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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 322 - (show annotations) (download)
Wed Mar 5 17:14:08 2008 UTC (6 years, 4 months ago) by ph10
File size: 151200 byte(s)
Craig's patch to pcrecpp.cc to restore ABI compatibility with pre-7.6 versions,
which defined a global no_arg variable instead of putting it in the RE class.

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

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