recursive directory scanning broke on some systems because the files
526 are not scanned in any specific order and on different systems the order
527 was different. A call to "sort" has been inserted into RunGrepTest for the
528 approprate test as a short-term fix. In the longer term there may be an
529 alternative.
530
531 3. I had an email from Eric Raymond about problems translating some of PCRE's
532 man pages to HTML (despite the fact that I distribute HTML pages, some
533 people do their own conversions for various reasons). The problems
534 concerned the use of low-level troff macros .br and .in. I have therefore
535 removed all such uses from the man pages (some were redundant, some could
536 be replaced by .nf/.fi pairs). The 132html script that I use to generate
537 HTML has been updated to handle .nf/.fi and to complain if it encounters
538 .br or .in.
539
540 4. Updated comments in configure.ac that get placed in config.h.in and also
541 arranged for config.h to be included in the distribution, with the name
542 config.h.generic, for the benefit of those who have to compile without
543 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
544
545 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
546 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
547 makevp.bat for the current PCRE, using the additional files
548 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
549
550 6. A Windows user reported a minor discrepancy with test 2, which turned out
551 to be caused by a trailing space on an input line that had got lost in his
552 copy. The trailing space was an accident, so I've just removed it.
553
554 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
555 that is needed.
556
557 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
558 as "const" (a) because they are and (b) because it helps the PHP
559 maintainers who have recently made a script to detect big data structures
560 in the php code that should be moved to the .rodata section. I remembered
561 to update Builducptable as well, so it won't revert if ucptable.h is ever
562 re-created.
563
564 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
565 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
566 order to be able to cut out the UTF-8 tables in the latter when UTF-8
567 support is not required. This saves 1.5-2K of code, which is important in
568 some applications.
569
570 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
571 so as not to refer to the tables, even though these functions will never be
572 called when UTF-8 support is disabled. Otherwise there are problems with a
573 shared library.
574
575 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
576
577 (a) It was defining its arguments as char * instead of void *.
578
579 (b) It was assuming that all moves were upwards in memory; this was true
580 a long time ago when I wrote it, but is no longer the case.
581
582 The emulated memove() is provided for those environments that have neither
583 memmove() nor bcopy(). I didn't think anyone used it these days, but that
584 is clearly not the case, as these two bugs were recently reported.
585
586 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
587 and Detrail to create the HTML documentation, the .txt form of the man
588 pages, and it removes trailing spaces from listed files. It also creates
589 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
590 case, it wraps all the #defines with #ifndefs. This script should be run
591 before "make dist".
592
593 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
594 with Unicode property support.
595
596 (a) For a maximizing quantifier, if the two different cases of the
597 character were of different lengths in their UTF-8 codings (there are
598 some cases like this - I found 11), and the matching function had to
599 back up over a mixture of the two cases, it incorrectly assumed they
600 were both the same length.
601
602 (b) When PCRE was configured to use the heap rather than the stack for
603 recursion during matching, it was not correctly preserving the data for
604 the other case of a UTF-8 character when checking ahead for a match
605 while processing a minimizing repeat. If the check also involved
606 matching a wide character, but failed, corruption could cause an
607 erroneous result when trying to check for a repeat of the original
608 character.
609
610 13. Some tidying changes to the testing mechanism:
611
612 (a) The RunTest script now detects the internal link size and whether there
613 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
614 values substituted by "configure". (The RunGrepTest script already did
615 this for UTF-8.) The configure.ac script no longer substitutes the
616 relevant variables.
617
618 (b) The debugging options /B and /D in pcretest show the compiled bytecode
619 with length and offset values. This means that the output is different
620 for different internal link sizes. Test 2 is skipped for link sizes
621 other than 2 because of this, bypassing the problem. Unfortunately,
622 there was also a test in test 3 (the locale tests) that used /B and
623 failed for link sizes other than 2. Rather than cut the whole test out,
624 I have added a new /Z option to pcretest that replaces the length and
625 offset values with spaces. This is now used to make test 3 independent
626 of link size. (Test 2 will be tidied up later.)
627
628 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
629 segmentation fault instead of returning the appropriate error message.
630
631 15. In multiline mode when the newline sequence was set to "any", the pattern
632 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
633 This doesn't seem right; it now treats the CRLF combination as the line
634 ending, and so does not match in that case. It's only a pattern such as ^$
635 that would hit this one: something like ^ABC$ would have failed after \r
636 and then tried again after \r\n.
637
638 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
639 in an attempt to make files that differ only in their line terminators
640 compare equal. This works on Linux.
641
642 17. Under certain error circumstances pcregrep might try to free random memory
643 as it exited. This is now fixed, thanks to valgrind.
644
645 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
646 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
647 was because its rules for how to advance for /g after matching an empty
648 string at the end of a line did not allow for this case. They now check for
649 it specially.
650
651 20. pcretest is supposed to handle patterns and data of any length, by
652 extending its buffers when necessary. It was getting this wrong when the
653 buffer for a data line had to be extended.
654
655 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
656 CRLF as a newline sequence.
657
658 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
659 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
660 I have nevertheless tidied it up.
661
662 23. Added some casts to kill warnings from HP-UX ia64 compiler.
663
664 24. Added a man page for pcre-config.
665
666
667 Version 7.0 19-Dec-06
668 ---------------------
669
670 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
671 moving to gcc 4.1.1.
672
673 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
674 sys/time.h, which is documented as needed for this function. It doesn't
675 seem to matter on Linux, but it showed up on some releases of OS X.
676
677 3. It seems that there are systems where bytes whose values are greater than
678 127 match isprint() in the "C" locale. The "C" locale should be the
679 default when a C program starts up. In most systems, only ASCII printing
680 characters match isprint(). This difference caused the output from pcretest
681 to vary, making some of the tests fail. I have changed pcretest so that:
682
683 (a) When it is outputting text in the compiled version of a pattern, bytes
684 other than 32-126 are always shown as hex escapes.
685
686 (b) When it is outputting text that is a matched part of a subject string,
687 it does the same, unless a different locale has been set for the match
688 (using the /L modifier). In this case, it uses isprint() to decide.
689
690 4. Fixed a major bug that caused incorrect computation of the amount of memory
691 required for a compiled pattern when options that changed within the
692 pattern affected the logic of the preliminary scan that determines the
693 length. The relevant options are -x, and -i in UTF-8 mode. The result was
694 that the computed length was too small. The symptoms of this bug were
695 either the PCRE error "internal error: code overflow" from pcre_compile(),
696 or a glibc crash with a message such as "pcretest: free(): invalid next
697 size (fast)". Examples of patterns that provoked this bug (shown in
698 pcretest format) are:
699
700 /(?-x: )/x
701 /(?x)(?-x: \s*#\s*)/
702 /((?i)[\x{c0}])/8
703 /(?i:[\x{c0}])/8
704
705 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
706 is now done differently.
707
708 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
709 wrapper classes; (b) implement a new function in the C++ scanner that is
710 more efficient than the old way of doing things because it avoids levels of
711 recursion in the regex matching; (c) add a paragraph to the documentation
712 for the FullMatch() function.
713
714 6. The escape sequence \n was being treated as whatever was defined as
715 "newline". Not only was this contrary to the documentation, which states
716 that \n is character 10 (hex 0A), but it also went horribly wrong when
717 "newline" was defined as CRLF. This has been fixed.
718
719 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
720 was being set to -1 for the "end of line" case (supposedly a value that no
721 character can have). Though this value is never used (the check for end of
722 line is "zero bytes in current character"), it caused compiler complaints.
723 I've changed it to 0xffffffff.
724
725 8. In pcre_version.c, the version string was being built by a sequence of
726 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
727 string (as it is for production releases) called a macro with an empty
728 argument. The C standard says the result of this is undefined. The gcc
729 compiler treats it as an empty string (which was what was wanted) but it is
730 reported that Visual C gives an error. The source has been hacked around to
731 avoid this problem.
732
733 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
734 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
735 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
736 of them did).
737
738 10. Originally, pcretest opened its input and output without "b"; then I was
739 told that "b" was needed in some environments, so it was added for release
740 5.0 to both the input and output. (It makes no difference on Unix-like
741 systems.) Later I was told that it is wrong for the input on Windows. I've
742 now abstracted the modes into two macros, to make it easier to fiddle with
743 them, and removed "b" from the input mode under Windows.
744
745 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
746
747 12. Added -help and --help to pcretest as an official way of being reminded
748 of the options.
749
750 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
751 and pcrecpp.cc because they annoy compilers at high warning levels.
752
753 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
754
755 15. Fixed an occurrence of == in configure.ac that should have been = (shell
756 scripts are not C programs :-) and which was not noticed because it works
757 on Linux.
758
759 16. pcretest is supposed to handle any length of pattern and data line (as one
760 line or as a continued sequence of lines) by extending its input buffer if
761 necessary. This feature was broken for very long pattern lines, leading to
762 a string of junk being passed to pcre_compile() if the pattern was longer
763 than about 50K.
764
765 17. I have done a major re-factoring of the way pcre_compile() computes the
766 amount of memory needed for a compiled pattern. Previously, there was code
767 that made a preliminary scan of the pattern in order to do this. That was
768 OK when PCRE was new, but as the facilities have expanded, it has become
769 harder and harder to keep it in step with the real compile phase, and there
770 have been a number of bugs (see for example, 4 above). I have now found a
771 cunning way of running the real compile function in a "fake" mode that
772 enables it to compute how much memory it would need, while actually only
773 ever using a few hundred bytes of working memory and without too many
774 tests of the mode. This should make future maintenance and development
775 easier. A side effect of this work is that the limit of 200 on the nesting
776 depth of parentheses has been removed (though this was never a serious
777 limitation, I suspect). However, there is a downside: pcre_compile() now
778 runs more slowly than before (30% or more, depending on the pattern). I
779 hope this isn't a big issue. There is no effect on runtime performance.
780
781 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
782 newline (only possible for the last line of a file) and it was a
783 pattern that set a locale (followed by /Lsomething), pcretest crashed.
784
785 19. Added additional timing features to pcretest. (1) The -tm option now times
786 matching only, not compiling. (2) Both -t and -tm can be followed, as a
787 separate command line item, by a number that specifies the number of
788 repeats to use when timing. The default is 50000; this gives better
789 precision, but takes uncomfortably long for very large patterns.
790
791 20. Extended pcre_study() to be more clever in cases where a branch of a
792 subpattern has no definite first character. For example, (a*|b*)[cd] would
793 previously give no result from pcre_study(). Now it recognizes that the
794 first character must be a, b, c, or d.
795
796 21. There was an incorrect error "recursive call could loop indefinitely" if
797 a subpattern (or the entire pattern) that was being tested for matching an
798 empty string contained only one non-empty item after a nested subpattern.
799 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
800 incorrectly, because the \d was being skipped in the check.
801
802 22. The pcretest program now has a new pattern option /B and a command line
803 option -b, which is equivalent to adding /B to every pattern. This causes
804 it to show the compiled bytecode, without the additional information that
805 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
806 is the same as /B/I).
807
808 23. A new optimization is now able automatically to treat some sequences such
809 as a*b as a*+b. More specifically, if something simple (such as a character
810 or a simple class like \d) has an unlimited quantifier, and is followed by
811 something that cannot possibly match the quantified thing, the quantifier
812 is automatically "possessified".
813
814 24. A recursive reference to a subpattern whose number was greater than 39
815 went wrong under certain circumstances in UTF-8 mode. This bug could also
816 have affected the operation of pcre_study().
817
818 25. Realized that a little bit of performance could be had by replacing
819 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
820
821 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
822
823 27. Possessive quantifiers such as a++ were previously implemented by turning
824 them into atomic groups such as ($>a+). Now they have their own opcodes,
825 which improves performance. This includes the automatically created ones
826 from 23 above.
827
828 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
829 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
830 the first matched character to be a colon. This applied both to named and
831 numbered groups.
832
833 29. The ucpinternal.h header file was missing its idempotency #ifdef.
834
835 30. I was sent a "project" file called libpcre.a.dev which I understand makes
836 building PCRE on Windows easier, so I have included it in the distribution.
837
838 31. There is now a check in pcretest against a ridiculously large number being
839 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
840 loop, the loop is abandoned.
841
842 32. Forward references to subpatterns in conditions such as (?(2)...) where
843 subpattern 2 is defined later cause pcre_compile() to search forwards in
844 the pattern for the relevant set of parentheses. This search went wrong
845 when there were unescaped parentheses in a character class, parentheses
846 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
847
848 33. "Subroutine" calls and backreferences were previously restricted to
849 referencing subpatterns earlier in the regex. This restriction has now
850 been removed.
851
852 34. Added a number of extra features that are going to be in Perl 5.10. On the
853 whole, these are just syntactic alternatives for features that PCRE had
854 previously implemented using the Python syntax or my own invention. The
855 other formats are all retained for compatibility.
856
857 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
858 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
859 also .NET compatible.
860
861 (b) A recursion or subroutine call to a named group can now be defined as
862 (?&name) as well as (?P>name).
863
864 (c) A backreference to a named group can now be defined as \k<name> or
865 \k'name' as well as (?P=name). The new forms, as well as being in Perl
866 5.10, are also .NET compatible.
867
868 (d) A conditional reference to a named group can now use the syntax
869 (?(<name>) or (?('name') as well as (?(name).
870
871 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
872 groups (named and numbered) that are never evaluated inline, but can be
873 called as "subroutines" from elsewhere. In effect, the DEFINE condition
874 is always false. There may be only one alternative in such a group.
875
876 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
877 as the simple (?(R). The condition is true only if the most recent
878 recursion is that of the given number or name. It does not search out
879 through the entire recursion stack.
880
881 (g) The escape \gN or \g{N} has been added, where N is a positive or
882 negative number, specifying an absolute or relative reference.
883
884 35. Tidied to get rid of some further signed/unsigned compiler warnings and
885 some "unreachable code" warnings.
886
887 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
888 things, this adds five new scripts.
889
890 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
891 There were also incompatibilities regarding the handling of \Q..\E inside
892 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
893 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
894
895 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
896 matches an empty string, and forcibly breaks the loop. There were bugs in
897 this code in non-simple cases. For a pattern such as ^(a()*)* matched
898 against aaaa the result was just "a" rather than "aaaa", for example. Two
899 separate and independent bugs (that affected different cases) have been
900 fixed.
901
902 39. Refactored the code to abolish the use of different opcodes for small
903 capturing bracket numbers. This is a tidy that I avoided doing when I
904 removed the limit on the number of capturing brackets for 3.5 back in 2001.
905 The new approach is not only tidier, it makes it possible to reduce the
906 memory needed to fix the previous bug (38).
907
908 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
909 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
910 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
911 mode.
912
913 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
914 report.
915
916 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
917 copy construction and assignment in the C++ wrapper.
918
919 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
920 couple of bugs that could have given wrong results in the "--newline=crlf"
921 case.
922
923 44. Added a number of casts and did some reorganization of signed/unsigned int
924 variables following suggestions from Dair Grant. Also renamed the variable
925 "this" as "item" because it is a C++ keyword.
926
927 45. Arranged for dftables to add
928
929 #include "pcre_internal.h"
930
931 to pcre_chartables.c because without it, gcc 4.x may remove the array
932 definition from the final binary if PCRE is built into a static library and
933 dead code stripping is activated.
934
935 46. For an unanchored pattern, if a match attempt fails at the start of a
936 newline sequence, and the newline setting is CRLF or ANY, and the next two
937 characters are CRLF, advance by two characters instead of one.
938
939
940 Version 6.7 04-Jul-06
941 ---------------------
942
943 1. In order to handle tests when input lines are enormously long, pcretest has
944 been re-factored so that it automatically extends its buffers when
945 necessary. The code is crude, but this _is_ just a test program. The
946 default size has been increased from 32K to 50K.
947
948 2. The code in pcre_study() was using the value of the re argument before
949 testing it for NULL. (Of course, in any sensible call of the function, it
950 won't be NULL.)
951
952 3. The memmove() emulation function in pcre_internal.h, which is used on
953 systems that lack both memmove() and bcopy() - that is, hardly ever -
954 was missing a "static" storage class specifier.
955
956 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
957 containing an extended class (one that cannot be represented by a bitmap
958 because it contains high-valued characters or Unicode property items, e.g.
959 [\pZ]). Almost always one would set UTF-8 mode when processing such a
960 pattern, but PCRE should not loop if you do not (it no longer does).
961 [Detail: two cases were found: (a) a repeated subpattern containing an
962 extended class; (b) a recursive reference to a subpattern that followed a
963 previous extended class. It wasn't skipping over the extended class
964 correctly when UTF-8 mode was not set.]
965
966 5. A negated single-character class was not being recognized as fixed-length
967 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
968 compile error "lookbehind assertion is not fixed length".
969
970 6. The RunPerlTest auxiliary script was showing an unexpected difference
971 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
972 write a Perl script that can interpret lines of an input file either as
973 byte characters or as UTF-8, which is what "perltest" was being required to
974 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
975 can't do is switch easily at run time between having the "use utf8;" pragma
976 or not. In the end, I fudged it by using the RunPerlTest script to insert
977 "use utf8;" explicitly for the UTF-8 tests.
978
979 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
980 the end of the subject string, contrary to the documentation and to what
981 Perl does. This was true of both matching functions. Now it matches only at
982 the start of the subject and immediately after *internal* newlines.
983
984 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
985 a pointer to an int instead of a pointer to an unsigned long int. This
986 caused problems on 64-bit systems.
987
988 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
989 instance of the 'standard' template library not being so standard".
990
991 10. There was no check on the number of named subpatterns nor the maximum
992 length of a subpattern name. The product of these values is used to compute
993 the size of the memory block for a compiled pattern. By supplying a very
994 long subpattern name and a large number of named subpatterns, the size
995 computation could be caused to overflow. This is now prevented by limiting
996 the length of names to 32 characters, and the number of named subpatterns
997 to 10,000.
998
999 11. Subpatterns that are repeated with specific counts have to be replicated in
1000 the compiled pattern. The size of memory for this was computed from the
1001 length of the subpattern and the repeat count. The latter is limited to
1002 65535, but there was no limit on the former, meaning that integer overflow
1003 could in principle occur. The compiled length of a repeated subpattern is
1004 now limited to 30,000 bytes in order to prevent this.
1005
1006 12. Added the optional facility to have named substrings with the same name.
1007
1008 13. Added the ability to use a named substring as a condition, using the
1009 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1010 are numbers (not recommended). Forward references are permitted.
1011
1012 14. Added forward references in named backreferences (if you see what I mean).
1013
1014 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1015 pattern could run off the end of the subject. For example, the pattern
1016 "(?s)(.{1,5})"8 did this with the subject "ab".
1017
1018 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1019 PCRE_CASELESS was set when matching characters that were quantified with ?
1020 or *.
1021
1022 17. A character class other than a single negated character that had a minimum
1023 but no maximum quantifier - for example [ab]{6,} - was not handled
1024 correctly by pce_dfa_exec(). It would match only one character.
1025
1026 18. A valid (though odd) pattern that looked like a POSIX character
1027 class but used an invalid character after [ (for example [[,abc,]]) caused
1028 pcre_compile() to give the error "Failed: internal error: code overflow" or
1029 in some cases to crash with a glibc free() error. This could even happen if
1030 the pattern terminated after [[ but there just happened to be a sequence of
1031 letters, a binary zero, and a closing ] in the memory that followed.
1032
1033 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1034 over the years. Originally (before any Unicode support), just the bottom 8
1035 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1036 output from "man perlunicode" includes this:
1037
1038 The regular expression compiler produces polymorphic opcodes. That
1039 is, the pattern adapts to the data and automatically switches to
1040 the Unicode character scheme when presented with Unicode data--or
1041 instead uses a traditional byte scheme when presented with byte
1042 data.
1043
1044 Sadly, a wide octal escape does not cause a switch, and in a string with
1045 no other multibyte characters, these octal escapes are treated as before.
1046 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1047 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1048 Unicode string.
1049
1050 I have not perpetrated such confusion in PCRE. Up till now, it took just
1051 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1052 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1053 translate to the appropriate multibyte character.
1054
1055 29. Applied some refactoring to reduce the number of warnings from Microsoft
1056 and Borland compilers. This has included removing the fudge introduced
1057 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1058 a warning about an unused variable.
1059
1060 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1061 characters since release 4.0, because Perl (from release 5.004) does not.
1062 [Or at least, is documented not to: some releases seem to be in conflict
1063 with the documentation.] However, when a pattern was studied with
1064 pcre_study() and all its branches started with \s, PCRE still included VT
1065 as a possible starting character. Of course, this did no harm; it just
1066 caused an unnecessary match attempt.
1067
1068 22. Removed a now-redundant internal flag bit that recorded the fact that case
1069 dependency changed within the pattern. This was once needed for "required
1070 byte" processing, but is no longer used. This recovers a now-scarce options
1071 bit. Also moved the least significant internal flag bit to the most-
1072 significant bit of the word, which was not previously used (hangover from
1073 the days when it was an int rather than a uint) to free up another bit for
1074 the future.
1075
1076 23. Added support for CRLF line endings as well as CR and LF. As well as the
1077 default being selectable at build time, it can now be changed at runtime
1078 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1079 specify that it is scanning data with non-default line endings.
1080
1081 24. Changed the definition of CXXLINK to make it agree with the definition of
1082 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1083
1084 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1085 recursions. This makes a big different to stack usage for some patterns.
1086
1087 26. If a subpattern containing a named recursion or subroutine reference such
1088 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1089 the space required for the compiled pattern went wrong and gave too small a
1090 value. Depending on the environment, this could lead to "Failed: internal
1091 error: code overflow at offset 49" or "glibc detected double free or
1092 corruption" errors.
1093
1094 27. Applied patches from Google (a) to support the new newline modes and (b) to
1095 advance over multibyte UTF-8 characters in GlobalReplace.
1096
1097 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1098 difference for some implementation of PCRE in some Windows version.
1099
1100 29. Added some extra testing facilities to pcretest:
1101
1102 \q<number> in a data line sets the "match limit" value
1103 \Q<number> in a data line sets the "match recursion limt" value
1104 -S <number> sets the stack size, where <number> is in megabytes
1105
1106 The -S option isn't available for Windows.
1107
1108
1109 Version 6.6 06-Feb-06
1110 ---------------------
1111
1112 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1113 in pcreposix.h. I have copied the definition from pcre.h.
1114
1115 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1116 because pcre.h is no longer a built file.
1117
1118 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1119 not normally included in the compiled code.
1120
1121
1122 Version 6.5 01-Feb-06
1123 ---------------------
1124
1125 1. When using the partial match feature with pcre_dfa_exec(), it was not
1126 anchoring the second and subsequent partial matches at the new starting
1127 point. This could lead to incorrect results. For example, with the pattern
1128 /1234/, partially matching against "123" and then "a4" gave a match.
1129
1130 2. Changes to pcregrep:
1131
1132 (a) All non-match returns from pcre_exec() were being treated as failures
1133 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1134 error message is output. Some extra information is given for the
1135 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1136 probably the only errors that are likely to be caused by users (by
1137 specifying a regex that has nested indefinite repeats, for instance).
1138 If there are more than 20 of these errors, pcregrep is abandoned.
1139
1140 (b) A binary zero was treated as data while matching, but terminated the
1141 output line if it was written out. This has been fixed: binary zeroes
1142 are now no different to any other data bytes.
1143
1144 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1145 used to set a locale for matching. The --locale=xxxx long option has
1146 been added (no short equivalent) to specify a locale explicitly on the
1147 pcregrep command, overriding the environment variables.
1148
1149 (d) When -B was used with -n, some line numbers in the output were one less
1150 than they should have been.
1151
1152 (e) Added the -o (--only-matching) option.
1153
1154 (f) If -A or -C was used with -c (count only), some lines of context were
1155 accidentally printed for the final match.
1156
1157 (g) Added the -H (--with-filename) option.
1158
1159 (h) The combination of options -rh failed to suppress file names for files
1160 that were found from directory arguments.
1161
1162 (i) Added the -D (--devices) and -d (--directories) options.
1163
1164 (j) Added the -F (--fixed-strings) option.
1165
1166 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1167
1168 (l) Added the --colo(u)r option.
1169
1170 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1171 is not present by default.
1172
1173 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1174 items such as (?R) or (?1), when the recursion could match a number of
1175 alternatives. If it matched one of the alternatives, but subsequently,
1176 outside the recursion, there was a failure, the code tried to back up into
1177 the recursion. However, because of the way PCRE is implemented, this is not
1178 possible, and the result was an incorrect result from the match.
1179
1180 In order to prevent this happening, the specification of recursion has
1181 been changed so that all such subpatterns are automatically treated as
1182 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1183
1184 4. I had overlooked the fact that, in some locales, there are characters for
1185 which isalpha() is true but neither isupper() nor islower() are true. In
1186 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1187 and ordfeminine) are like this. This affected the treatment of \w and \W
1188 when they appeared in character classes, but not when they appeared outside
1189 a character class. The bit map for "word" characters is now created
1190 separately from the results of isalnum() instead of just taking it from the
1191 upper, lower, and digit maps. (Plus the underscore character, of course.)
1192
1193 5. The above bug also affected the handling of POSIX character classes such as
1194 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1195 permanent tables. Instead, the bit maps for such a class were previously
1196 created as the appropriate unions of the upper, lower, and digit bitmaps.
1197 Now they are created by subtraction from the [[:word:]] class, which has
1198 its own bitmap.
1199
1200 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1201 It is created by subtracting the vertical space characters (\x09, \x0a,
1202 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1203 subtraction was done in the overall bitmap for a character class, meaning
1204 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1205 be recognized. This bug has been fixed.
1206
1207 7. Patches from the folks at Google:
1208
1209 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1210 real life, but is still worth protecting against".
1211
1212 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1213 regular expressions".
1214
1215 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1216 have it.
1217
1218 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1219 "configure" and the latter not, in order to fix a problem somebody had
1220 with compiling the Arg class on HP-UX.
1221
1222 (e) Improve the error-handling of the C++ wrapper a little bit.
1223
1224 (f) New tests for checking recursion limiting.
1225
1226 8. The pcre_memmove() function, which is used only if the environment does not
1227 have a standard memmove() function (and is therefore rarely compiled),
1228 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1229 returning a result (though PCRE never actually uses the result).
1230
1231 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1232 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1233 returned instead of calling malloc() with an overflowing number that would
1234 most likely cause subsequent chaos.
1235
1236 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1237
1238 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1239 with this option is matched, the nmatch and pmatch options of regexec() are
1240 ignored.
1241
1242 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1243 provided in case anyone wants to the the POSIX interface with UTF-8
1244 strings.
1245
1246 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1247 C++ linking (needed for some HP-UX environments).
1248
1249 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1250 (unused parameter) and in the pcre_printint() function (omitted "default"
1251 switch label when the default is to do nothing).
1252
1253 15. Added some code to make it possible, when PCRE is compiled as a C++
1254 library, to replace subject pointers for pcre_exec() with a smart pointer
1255 class, thus making it possible to process discontinuous strings.
1256
1257 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1258 much the same function. They were added by different people who were trying
1259 to make PCRE easy to compile on non-Unix systems. It has been suggested
1260 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1261 for compiling on Windows systems. I have therefore replaced it with
1262 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1263 defaults to "extern" for C or "extern C" for C++, which works fine on
1264 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1265 SCOPE with something explicit in config.h. In addition:
1266
1267 (a) pcreposix.h still had just "extern" instead of either of these macros;
1268 I have replaced it with PCRE_DATA_SCOPE.
1269
1270 (b) Functions such as _pcre_xclass(), which are internal to the library,
1271 but external in the C sense, all had PCRE_EXPORT in their definitions.
1272 This is apparently wrong for the Windows case, so I have removed it.
1273 (It makes no difference on Unix-like systems.)
1274
1275 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1276 of recursive calls to match(). This is different to MATCH_LIMIT because
1277 that limits the total number of calls to match(), not all of which increase
1278 the depth of recursion. Limiting the recursion depth limits the amount of
1279 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1280 when PCRE is compiled, and changed at run time. A patch from Google adds
1281 this functionality to the C++ interface.
1282
1283 18. Changes to the handling of Unicode character properties:
1284
1285 (a) Updated the table to Unicode 4.1.0.
1286
1287 (b) Recognize characters that are not in the table as "Cn" (undefined).
1288
1289 (c) I revised the way the table is implemented to a much improved format
1290 which includes recognition of ranges. It now supports the ranges that
1291 are defined in UnicodeData.txt, and it also amalgamates other
1292 characters into ranges. This has reduced the number of entries in the
1293 table from around 16,000 to around 3,000, thus reducing its size
1294 considerably. I realized I did not need to use a tree structure after
1295 all - a binary chop search is just as efficient. Having reduced the
1296 number of entries, I extended their size from 6 bytes to 8 bytes to
1297 allow for more data.
1298
1299 (d) Added support for Unicode script names via properties such as \p{Han}.
1300
1301 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1302 matching that character.
1303
1304 20. When matching a repeated Unicode property with a minimum greater than zero,
1305 (for example \pL{2,}), PCRE could look past the end of the subject if it
1306 reached it while seeking the minimum number of characters. This could
1307 happen only if some of the characters were more than one byte long, because
1308 there is a check for at least the minimum number of bytes.
1309
1310 21. Refactored the implementation of \p and \P so as to be more general, to
1311 allow for more different types of property in future. This has changed the
1312 compiled form incompatibly. Anybody with saved compiled patterns that use
1313 \p or \P will have to recompile them.
1314
1315 22. Added "Any" and "L&" to the supported property types.
1316
1317 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1318 but give a compile time error if the value is greater than 0xff.
1319
1320 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1321 accidentally not being installed or uninstalled.
1322
1323 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1324 made were to insert the current release number. This seemed silly, because
1325 it made things harder for people building PCRE on systems that don't run
1326 "configure". I have turned pcre.h into a distributed file, no longer built
1327 by "configure", with the version identification directly included. There is
1328 no longer a pcre.h.in file.
1329
1330 However, this change necessitated a change to the pcre-config script as
1331 well. It is built from pcre-config.in, and one of the substitutions was the
1332 release number. I have updated configure.ac so that ./configure now finds
1333 the release number by grepping pcre.h.
1334
1335 26. Added the ability to run the tests under valgrind.
1336
1337
1338 Version 6.4 05-Sep-05
1339 ---------------------
1340
1341 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1342 "--" to be printed when multiple files were scanned, even when none of the
1343 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1344 consider it to be a bug, and have restored the previous behaviour.
1345
1346 2. A couple of code tidies to get rid of compiler warnings.
1347
1348 3. The pcretest program used to cheat by referring to symbols in the library
1349 whose names begin with _pcre_. These are internal symbols that are not
1350 really supposed to be visible externally, and in some environments it is
1351 possible to suppress them. The cheating is now confined to including
1352 certain files from the library's source, which is a bit cleaner.
1353
1354 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1355 file's purpose clearer.
1356
1357 5. Reorganized pcre_ucp_findchar().
1358
1359
1360 Version 6.3 15-Aug-05
1361 ---------------------
1362
1363 1. The file libpcre.pc.in did not have general read permission in the tarball.
1364
1365 2. There were some problems when building without C++ support:
1366
1367 (a) If C++ support was not built, "make install" and "make test" still
1368 tried to test it.
1369
1370 (b) There were problems when the value of CXX was explicitly set. Some
1371 changes have been made to try to fix these, and ...
1372
1373 (c) --disable-cpp can now be used to explicitly disable C++ support.
1374
1375 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1376 backslash in a target when C++ was disabled. This confuses some
1377 versions of "make", apparently. Using an intermediate variable solves
1378 this. (Same for CPP_LOBJ.)
1379
1380 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1381 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1382 necessary on certain architectures.
1383
1384 4. Added a setting of -export-symbols-regex to the link command to remove
1385 those symbols that are exported in the C sense, but actually are local
1386 within the library, and not documented. Their names all begin with
1387 "_pcre_". This is not a perfect job, because (a) we have to except some
1388 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1389 available (and never for static libraries). I have made a note to try to
1390 find a way round (a) in the future.
1391
1392
1393 Version 6.2 01-Aug-05
1394 ---------------------
1395
1396 1. There was no test for integer overflow of quantifier values. A construction
1397 such as {1111111111111111} would give undefined results. What is worse, if
1398 a minimum quantifier for a parenthesized subpattern overflowed and became
1399 negative, the calculation of the memory size went wrong. This could have
1400 led to memory overwriting.
1401
1402 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1403
1404 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1405 operating environments where this matters.
1406
1407 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1408 PCRE options from within the C++ wrapper.
1409
1410 5. Named capturing subpatterns were not being correctly counted when a pattern
1411 was compiled. This caused two problems: (a) If there were more than 100
1412 such subpatterns, the calculation of the memory needed for the whole
1413 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1414 back references of the form \12, where the number was greater than 9, were
1415 not recognized as back references, even though there were sufficient
1416 previous subpatterns.
1417
1418 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1419 versions of gcc, e.g. 2.95.4.
1420
1421
1422 Version 6.1 21-Jun-05
1423 ---------------------
1424
1425 1. There was one reference to the variable "posix" in pcretest.c that was not
1426 surrounded by "#if !defined NOPOSIX".
1427
1428 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1429 the cross-check on the old pcre_info() function, for the benefit of the
1430 cut-down version of PCRE that is currently imported into Exim.
1431
1432 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1433 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1434 patterns that start (?i) but I don't think that matters. The use of (?i) is
1435 just an example; this all applies to the other options as well.
1436
1437 4. Since libtool seems to echo the compile commands it is issuing, the output
1438 from "make" can be reduced a bit by putting "@" in front of each libtool
1439 compile command.
1440
1441 5. Patch from the folks at Google for configure.in to be a bit more thorough
1442 in checking for a suitable C++ installation before trying to compile the
1443 C++ stuff. This should fix a reported problem when a compiler was present,
1444 but no suitable headers.
1445
1446 6. The man pages all had just "PCRE" as their title. I have changed them to
1447 be the relevant file name. I have also arranged that these names are
1448 retained in the file doc/pcre.txt, which is a concatenation in text format
1449 of all the man pages except the little individual ones for each function.
1450
1451 7. The NON-UNIX-USE file had not been updated for the different set of source
1452 files that come with release 6. I also added a few comments about the C++
1453 wrapper.
1454
1455
1456 Version 6.0 07-Jun-05
1457 ---------------------
1458
1459 1. Some minor internal re-organization to help with my DFA experiments.
1460
1461 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1462 didn't matter for the library itself when fully configured, but did matter
1463 when compiling without UCP support, or within Exim, where the ucp files are
1464 not imported.
1465
1466 3. Refactoring of the library code to split up the various functions into
1467 different source modules. The addition of the new DFA matching code (see
1468 below) to a single monolithic source would have made it really too
1469 unwieldy, quite apart from causing all the code to be include in a
1470 statically linked application, when only some functions are used. This is
1471 relevant even without the DFA addition now that patterns can be compiled in
1472 one application and matched in another.
1473
1474 The downside of splitting up is that there have to be some external
1475 functions and data tables that are used internally in different modules of
1476 the library but which are not part of the API. These have all had their
1477 names changed to start with "_pcre_" so that they are unlikely to clash
1478 with other external names.
1479
1480 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1481 a different (DFA) algorithm. Although it is slower than the original
1482 function, it does have some advantages for certain types of matching
1483 problem.
1484
1485 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1486 including restarting after a partial match.
1487
1488 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1489 defined when compiling for Windows was sent to me. I have put it into the
1490 code, though I have no means of testing or verifying it.
1491
1492 7. Added the pcre_refcount() auxiliary function.
1493
1494 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1495 match before or at the first newline in the subject string. In pcretest,
1496 the /f option on a pattern can be used to set this.
1497
1498 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1499 would behave wrongly. This has been present in PCRE since release 4.0.
1500
1501 10. A number of changes to the pcregrep command:
1502
1503 (a) Refactored how -x works; insert ^(...)$ instead of setting
1504 PCRE_ANCHORED and checking the length, in preparation for adding
1505 something similar for -w.
1506
1507 (b) Added the -w (match as a word) option.
1508
1509 (c) Refactored the way lines are read and buffered so as to have more
1510 than one at a time available.
1511
1512 (d) Implemented a pcregrep test script.
1513
1514 (e) Added the -M (multiline match) option. This allows patterns to match
1515 over several lines of the subject. The buffering ensures that at least
1516 8K, or the rest of the document (whichever is the shorter) is available
1517 for matching (and similarly the previous 8K for lookbehind assertions).
1518
1519 (f) Changed the --help output so that it now says
1520
1521 -w, --word-regex(p)
1522
1523 instead of two lines, one with "regex" and the other with "regexp"
1524 because that confused at least one person since the short forms are the
1525 same. (This required a bit of code, as the output is generated
1526 automatically from a table. It wasn't just a text change.)
1527
1528 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1529 option but starts with a hyphen. Could be a pattern or a path name
1530 starting with a hyphen, for instance.
1531
1532 (h) "-" can be given as a file name to represent stdin.
1533
1534 (i) When file names are being printed, "(standard input)" is used for
1535 the standard input, for compatibility with GNU grep. Previously
1536 "<stdin>" was used.
1537
1538 (j) The option --label=xxx can be used to supply a name to be used for
1539 stdin when file names are being printed. There is no short form.
1540
1541 (k) Re-factored the options decoding logic because we are going to add
1542 two more options that take data. Such options can now be given in four
1543 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1544
1545 (l) Added the -A, -B, and -C options for requesting that lines of context
1546 around matches be printed.
1547
1548 (m) Added the -L option to print the names of files that do not contain
1549 any matching lines, that is, the complement of -l.
1550
1551 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1552 continue to scan other files.
1553
1554 (o) The -s option was incorrectly implemented. For compatibility with other
1555 greps, it now suppresses the error message for a non-existent or non-
1556 accessible file (but not the return code). There is a new option called
1557 -q that suppresses the output of matching lines, which was what -s was
1558 previously doing.
1559
1560 (p) Added --include and --exclude options to specify files for inclusion
1561 and exclusion when recursing.
1562
1563 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1564 Hopefully, it now does.
1565
1566 12. Missing cast in pcre_study().
1567
1568 13. Added an "uninstall" target to the makefile.
1569
1570 14. Replaced "extern" in the function prototypes in Makefile.in with
1571 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1572 world, but is set differently for Windows.
1573
1574 15. Added a second compiling function called pcre_compile2(). The only
1575 difference is that it has an extra argument, which is a pointer to an
1576 integer error code. When there is a compile-time failure, this is set
1577 non-zero, in addition to the error test pointer being set to point to an
1578 error message. The new argument may be NULL if no error number is required
1579 (but then you may as well call pcre_compile(), which is now just a
1580 wrapper). This facility is provided because some applications need a
1581 numeric error indication, but it has also enabled me to tidy up the way
1582 compile-time errors are handled in the POSIX wrapper.
1583
1584 16. Added VPATH=.libs to the makefile; this should help when building with one
1585 prefix path and installing with another. (Or so I'm told by someone who
1586 knows more about this stuff than I do.)
1587
1588 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1589 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1590 match everything, including newlines. This is not POSIX-compatible, but
1591 somebody wanted the feature. From pcretest it can be activated by using
1592 both the P and the s flags.
1593
1594 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1595
1596 19. libpcre.pc was being incorrectly installed as executable.
1597
1598 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1599 it now also looks for '\r' so that it will work unmodified on Windows.
1600
1601 21. Added Google's contributed C++ wrapper to the distribution.
1602
1603 22. Added some untidy missing memory free() calls in pcretest, to keep
1604 Electric Fence happy when testing.
1605
1606
1607
1608 Version 5.0 13-Sep-04
1609 ---------------------
1610
1611 1. Internal change: literal characters are no longer packed up into items
1612 containing multiple characters in a single byte-string. Each character
1613 is now matched using a separate opcode. However, there may be more than one
1614 byte in the character in UTF-8 mode.
1615
1616 2. The pcre_callout_block structure has two new fields: pattern_position and
1617 next_item_length. These contain the offset in the pattern to the next match
1618 item, and its length, respectively.
1619
1620 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1621 insertion of callouts before each pattern item. Added the /C option to
1622 pcretest to make use of this.
1623
1624 4. On the advice of a Windows user, the lines
1625
1626 #if defined(_WIN32) || defined(WIN32)
1627 _setmode( _fileno( stdout ), 0x8000 );
1628 #endif /* defined(_WIN32) || defined(WIN32) */
1629
1630 have been added to the source of pcretest. This apparently does useful
1631 magic in relation to line terminators.
1632
1633 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1634 for the benefit of those environments where the "b" makes a difference.
1635
1636 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1637 to know about it. I have put a hack into configure.in that adds in code
1638 to set GCC=yes if CC=icc. This seems to end up at a point in the
1639 generated configure script that is early enough to affect the setting of
1640 compiler options, which is what is needed, but I have no means of testing
1641 whether it really works. (The user who reported this had patched the
1642 generated configure script, which of course I cannot do.)
1643
1644 LATER: After change 22 below (new libtool files), the configure script
1645 seems to know about icc (and also ecc). Therefore, I have commented out
1646 this hack in configure.in.
1647
1648 7. Added support for pkg-config (2 patches were sent in).
1649
1650 8. Negated POSIX character classes that used a combination of internal tables
1651 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1652 [[:^ascii]]. Typically, they would match almost any characters. The other
1653 POSIX classes were not broken in this way.
1654
1655 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1656 to find the match, as PCRE was deluded into thinking that the match had to
1657 start at the start point or following a newline. The same bug applied to
1658 patterns with negative forward assertions or any backward assertions
1659 preceding ".*" at the start, unless the pattern required a fixed first
1660 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1661
1662 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1663 starting at the last subject character, bytes beyond the end of the subject
1664 string were read.
1665
1666 11. Renamed the variable "class" as "classbits" to make life easier for C++
1667 users. (Previously there was a macro definition, but it apparently wasn't
1668 enough.)
1669
1670 12. Added the new field "tables" to the extra data so that tables can be passed
1671 in at exec time, or the internal tables can be re-selected. This allows
1672 a compiled regex to be saved and re-used at a later time by a different
1673 program that might have everything at different addresses.
1674
1675 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1676 -R library as well as a -L library.
1677
1678 14. The debugging options of pcretest (-d on the command line or D on a
1679 pattern) showed incorrect output for anything following an extended class
1680 that contained multibyte characters and which was followed by a quantifier.
1681
1682 15. Added optional support for general category Unicode character properties
1683 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1684 support. It adds about 90K to the size of the library. The meanings of the
1685 inbuilt class escapes such as \d and \s have NOT been changed.
1686
1687 16. Updated pcredemo.c to include calls to free() to release the memory for the
1688 compiled pattern.
1689
1690 17. The generated file chartables.c was being created in the source directory
1691 instead of in the building directory. This caused the build to fail if the
1692 source directory was different from the building directory, and was
1693 read-only.
1694
1695 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1696 file. No doubt somebody will tell me if they don't make sense... Also added
1697 Dan Mooney's comments about building on OpenVMS.
1698
1699 19. Added support for partial matching via the PCRE_PARTIAL option for
1700 pcre_exec() and the \P data escape in pcretest.
1701
1702 20. Extended pcretest with 3 new pattern features:
1703
1704 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1705 write the compiled pattern to the file whose name is "rest-of-line".
1706 This is a straight binary dump of the data, with the saved pointer to
1707 the character tables forced to be NULL. The study data, if any, is
1708 written too. After writing, pcretest reads a new pattern.
1709
1710 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1711 compiled pattern from the given file. There must not be any
1712 occurrences of "<" in the file name (pretty unlikely); if there are,
1713 pcretest will instead treat the initial "<" as a pattern delimiter.
1714 After reading in the pattern, pcretest goes on to read data lines as
1715 usual.
1716
1717 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1718 and 16-bit fields in a compiled pattern, to simulate a pattern that
1719 was compiled on a host of opposite endianness.
1720
1721 21. The pcre-exec() function can now cope with patterns that were compiled on
1722 hosts of opposite endianness, with this restriction:
1723
1724 As for any compiled expression that is saved and used later, the tables
1725 pointer field cannot be preserved; the extra_data field in the arguments
1726 to pcre_exec() should be used to pass in a tables address if a value
1727 other than the default internal tables were used at compile time.
1728
1729 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1730 now diagnosed as an error. Previously, most of the time, a negative number
1731 would have been treated as zero, but if in addition "ovector" was passed as
1732 NULL, a crash could occur.
1733
1734 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1735 new versions from the libtool 1.5 distribution (the last one is a copy of
1736 a file called libtool.m4). This seems to have fixed the need to patch
1737 "configure" to support Darwin 1.3 (which I used to do). However, I still
1738 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1739 workstation).
1740
1741 24. Changed the PCRE licence to be the more standard "BSD" licence.
1742
1743
1744 Version 4.5 01-Dec-03
1745 ---------------------
1746
1747 1. There has been some re-arrangement of the code for the match() function so
1748 that it can be compiled in a version that does not call itself recursively.
1749 Instead, it keeps those local variables that need separate instances for
1750 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1751 needs to "recurse". Keeping track of where control must go is done by means
1752 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1753 hide most of the details from the main code, and operates only if
1754 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1755 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1756 operating.
1757
1758 To make it easier for callers to provide specially tailored get/free
1759 functions for this usage, two new functions, pcre_stack_malloc, and
1760 pcre_stack_free, are used. They are always called in strict stacking order,
1761 and the size of block requested is always the same.
1762
1763 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1764 PCRE has been compiled to use the stack or the heap for recursion. The
1765 -C option of pcretest uses this to show which version is compiled.
1766
1767 A new data escape \S, is added to pcretest; it causes the amounts of store
1768 obtained and freed by both kinds of malloc/free at match time to be added
1769 to the output.
1770
1771 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1772 what's available on my current Linux desktop machine.
1773
1774 3. When matching a UTF-8 string, the test for a valid string at the start has
1775 been extended. If start_offset is not zero, PCRE now checks that it points
1776 to a byte that is the start of a UTF-8 character. If not, it returns
1777 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1778 this is necessary because there may be backward assertions in the pattern.
1779 When matching the same subject several times, it may save resources to use
1780 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1781
1782 4. The code for checking the validity of UTF-8 strings has been tightened so
1783 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1784 containing "overlong sequences".
1785
1786 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1787 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1788 should have been "&", but it just so happened that all the cases this let
1789 through by mistake were picked up later in the function.
1790
1791 6. I had used a variable called "isblank" - this is a C99 function, causing
1792 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1793
1794 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1795 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1796 so the version is shown; (c) stop "make test" echoing "./RunTest".
1797
1798 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1799
1800 9. The prototype for memmove() for systems that don't have it was using
1801 size_t, but the inclusion of the header that defines size_t was later. I've
1802 moved the #includes for the C headers earlier to avoid this.
1803
1804 10. Added some adjustments to the code to make it easier to compiler on certain
1805 special systems:
1806
1807 (a) Some "const" qualifiers were missing.
1808 (b) Added the macro EXPORT before all exported functions; by default this
1809 is defined to be empty.
1810 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1811 that it reads its output file name as an argument instead of writing
1812 to the standard output and assuming this can be redirected.
1813
1814 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1815 class containing characters with values greater than 255, PCRE compilation
1816 went into a loop.
1817
1818 12. A recursive reference to a subpattern that was within another subpattern
1819 that had a minimum quantifier of zero caused PCRE to crash. For example,
1820 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1821 recursion. If the recursively-called subpattern itself had a zero repeat,
1822 that was OK.
1823
1824 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1825 buffer into which it was copied (for escape processing) was still set at
1826 1024, so long lines caused crashes.
1827
1828 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1829 "internal error: code overflow...". This applied to any character class
1830 that was followed by a possessive quantifier.
1831
1832 15. Modified the Makefile to add libpcre.la as a prerequisite for
1833 libpcreposix.la because I was told this is needed for a parallel build to
1834 work.
1835
1836 16. If a pattern that contained .* following optional items at the start was
1837 studied, the wrong optimizing data was generated, leading to matching
1838 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1839 matching string must start with a or b or c. The correct conclusion for
1840 this pattern is that a match can start with any character.
1841
1842
1843 Version 4.4 13-Aug-03
1844 ---------------------
1845
1846 1. In UTF-8 mode, a character class containing characters with values between
1847 127 and 255 was not handled correctly if the compiled pattern was studied.
1848 In fixing this, I have also improved the studying algorithm for such
1849 classes (slightly).
1850
1851 2. Three internal functions had redundant arguments passed to them. Removal
1852 might give a very teeny performance improvement.
1853
1854 3. Documentation bug: the value of the capture_top field in a callout is *one
1855 more than* the number of the hightest numbered captured substring.
1856
1857 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1858 in incorrectly linking with a previously installed version. They now link
1859 explicitly with libpcre.la.
1860
1861 5. configure.in no longer needs to recognize Cygwin specially.
1862
1863 6. A problem in pcre.in for Windows platforms is fixed.
1864
1865 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1866 pcretest, it used to include the size of the study block as part of its
1867 output. Unfortunately, the structure contains a field that has a different
1868 size on different hardware architectures. This meant that the tests that
1869 showed this size failed. As the block is currently always of a fixed size,
1870 this information isn't actually particularly useful in pcretest output, so
1871 I have just removed it.
1872
1873 8. Three pre-processor statements accidentally did not start in column 1.
1874 Sadly, there are *still* compilers around that complain, even though
1875 standard C has not required this for well over a decade. Sigh.
1876
1877 9. In pcretest, the code for checking callouts passed small integers in the
1878 callout_data field, which is a void * field. However, some picky compilers
1879 complained about the casts involved for this on 64-bit systems. Now
1880 pcretest passes the address of the small integer instead, which should get
1881 rid of the warnings.
1882
1883 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1884 both compile and run time, and gives an error if an invalid UTF-8 sequence
1885 is found. There is a option for disabling this check in cases where the
1886 string is known to be correct and/or the maximum performance is wanted.
1887
1888 11. In response to a bug report, I changed one line in Makefile.in from
1889
1890 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1891 to
1892 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1893
1894 to look similar to other lines, but I have no way of telling whether this
1895 is the right thing to do, as I do not use Windows. No doubt I'll get told
1896 if it's wrong...
1897
1898
1899 Version 4.3 21-May-03
1900 ---------------------
1901
1902 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1903 Makefile.
1904
1905 2. Some refactoring to improve the quality of the code:
1906
1907 (i) The utf8_table... variables are now declared "const".
1908
1909 (ii) The code for \cx, which used the "case flipping" table to upper case
1910 lower case letters, now just substracts 32. This is ASCII-specific,
1911 but the whole concept of \cx is ASCII-specific, so it seems
1912 reasonable.
1913
1914 (iii) PCRE was using its character types table to recognize decimal and
1915 hexadecimal digits in the pattern. This is silly, because it handles
1916 only 0-9, a-f, and A-F, but the character types table is locale-
1917 specific, which means strange things might happen. A private
1918 table is now used for this - though it costs 256 bytes, a table is
1919 much faster than multiple explicit tests. Of course, the standard
1920 character types table is still used for matching digits in subject
1921 strings against \d.
1922
1923 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1924 ending in _t are). So I've renamed it as ESC_tee.
1925
1926 3. The first argument for regexec() in the POSIX wrapper should have been
1927 defined as "const".
1928
1929 4. Changed pcretest to use malloc() for its buffers so that they can be
1930 Electric Fenced for debugging.
1931
1932 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1933 to read one or more bytes before the start of the subject string. Often this
1934 had no effect on PCRE's behaviour, but in some circumstances it could
1935 provoke a segmentation fault.
1936
1937 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1938 to try to read one or more bytes before the start of the subject string.
1939
1940 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1941 UTF-8 support could misbehave in various ways if the subject string
1942 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1943 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1944 back over UTF-8 characters.)
1945
1946
1947 Version 4.2 14-Apr-03
1948 ---------------------
1949
1950 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1951
1952 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1953 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1954 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1955 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1956 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1957 and BUILD_EXEEXT
1958 Note: automatic setting of the BUILD variables is not yet working
1959 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1960 compile-time but not at link-time
1961 [LINK]: use for linking executables only
1962 make different versions for Windows and non-Windows
1963 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1964 libraries
1965 [LINK_FOR_BUILD]: new variable
1966 [OBJEXT]: use throughout
1967 [EXEEXT]: use throughout
1968 <winshared>: new target
1969 <wininstall>: new target
1970 <dftables.o>: use native compiler
1971 <dftables>: use native linker
1972 <install>: handle Windows platform correctly
1973 <clean>: ditto
1974 <check>: ditto
1975 copy DLL to top builddir before testing
1976
1977 As part of these changes, -no-undefined was removed again. This was reported
1978 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1979 in any case.
1980
1981 3. Some tidies to get rid of compiler warnings:
1982
1983 . In the match_data structure, match_limit was an unsigned long int, whereas
1984 match_call_count was an int. I've made them both unsigned long ints.
1985
1986 . In pcretest the fact that a const uschar * doesn't automatically cast to
1987 a void * provoked a warning.
1988
1989 . Turning on some more compiler warnings threw up some "shadow" variables
1990 and a few more missing casts.
1991
1992 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1993 option, a class that contained a single character with a value between 128
1994 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1995
1996 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1997 option, a class that contained several characters, but with at least one
1998 whose value was between 128 and 255 caused PCRE to crash.
1999
2000
2001 Version 4.1 12-Mar-03
2002 ---------------------
2003
2004 1. Compiling with gcc -pedantic found a couple of places where casts were
2005 needed, and a string in dftables.c that was longer than standard compilers are
2006 required to support.
2007
2008 2. Compiling with Sun's compiler found a few more places where the code could
2009 be tidied up in order to avoid warnings.
2010
2011 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2012 first of these names is deprecated in the latest Autoconf in favour of the name
2013 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2014 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2015 analogy I have changed it to CFLAGS_FOR_BUILD.
2016
2017 4. Added -no-undefined to the linking command in the Makefile, because this is
2018 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2019 linking step for the pcreposix library.
2020
2021 5. PCRE was failing to diagnose the case of two named groups with the same
2022 name.
2023
2024 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2025 literal character that is needed in the subject for a match, and scans along to
2026 ensure that it is present before embarking on the full matching process. This
2027 saves time in cases of nested unlimited repeats that are never going to match.
2028 Problem: the scan can take a lot of time if the subject is very long (e.g.
2029 megabytes), thus penalizing straightforward matches. It is now done only if the
2030 amount of subject to be scanned is less than 1000 bytes.
2031
2032 7. A lesser problem with the same optimization is that it was recording the
2033 first character of an anchored pattern as "needed", thus provoking a search
2034 right along the subject, even when the first match of the pattern was going to
2035 fail. The "needed" character is now not set for anchored patterns, unless it
2036 follows something in the pattern that is of non-fixed length. Thus, it still
2037 fulfils its original purpose of finding quick non-matches in cases of nested
2038 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2039
2040
2041 Version 4.0 17-Feb-03
2042 ---------------------
2043
2044 1. If a comment in an extended regex that started immediately after a meta-item
2045 extended to the end of string, PCRE compiled incorrect data. This could lead to
2046 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2047
2048 2. Moved to autoconf 2.53 and libtool 1.4.2.
2049
2050 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2051 the special perltest8 script is no longer needed - all the tests can be run
2052 from a single perltest script.
2053
2054 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2055 by \s. It has now been removed in PCRE. This means it isn't recognized as
2056 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2057 class [:space:] *does* include VT, thereby creating a mess.
2058
2059 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2060 space and tab.
2061
2062 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2063 its new features into the main test script, reducing the number of scripts.
2064
2065 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2066 were backward compatible, and made the (?i) apply to the whole pattern, as if
2067 /i were given. Now it behaves more logically, and applies the option setting
2068 only to what follows. PCRE has been changed to follow suit. However, if it
2069 finds options settings right at the start of the pattern, it extracts them into
2070 the global options, as before. Thus, they show up in the info data.
2071
2072 8. Added support for the \Q...\E escape sequence. Characters in between are
2073 treated as literals. This is slightly different from Perl in that $ and @ are
2074 also handled as literals inside the quotes. In Perl, they will cause variable
2075 interpolation. Note the following examples:
2076
2077 Pattern PCRE matches Perl matches
2078
2079 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2080 \Qabc\$xyz\E abc\$xyz abc\$xyz
2081 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2082
2083 For compatibility with Perl, \Q...\E sequences are recognized inside character
2084 classes as well as outside them.
2085
2086 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2087 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2088 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2089 signed/unsigned warnings.
2090
2091 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2092 option for pcretest, so I've replaced it by a simple function that does just
2093 that job.
2094
2095 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2096 "pcregrep -".
2097
2098 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2099 Java package. This provides some syntactic sugar for simple cases of what my
2100 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2101 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2102 item, you can use this simplified notation. Note that only makes sense with
2103 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2104 greediness, whatever the setting of the PCRE_UNGREEDY option.
2105
2106 13. A change of greediness default within a pattern was not taking effect at
2107 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2108 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2109 was abstracted outside.
2110
2111 14. PCRE now supports the \G assertion. It is true when the current matching
2112 position is at the start point of the match. This differs from \A when the
2113 starting offset is non-zero. Used with the /g option of pcretest (or similar
2114 code), it works in the same way as it does for Perl's /g option. If all
2115 alternatives of a regex begin with \G, the expression is anchored to the start
2116 match position, and the "anchored" flag is set in the compiled expression.
2117
2118 15. Some bugs concerning the handling of certain option changes within patterns
2119 have been fixed. These applied to options other than (?ims). For example,
2120 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2121 been the other way round. Some of this was related to change 7 above.
2122
2123 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2124 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2125 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2126 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2127
2128 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2129 mode. Unlike ".", it always matches newline, whatever the setting of
2130 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2131 assertions. Perl allows it, but it doesn't (in general) work because it can't
2132 calculate the length of the lookbehind. At least, that's the case for Perl
2133 5.8.0 - I've been told they are going to document that it doesn't work in
2134 future.
2135
2136 18. Added an error diagnosis for escapes that PCRE does not support: these are
2137 \L, \l, \N, \P, \p, \U, \u, and \X.
2138
2139 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2140 reading past the end of the pattern in cases such as /[abcd/.
2141
2142 20. PCRE was getting more memory than necessary for patterns with classes that
2143 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2144
2145 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2146 compiling PCRE for use with Virtual Pascal.
2147
2148 22. Small fix to the Makefile to make it work properly if the build is done
2149 outside the source tree.
2150
2151 23. Added a new extension: a condition to go with recursion. If a conditional
2152 subpattern starts with (?(R) the "true" branch is used if recursion has
2153 happened, whereas the "false" branch is used only at the top level.
2154
2155 24. When there was a very long string of literal characters (over 255 bytes
2156 without UTF support, over 250 bytes with UTF support), the computation of how
2157 much memory was required could be incorrect, leading to segfaults or other
2158 strange effects.
2159
2160 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2161 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2162 there was a subsequent back reference to those brackets. This meant that, for
2163 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2164 possible to check for precisely this case. All we can do is abandon the
2165 optimization if .* occurs inside capturing brackets when there are any back
2166 references whatsoever. (See below for a better fix that came later.)
2167
2168 26. The handling of the optimization for finding the first character of a
2169 non-anchored pattern, and for finding a character that is required later in the
2170 match were failing in some cases. This didn't break the matching; it just
2171 failed to optimize when it could. The way this is done has been re-implemented.
2172
2173 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2174
2175 28. Added a new feature that provides some of the functionality that Perl
2176 provides with (?{...}). The facility is termed a "callout". The way it is done
2177 in PCRE is for the caller to provide an optional function, by setting
2178 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2179 global variable. By default it is unset, which disables all calling out. To get
2180 the function called, the regex must include (?C) at appropriate points. This
2181 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2182 This provides a means of identifying different callout points. When PCRE
2183 reaches such a point in the regex, if pcre_callout has been set, the external
2184 function is called. It is provided with data in a structure called
2185 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2186 matching continues; if it returns a non-zero value, the match at the current
2187 point fails. However, backtracking will occur if possible. [This was changed
2188 later and other features added - see item 49 below.]
2189
2190 29. pcretest is upgraded to test the callout functionality. It provides a
2191 callout function that displays information. By default, it shows the start of
2192 the match and the current position in the text. There are some new data escapes
2193 to vary what happens:
2194
2195 \C+ in addition, show current contents of captured substrings
2196 \C- do not supply a callout function
2197 \C!n return 1 when callout number n is reached
2198 \C!n!m return 1 when callout number n is reached for the mth time
2199
2200 30. If pcregrep was called with the -l option and just a single file name, it
2201 output "<stdin>" if a match was found, instead of the file name.
2202
2203 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2204 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2205 pcre_exec(). This saves a malloc/free per call. The default value of
2206 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2207 when configuring.
2208
2209 32. The default maximum size of a compiled pattern is 64K. There have been a
2210 few cases of people hitting this limit. The code now uses macros to handle the
2211 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2212 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2213 configuring. Tests 2 and 5 work only with 2-byte links because they output
2214 debugging information about compiled patterns.
2215
2216 33. Internal code re-arrangements:
2217
2218 (a) Moved the debugging function for printing out a compiled regex into
2219 its own source file (printint.c) and used #include to pull it into
2220 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2221 separate copies.
2222
2223 (b) Defined the list of op-code names for debugging as a macro in
2224 internal.h so that it is next to the definition of the opcodes.
2225
2226 (c) Defined a table of op-code lengths for simpler skipping along compiled
2227 code. This is again a macro in internal.h so that it is next to the
2228 definition of the opcodes.
2229
2230 34. Added support for recursive calls to individual subpatterns, along the
2231 lines of Robin Houston's patch (but implemented somewhat differently).
2232
2233 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2234 allow it to read and process whole directories in Win32. This code was
2235 contributed by Lionel Fourquaux; it has not been tested by me.
2236
2237 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2238 used to name a group. Names consist of alphanumerics and underscores, and must
2239 be unique. Back references use the syntax (?P=name) and recursive calls use
2240 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2241 numbers. The function pcre_fullinfo() can be used after compilation to extract
2242 a name/number map. There are three relevant calls:
2243
2244 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2245 PCRE_INFO_NAMECOUNT yields the number of entries
2246 PCRE_INFO_NAMETABLE yields a pointer to the map.
2247
2248 The map is a vector of fixed-size entries. The size of each entry depends on
2249 the length of the longest name used. The first two bytes of each entry are the
2250 group number, most significant byte first. There follows the corresponding
2251 name, zero terminated. The names are in alphabetical order.
2252
2253 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2254 case instead of 255. Making it the same both with and without UTF-8 support
2255 means that the same test output works with both.
2256
2257 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2258 calling malloc() with a zero argument.
2259
2260 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2261 optimization. I've improved things by keeping a bitmap of backreferences with
2262 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2263 fact referenced, the optimization can be applied. It is unlikely that a
2264 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2265 the match to follow \n) will appear inside brackets with a number greater than
2266 31, but if it does, any back reference > 31 suppresses the optimization.
2267
2268 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2269 of disabling numbered capturing parentheses. Any opening parenthesis that is
2270 not followed by ? behaves as if it were followed by ?: but named parentheses
2271 can still be used for capturing (and they will acquire numbers in the usual
2272 way).
2273
2274 41. Redesigned the return codes from the match() function into yes/no/error so
2275 that errors can be passed back from deep inside the nested calls. A malloc
2276 failure while inside a recursive subpattern call now causes the
2277 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2278
2279 42. It is now possible to set a limit on the number of times the match()
2280 function is called in a call to pcre_exec(). This facility makes it possible to
2281 limit the amount of recursion and backtracking, though not in a directly
2282 obvious way, because the match() function is used in a number of different
2283 circumstances. The count starts from zero for each position in the subject
2284 string (for non-anchored patterns). The default limit is, for compatibility, a
2285 large number, namely 10 000 000. You can change this in two ways:
2286
2287 (a) When configuring PCRE before making, you can use --with-match-limit=n
2288 to set a default value for the compiled library.
2289
2290 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2291 a different value is set. See 45 below.
2292
2293 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2294
2295 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2296 of things that can be changed at compile time. The first argument specifies
2297 what is wanted and the second points to where the information is to be placed.
2298 The current list of available information is:
2299
2300 PCRE_CONFIG_UTF8
2301
2302 The output is an integer that is set to one if UTF-8 support is available;
2303 otherwise it is set to zero.
2304
2305 PCRE_CONFIG_NEWLINE
2306
2307 The output is an integer that it set to the value of the code that is used for
2308 newline. It is either LF (10) or CR (13).
2309
2310 PCRE_CONFIG_LINK_SIZE
2311
2312 The output is an integer that contains the number of bytes used for internal
2313 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2314
2315 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2316
2317 The output is an integer that contains the threshold above which the POSIX
2318 interface uses malloc() for output vectors. See item 31 above.
2319
2320 PCRE_CONFIG_MATCH_LIMIT
2321
2322 The output is an unsigned integer that contains the default limit of the number
2323 of match() calls in a pcre_exec() execution. See 42 above.
2324
2325 44. pcretest has been upgraded by the addition of the -C option. This causes it
2326 to extract all the available output from the new pcre_config() function, and to
2327 output it. The program then exits immediately.
2328
2329 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2330 order to support additional features. One way would have been to define
2331 pcre_exec2() (for example) with extra arguments, but this would not have been
2332 extensible, and would also have required all calls to the original function to
2333 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2334 is used for passing in "extra" data from pcre_study().
2335
2336 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2337 contains the following fields:
2338
2339 flags a bitmap indicating which of the following fields are set
2340 study_data opaque data from pcre_study()
2341 match_limit a way of specifying a limit on match() calls for a specific
2342 call to pcre_exec()
2343 callout_data data for callouts (see 49 below)
2344
2345 The flag bits are also defined in pcre.h, and are
2346
2347 PCRE_EXTRA_STUDY_DATA
2348 PCRE_EXTRA_MATCH_LIMIT
2349 PCRE_EXTRA_CALLOUT_DATA
2350
2351 The pcre_study() function now returns one of these new pcre_extra blocks, with
2352 the actual study data pointed to by the study_data field, and the
2353 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2354 before. That is, this change is entirely upwards-compatible and requires no
2355 change to existing code.
2356
2357 If you want to pass in additional data to pcre_exec(), you can either place it
2358 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2359 block.
2360
2361 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2362 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2363 times with different match limits, until it finds the minimum value needed for
2364 pcre_exec() to complete. The value is then output. This can be instructive; for
2365 most simple matches the number is quite small, but for pathological cases it
2366 gets very large very quickly.
2367
2368 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2369 returns the size of the data block pointed to by the study_data field in a
2370 pcre_extra block, that is, the value that was passed as the argument to
2371 pcre_malloc() when PCRE was getting memory in which to place the information
2372 created by pcre_study(). The fourth argument should point to a size_t variable.
2373 pcretest has been extended so that this information is shown after a successful
2374 pcre_study() call when information about the compiled regex is being displayed.
2375
2376 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2377 because what follows is always an absolute path. (Later: it turns out that this
2378 is more than cosmetic for MinGW, because it doesn't like empty path
2379 components.)
2380
2381 49. Some changes have been made to the callout feature (see 28 above):
2382
2383 (i) A callout function now has three choices for what it returns:
2384
2385 0 => success, carry on matching
2386 > 0 => failure at this point, but backtrack if possible
2387 < 0 => serious error, return this value from pcre_exec()
2388
2389 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2390 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2391 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2392 use by callout functions. It will never be used by PCRE itself.
2393
2394 (ii) The pcre_extra structure (see 45 above) has a void * field called
2395 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2396 pcre_callout_block structure has a field of the same name. The contents of
2397 the field passed in the pcre_extra structure are passed to the callout
2398 function in the corresponding field in the callout block. This makes it
2399 easier to use the same callout-containing regex from multiple threads. For
2400 testing, the pcretest program has a new data escape
2401
2402 \C*n pass the number n (may be negative) as callout_data
2403
2404 If the callout function in pcretest receives a non-zero value as
2405 callout_data, it returns that value.
2406
2407 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2408 there were some redundant $(CFLAGS) in commands that are now specified as
2409 $(LINK), which already includes $(CFLAGS).
2410
2411 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2412 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2413 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2414 one-byte characters throughout. Note that case-insensitive matching applies
2415 only to characters whose values are less than 256. PCRE doesn't support the
2416 notion of cases for higher-valued characters.
2417
2418 (i) A character class whose characters are all within 0-255 is handled as
2419 a bit map, and the map is inverted for negative classes. Previously, a
2420 character > 255 always failed to match such a class; however it should
2421 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2422
2423 (ii) A negated character class with a single character < 255 is coded as
2424 "not this character" (OP_NOT). This wasn't working properly when the test
2425 character was multibyte, either singly or repeated.
2426
2427 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2428 mode, for example: \x{100}{2,3}.
2429
2430 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2431 singly or repeated) now correctly test multibyte characters. However,
2432 PCRE doesn't recognize any characters with values greater than 255 as
2433 digits, spaces, or word characters. Such characters always match \D, \S,
2434 and \W, and never match \d, \s, or \w.
2435
2436 (v) Classes may now contain characters and character ranges with values
2437 greater than 255. For example: [ab\x{100}-\x{400}].
2438
2439 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2440 PCRE in UTF-8 mode.
2441
2442 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2443 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2444 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2445 value.)
2446
2447 53. The single man page has become too large. I have therefore split it up into
2448 a number of separate man pages. These also give rise to individual HTML pages;
2449 these are now put in a separate directory, and there is an index.html page that
2450 lists them all. Some hyperlinking between the pages has been installed.
2451
2452 54. Added convenience functions for handling named capturing parentheses.
2453
2454 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2455 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2456 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2457 are faulted.
2458
2459 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2460 calling configure. These values are used when compiling the dftables.c program
2461 which is run to generate the source of the default character tables. They
2462 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2463 you will need to set these values.
2464
2465 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2466
2467
2468 Version 3.9 02-Jan-02
2469 ---------------------
2470
2471 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2472
2473 2. If --disable-static was given, the building process failed when trying to
2474 build pcretest and pcregrep. (For some reason it was using libtool to compile
2475 them, which is not right, as they aren't part of the library.)
2476
2477
2478 Version 3.8 18-Dec-01
2479 ---------------------
2480
2481 1. The experimental UTF-8 code was completely screwed up. It was packing the
2482 bytes in the wrong order. How dumb can you get?
2483
2484
2485 Version 3.7 29-Oct-01
2486 ---------------------
2487
2488 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2489 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2490 this didn't happen under Solaris 8, where I normally test things.
2491
2492 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2493 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2494 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2495 this fix an hour or so after the initial 3.7 release.)
2496
2497
2498 Version 3.6 23-Oct-01
2499 ---------------------
2500
2501 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2502 offsets passed as NULL with zero offset count.
2503
2504 2. The config.guess and config.sub files had not been updated when I moved to
2505 the latest autoconf.
2506
2507
2508 Version 3.5 15-Aug-01
2509 ---------------------
2510
2511 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2512 had been forgotten.
2513
2514 2. By using declared but undefined structures, we can avoid using "void"
2515 definitions in pcre.h while keeping the internal definitions of the structures
2516 private.
2517
2518 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2519 user point of view, this means that both static and shared libraries are built
2520 by default, but this can be individually controlled. More of the work of
2521 handling this static/shared cases is now inside libtool instead of PCRE's make
2522 file.
2523
2524 4. The pcretest utility is now installed along with pcregrep because it is
2525 useful for users (to test regexs) and by doing this, it automatically gets
2526 relinked by libtool. The documentation has been turned into a man page, so
2527 there are now .1, .txt, and .html versions in /doc.
2528
2529 5. Upgrades to pcregrep:
2530 (i) Added long-form option names like gnu grep.
2531 (ii) Added --help to list all options with an explanatory phrase.
2532 (iii) Added -r, --recursive to recurse into sub-directories.
2533 (iv) Added -f, --file to read patterns from a file.
2534
2535 6. pcre_exec() was referring to its "code" argument before testing that
2536 argument for NULL (and giving an error if it was NULL).
2537
2538 7. Upgraded Makefile.in to allow for compiling in a different directory from
2539 the source directory.
2540
2541 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2542 options bits, the pointer it was passed was to an int instead of to an unsigned
2543 long int. This mattered only on 64-bit systems.
2544
2545 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2546 generated) instead of pcre.in, which it its source. Also made the same change
2547 in several of the .c files.
2548
2549 10. A new release of gcc defines printf() as a macro, which broke pcretest
2550 because it had an ifdef in the middle of a string argument for printf(). Fixed
2551 by using separate calls to printf().
2552
2553 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2554 script, to force use of CR or LF instead of \n in the source. On non-Unix
2555 systems, the value can be set in config.h.
2556
2557 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2558 absolute limit. Changed the text of the error message to make this clear, and
2559 likewise updated the man page.
2560
2561 13. The limit of 99 on the number of capturing subpatterns has been removed.
2562 The new limit is 65535, which I hope will not be a "real" limit.
2563
2564
2565 Version 3.4 22-Aug-00
2566 ---------------------
2567
2568 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2569
2570 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2571
2572
2573 Version 3.3 01-Aug-00
2574 ---------------------
2575
2576 1. If an octal character was given, but the value was greater than \377, it
2577 was not getting masked to the least significant bits, as documented. This could
2578 lead to crashes in some systems.
2579
2580 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2581 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2582
2583 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2584 These just pass their arguments on to (pcre_free)(), but they are provided
2585 because some uses of PCRE bind it to non-C systems that can call its functions,
2586 but cannot call free() or pcre_free() directly.
2587
2588 4. Add "make test" as a synonym for "make check". Corrected some comments in
2589 the Makefile.
2590
2591 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2592 Makefile.
2593
2594 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2595 command called pgrep for grepping around the active processes.
2596
2597 7. Added the beginnings of support for UTF-8 character strings.
2598
2599 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2600 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2601 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2602 out for the ar command.)
2603
2604
2605 Version 3.2 12-May-00
2606 ---------------------
2607
2608 This is purely a bug fixing release.
2609
2610 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2611 of ZA. This was just one example of several cases that could provoke this bug,
2612 which was introduced by change 9 of version 2.00. The code for breaking
2613 infinite loops after an iteration that matches an empty string was't working
2614 correctly.
2615
2616 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2617 when matched against abbab (for example). After matching an empty string, it
2618 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2619 caused it to match further down the string than it should.
2620
2621 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2622 was there because it doesn't seem to be needed, and it causes trouble on some
2623 systems, as it is not a Standard C header. It has been removed.
2624
2625 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2626 were reported on the Macintosh. The changes were from
2627
2628 while ((c = *(++ptr)) != 0 && c != '\n');
2629 to
2630 while ((c = *(++ptr)) != 0 && c != '\n') ;
2631
2632 Totally extraordinary, but if that's what it takes...
2633
2634 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2635 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2636 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2637 assumes the way PCRE uses memmove() (always moving upwards).
2638
2639 6. PCRE is being used in one environment where strchr() is not available. There
2640 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2641 faster code anyway.
2642
2643
2644 Version 3.1 09-Feb-00
2645 ---------------------
2646
2647 The only change in this release is the fixing of some bugs in Makefile.in for
2648 the "install" target:
2649
2650 (1) It was failing to install pcreposix.h.
2651
2652 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2653
2654
2655 Version 3.0 01-Feb-00
2656 ---------------------
2657
2658 1. Add support for the /+ modifier to perltest (to output $` like it does in
2659 pcretest).
2660
2661 2. Add support for the /g modifier to perltest.
2662
2663 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2664 matches null strings.
2665
2666 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2667 pattern. Perl treats empty patterns specially - it reuses the most recent
2668 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2669 effect.
2670
2671 5. The POSIX interface was broken in that it was just handing over the POSIX
2672 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2673 required a bigger vector, with some working space on the end. This means that
2674 the POSIX wrapper now has to get and free some memory, and copy the results.
2675
2676 6. Added some simple autoconf support, placing the test data and the
2677 documentation in separate directories, re-organizing some of the
2678 information files, and making it build pcre-config (a GNU standard). Also added
2679 libtool support for building PCRE as a shared library, which is now the
2680 default.
2681
2682 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2683 09 are not valid octal constants. Single digits will be used for minor values
2684 less than 10.
2685
2686 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2687 existing programs that set these in the POSIX interface can use PCRE without
2688 modification.
2689
2690 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2691 return all that pcre_info() returns, plus additional data. The pcre_info()
2692 function is retained for compatibility, but is considered to be obsolete.
2693
2694 10. Added experimental recursion feature (?R) to handle one common case that
2695 Perl 5.6 will be able to do with (?p{...}).
2696
2697 11. Added support for POSIX character classes like [:alpha:], which Perl is
2698 adopting.
2699
2700
2701 Version 2.08 31-Aug-99
2702 ----------------------
2703
2704 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2705 trying to match at the startoffset position, but instead was moving forward to
2706 the next newline as if a previous match had failed.
2707
2708 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2709 and could get into a loop if a null string was matched other than at the start
2710 of the subject.
2711
2712 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2713 be distinguished at compile time, and for completeness also added PCRE_DATE.
2714
2715 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2716 in GnuWin32 environments.
2717
2718
2719 Version 2.07 29-Jul-99
2720 ----------------------
2721
2722 1. The documentation is now supplied in plain text form and HTML as well as in
2723 the form of man page sources.
2724
2725 2. C++ compilers don't like assigning (void *) values to other pointer types.
2726 In particular this affects malloc(). Although there is no problem in Standard
2727 C, I've put in casts to keep C++ compilers happy.
2728
2729 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2730 should be (const char *).
2731
2732 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2733 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2734 However, I haven't made this a standard facility. The documentation doesn't
2735 mention it, and the Makefile doesn't support it.
2736
2737 5. The Makefile now contains an "install" target, with editable destinations at
2738 the top of the file. The pcretest program is not installed.
2739
2740 6. pgrep -V now gives the PCRE version number and date.
2741
2742 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2743 causing the entire string to be ignored, instead of just the last character.
2744
2745 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2746 non-matching string, it can take a very, very long time, even for strings of
2747 quite modest length, because of the nested recursion. PCRE now does better in
2748 some of these cases. It does this by remembering the last required literal
2749 character in the pattern, and pre-searching the subject to ensure it is present
2750 before running the real match. In other words, it applies a heuristic to detect
2751 some types of certain failure quickly, and in the above example, if presented
2752 with a string that has no trailing " it gives "no match" very quickly.
2753
2754 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2755 other alternatives are tried instead.
2756
2757
2758 Version 2.06 09-Jun-99
2759 ----------------------
2760
2761 1. Change pcretest's output for amount of store used to show just the code
2762 space, because the remainder (the data block) varies in size between 32-bit and
2763 64-bit systems.
2764
2765 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2766 start matching at. This allows lookbehinds to work when searching for multiple
2767 occurrences in a string.
2768
2769 3. Added additional options to pcretest for testing multiple occurrences:
2770
2771 /+ outputs the rest of the string that follows a match
2772 /g loops for multiple occurrences, using the new startoffset argument
2773 /G loops for multiple occurrences by passing an incremented pointer
2774
2775 4. PCRE wasn't doing the "first character" optimization for patterns starting
2776 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2777 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2778 the letter 'x'. On long subject strings, this gives a significant speed-up.
2779
2780
2781 Version 2.05 21-Apr-99
2782 ----------------------
2783
2784 1. Changed the type of magic_number from int to long int so that it works
2785 properly on 16-bit systems.
2786
2787 2. Fixed a bug which caused patterns starting with .* not to work correctly
2788 when the subject string contained newline characters. PCRE was assuming
2789 anchoring for such patterns in all cases, which is not correct because .* will
2790 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2791 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2792 must be retried after every newline in the subject.
2793
2794
2795 Version 2.04 18-Feb-99
2796 ----------------------
2797
2798 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2799 computation of the store needed to hold the pattern was incorrect (too large).
2800 If such patterns were nested a few deep, this could multiply and become a real
2801 problem.
2802
2803 2. Added /M option to pcretest to show the memory requirement of a specific
2804 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2805
2806 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2807 compiled in such a way that the backtracking after subsequent failure was
2808 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2809 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2810
2811
2812 Version 2.03 02-Feb-99
2813 ----------------------
2814
2815 1. Fixed typo and small mistake in man page.
2816
2817 2. Added 4th condition (GPL supersedes if conflict) and created separate
2818 LICENCE file containing the conditions.
2819
2820 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2821 Perl, that is the internal \ allows the delimiter to be included in the
2822 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2823 the final delimiter, add \ to the end of the pattern (to test the error).
2824
2825 4. Added the convenience functions for extracting substrings after a successful
2826 match. Updated pcretest to make it able to test these functions.
2827
2828
2829 Version 2.02 14-Jan-99
2830 ----------------------
2831
2832 1. Initialized the working variables associated with each extraction so that
2833 their saving and restoring doesn't refer to uninitialized store.
2834
2835 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2836 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2837 fix the problem.
2838
2839 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2840 calls, and wasn't printing the correct value for compiling calls. Increased the
2841 default value of LOOPREPEAT, and the number of significant figures in the
2842 times.
2843
2844 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2845
2846 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2847 a building problem on Windows NT with a FAT file system.
2848
2849
2850 Version 2.01 21-Oct-98
2851 ----------------------
2852
2853 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2854 to character tables built by pcre_maketables() in the current locale. If NULL
2855 is passed, the default tables are used.
2856
2857
2858 Version 2.00 24-Sep-98
2859 ----------------------
2860
2861 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2862 it any more.
2863
2864 2. Allow quantification of (?>) groups, and make it work correctly.
2865
2866 3. The first character computation wasn't working for (?>) groups.
2867
2868 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2869 end of the subject) and add 5.005's \z, which really does match only at the
2870 very end of the subject.
2871
2872 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2873
2874 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2875 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2876 localized options. All options to pcre_study() were also removed.
2877
2878 7. Add other new features from 5.005:
2879
2880 $(?<= positive lookbehind
2881 $(?<! negative lookbehind
2882 (?imsx-imsx) added the unsetting capability
2883 such a setting is global if at outer level; local otherwise
2884 (?imsx-imsx:) non-capturing groups with option setting
2885 (?(cond)re|re) conditional pattern matching
2886
2887 A backreference to itself in a repeated group matches the previous
2888 captured string.
2889
2890 8. General tidying up of studying (both automatic and via "study")
2891 consequential on the addition of new assertions.
2892
2893 9. As in 5.005, unlimited repeated groups that could match an empty substring
2894 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2895 runtime if any iteration does actually match an empty substring.
2896
2897 10. Include the RunTest script in the distribution.
2898
2899 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2900 discrepancies, some of which were old and were also with respect to 5.004. They
2901 have now been fixed.
2902
2903
2904 Version 1.09 28-Apr-98
2905 ----------------------
2906
2907 1. A negated single character class followed by a quantifier with a minimum
2908 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2909 program crashes, or just wrong answers. This did not apply to negated classes
2910 containing more than one character, or to minima other than one.
2911
2912
2913 Version 1.08 27-Mar-98
2914 ----------------------
2915
2916 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2917
2918 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2919 latter must appear before anything that relies on it in the pattern.
2920
2921
2922 Version 1.07 16-Feb-98
2923 ----------------------
2924
2925 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2926 repeat of a potentially empty string).
2927
2928
2929 Version 1.06 23-Jan-98
2930 ----------------------
2931
2932 1. Added Markus Oberhumer's little patches for C++.
2933
2934 2. Literal strings longer than 255 characters were broken.
2935
2936
2937 Version 1.05 23-Dec-97
2938 ----------------------
2939
2940 1. Negated character classes containing more than one character were failing if
2941 PCRE_CASELESS was set at run time.
2942
2943
2944 Version 1.04 19-Dec-97
2945 ----------------------
2946
2947 1. Corrected the man page, where some "const" qualifiers had been omitted.
2948
2949 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2950 input syntax.
2951
2952 3. Fixed memory leak which occurred when a regex with back references was
2953 matched with an offsets vector that wasn't big enough. The temporary memory
2954 that is used in this case wasn't being freed if the match failed.
2955
2956 4. Tidied pcretest to ensure it frees memory that it gets.
2957
2958 5. Temporary memory was being obtained in the case where the passed offsets
2959 vector was exactly big enough.
2960
2961 6. Corrected definition of offsetof() from change 5 below.
2962
2963 7. I had screwed up change 6 below and broken the rules for the use of
2964 setjmp(). Now fixed.
2965
2966
2967 Version 1.03 18-Dec-97
2968 ----------------------
2969
2970 1. A erroneous regex with a missing opening parenthesis was correctly
2971 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2972 on some systems.
2973
2974 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2975 it was reported that one broken compiler failed on the former because "code" is
2976 also an independent variable.
2977
2978 3. The erroneous regex a[]b caused an array overrun reference.
2979
2980 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2981 fail on data ending with that character. (It was going on too far, and checking
2982 the next character, typically a binary zero.) This was specific to the
2983 optimized code for single-character negative classes.
2984
2985 5. Added a contributed patch from the TIN world which does the following:
2986
2987 + Add an undef for memmove, in case the the system defines a macro for it.
2988
2989 + Add a definition of offsetof(), in case there isn't one. (I don't know
2990 the reason behind this - offsetof() is part of the ANSI standard - but
2991 it does no harm).
2992
2993 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2994 most of the places where whitespace preceded '#'. I have given up and
2995 allowed the remaining 2 cases to be at the margin.
2996
2997 + Rename some variables in pcre to eliminate shadowing. This seems very
2998 pedantic, but does no harm, of course.
2999
3000 6. Moved the call to setjmp() into its own function, to get rid of warnings
3001 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3002
3003 7. Constructs such as \d{8,} were compiling into the equivalent of
3004 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3005 outcome, but in this particular case used more store than had been allocated,
3006 which caused the bug to be discovered because it threw up an internal error.
3007
3008 8. The debugging code in both pcre and pcretest for outputting the compiled
3009 form of a regex was going wrong in the case of back references followed by
3010 curly-bracketed repeats.
3011
3012
3013 Version 1.02 12-Dec-97
3014 ----------------------
3015
3016 1. Typos in pcre.3 and comments in the source fixed.
3017
3018 2. Applied a contributed patch to get rid of places where it used to remove
3019 'const' from variables, and fixed some signed/unsigned and uninitialized
3020 variable warnings.
3021
3022 3. Added the "runtest" target to Makefile.
3023
3024 4. Set default compiler flag to -O2 rather than just -O.
3025
3026
3027 Version 1.01 19-Nov-97
3028 ----------------------
3029
3030 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3031 like /([ab]*)*/, that is, for classes with more than one character in them.
3032
3033 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3034 as /((?>a*))*/ (a PCRE_EXTRA facility).
3035
3036
3037 Version 1.00 18-Nov-97
3038 ----------------------
3039
3040 1. Added compile-time macros to support systems such as SunOS4 which don't have
3041 memmove() or strerror() but have other things that can be used instead.
3042
3043 2. Arranged that "make clean" removes the executables.
3044
3045
3046 Version 0.99 27-Oct-97
3047 ----------------------
3048
3049 1. Fixed bug in code for optimizing classes with only one character. It was
3050 initializing a 32-byte map regardless, which could cause it to run off the end
3051 of the memory it had got.
3052
3053 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3054
3055
3056 Version 0.98 22-Oct-97
3057 ----------------------
3058
3059 1. Fixed bug in code for handling temporary memory usage when there are more
3060 back references than supplied space in the ovector. This could cause segfaults.
3061
3062
3063 Version 0.97 21-Oct-97
3064 ----------------------
3065
3066 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3067
3068 2. Optimized negated single characters not to use a bit map.
3069
3070 3. Brought error texts together as macro definitions; clarified some of them;
3071 fixed one that was wrong - it said "range out of order" when it meant "invalid
3072 escape sequence".
3073
3074 4. Changed some char * arguments to const char *.
3075
3076 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3077
3078 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3079 pcretest.
3080
3081
3082 Version 0.96 16-Oct-97
3083 ----------------------
3084
3085 1. Added a simple "pgrep" utility to the distribution.
3086
3087 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3088 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3089 where "ddd" means "one or more decimal digits".
3090
3091 3. Fixed serious bug. If a pattern had a back reference, but the call to
3092 pcre_exec() didn't supply a large enough ovector to record the related
3093 identifying subpattern, the match always failed. PCRE now remembers the number
3094 of the largest back reference, and gets some temporary memory in which to save
3095 the offsets during matching if necessary, in order to ensure that
3096 backreferences always work.
3097
3098 4. Increased the compatibility with Perl in a number of ways:
3099
3100 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3101 to request this handling. The option can be set at compile or exec time.
3102
3103 (b) $ matches before a terminating newline by default; an option
3104 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3105 mode). The option can be set at compile or exec time.
3106
3107 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3108 the same as Perl's. If the decimal number it represents is less than 10
3109 or there aren't that many previous left capturing parentheses, an octal
3110 escape is read. Inside a character class, it's always an octal escape,
3111 even if it is a single digit.
3112
3113 (d) An escaped but undefined alphabetic character is taken as a literal,
3114 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3115 escapes.
3116
3117 (e) {0} is now permitted. (The previous item is removed from the compiled
3118 pattern).
3119
3120 5. Changed all the names of code files so that the basic parts are no longer
3121 than 10 characters, and abolished the teeny "globals.c" file.
3122
3123 6. Changed the handling of character classes; they are now done with a 32-byte
3124 bit map always.
3125
3126 7. Added the -d and /D options to pcretest to make it possible to look at the
3127 internals of compilation without having to recompile pcre.
3128
3129
3130 Version 0.95 23-Sep-97
3131 ----------------------
3132
3133 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3134 \x20 at the start of a run of normal characters. These were being treated as
3135 real characters, instead of the source characters being re-checked.
3136
3137
3138 Version 0.94 18-Sep-97
3139 ----------------------
3140
3141 1. The functions are now thread-safe, with the caveat that the global variables
3142 containing pointers to malloc() and free() or alternative functions are the
3143 same for all threads.
3144
3145 2. Get pcre_study() to generate a bitmap of initial characters for non-
3146 anchored patterns when this is possible, and use it if passed to pcre_exec().
3147
3148
3149 Version 0.93 15-Sep-97
3150 ----------------------
3151
3152 1. /(b)|(:+)/ was computing an incorrect first character.
3153
3154 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3155 but not actually doing anything yet.
3156
3157 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3158 as Perl does (e.g. [-az] or [az-]).
3159
3160 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3161 all possible positions.
3162
3163 5. Split up into different modules to avoid including unneeded functions in a
3164 compiled binary. However, compile and exec are still in one module. The "study"
3165 function is split off.
3166
3167 6. The character tables are now in a separate module whose source is generated
3168 by an auxiliary program - but can then be edited by hand if required. There are
3169 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3170 toupper() in the code.
3171
3172 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3173 make them global. Abolish the function for setting them, as the caller can now
3174 set them directly.
3175
3176
3177 Version 0.92 11-Sep-97
3178 ----------------------
3179
3180 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3181 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3182
3183 2. Caseless matching was not working in character classes if the characters in
3184 the pattern were in upper case.
3185
3186 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3187
3188 4. Make PCRE_ANCHORED public and accept as a compile option.
3189
3190 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3191 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3192 pass them.
3193
3194 6. Give an error if bad option bits passed at compile or run time.
3195
3196 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3197 pcretest to cause it to pass that flag.
3198
3199 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3200 options, and the first character, if set.
3201
3202 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3203
3204
3205 Version 0.91 10-Sep-97
3206 ----------------------
3207
3208 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3209 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3210
3211 2. PCRE was looping on encountering an indefinitely repeated back reference to
3212 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3213 Perl does - treats the match as successful.
3214
3215 ****

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