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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 321 - (show annotations) (download)
Wed Mar 5 17:10:45 2008 UTC (6 years, 7 months ago) by ph10
File size: 151017 byte(s)
Craig's patch to sort out a long long problem: "If we can't convert 
a string to a long long, pretend we don't even have a long long."

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

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