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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 327 - (show annotations) (download)
Sat Mar 8 19:38:30 2008 UTC (6 years, 8 months ago) by ph10
File size: 152438 byte(s)
Added some missing "const"s to declarations of static tables in 
pcre_compile.c and pcre_dfa_exec.c.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.7 05-Mar-08
5 ---------------------
6
7 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
8 a string to a long long, pretend we don't even have a long long." This is
9 done by checking for the strtoq, strtoll, and _strtoi64 functions.
10
11 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
12 pre-7.6 versions, which defined a global no_arg variable instead of putting
13 it in the RE class.
14
15 3. Remove a line of dead code, identified by coverity and reported by Nuno
16 Lopes.
17
18 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
19
20 (1) The include/exclude patterns were being applied to the whole pathnames
21 of files, instead of just to the final components.
22
23 (2) If there was more than one level of directory, the subdirectories were
24 skipped unless they satisfied the include/exclude conditions. This is
25 inconsistent with GNU grep (and could even be seen as contrary to the
26 pcregrep specification - which I improved to make it absolutely clear).
27 The action now is always to scan all levels of directory, and just
28 apply the include/exclude patterns to regular files.
29
30 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
31 --exclude_dir in the tests to avoid scanning .svn directories.
32
33 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
34 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
35 doesn't support NULs in patterns.
36
37 7. Added some missing "const"s to declarations of static tables in
38 pcre_compile.c and pcre_dfa_exec.c.
39
40
41 Version 7.6 28-Jan-08
42 ---------------------
43
44 1. A character class containing a very large number of characters with
45 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
46 overflow.
47
48 2. Patch to cut out the "long long" test in pcrecpp_unittest when
49 HAVE_LONG_LONG is not defined.
50
51 3. Applied Christian Ehrlicher's patch to update the CMake build files to
52 bring them up to date and include new features. This patch includes:
53
54 - Fixed PH's badly added libz and libbz2 support.
55 - Fixed a problem with static linking.
56 - Added pcredemo. [But later removed - see 7 below.]
57 - Fixed dftables problem and added an option.
58 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
59 HAVE_LONG_LONG.
60 - Added readline support for pcretest.
61 - Added an listing of the option settings after cmake has run.
62
63 4. A user submitted a patch to Makefile that makes it easy to create
64 "pcre.dll" under mingw when using Configure/Make. I added stuff to
65 Makefile.am that cause it to include this special target, without
66 affecting anything else. Note that the same mingw target plus all
67 the other distribution libraries and programs are now supported
68 when configuring with CMake (see 6 below) instead of with
69 Configure/Make.
70
71 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
72 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
73 exported in the Windows port". It has not yet been confirmed that the patch
74 solves the problem, but it does no harm.
75
76 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
77 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
78 with CMake, and also correct the comment about stack recursion.
79
80 7. Remove the automatic building of pcredemo from the ./configure system and
81 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
82 of a program that users should build themselves after PCRE is installed, so
83 building it automatically is not really right. What is more, it gave
84 trouble in some build environments.
85
86 8. Further tidies to CMakeLists.txt from Sheri and Christian.
87
88
89 Version 7.5 10-Jan-08
90 ---------------------
91
92 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
93 values in parens when parsing an RE using the C++ wrapper."
94
95 2. Negative specials like \S did not work in character classes in UTF-8 mode.
96 Characters greater than 255 were excluded from the class instead of being
97 included.
98
99 3. The same bug as (2) above applied to negated POSIX classes such as
100 [:^space:].
101
102 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
103 defined or documented. It seems to have been a typo for PCRE_STATIC, so
104 I have changed it.
105
106 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
107 first named subpattern) and a construct such as (?&a) would reference the
108 first named subpattern whose name started with "a" (in other words, the
109 length check was missing). Both these problems are fixed. "Subpattern name
110 expected" is now given for (?&) (a zero-length name), and this patch also
111 makes it give the same error for \k'' (previously it complained that that
112 was a reference to a non-existent subpattern).
113
114 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
115 this is right because (?- can be followed by option settings as well as by
116 digits. I have, however, made the messages clearer.
117
118 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
119 than the number used in the conditional) now cause a compile-time error.
120 This is actually not compatible with Perl, which accepts such patterns, but
121 treats the conditional as always being FALSE (as PCRE used to), but it
122 seems to me that giving a diagnostic is better.
123
124 8. Change "alphameric" to the more common word "alphanumeric" in comments
125 and messages.
126
127 9. Fix two occurrences of "backslash" in comments that should have been
128 "backspace".
129
130 10. Remove two redundant lines of code that can never be obeyed (their function
131 was moved elsewhere).
132
133 11. The program that makes PCRE's Unicode character property table had a bug
134 which caused it to generate incorrect table entries for sequences of
135 characters that have the same character type, but are in different scripts.
136 It amalgamated them into a single range, with the script of the first of
137 them. In other words, some characters were in the wrong script. There were
138 thirteen such cases, affecting characters in the following ranges:
139
140 U+002b0 - U+002c1
141 U+0060c - U+0060d
142 U+0061e - U+00612
143 U+0064b - U+0065e
144 U+0074d - U+0076d
145 U+01800 - U+01805
146 U+01d00 - U+01d77
147 U+01d9b - U+01dbf
148 U+0200b - U+0200f
149 U+030fc - U+030fe
150 U+03260 - U+0327f
151 U+0fb46 - U+0fbb1
152 U+10450 - U+1049d
153
154 12. The -o option (show only the matching part of a line) for pcregrep was not
155 compatible with GNU grep in that, if there was more than one match in a
156 line, it showed only the first of them. It now behaves in the same way as
157 GNU grep.
158
159 13. If the -o and -v options were combined for pcregrep, it printed a blank
160 line for every non-matching line. GNU grep prints nothing, and pcregrep now
161 does the same. The return code can be used to tell if there were any
162 non-matching lines.
163
164 14. Added --file-offsets and --line-offsets to pcregrep.
165
166 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
167 infinitely looping recursion. The bug was that positive lookaheads were not
168 being skipped when checking for a possible empty match (negative lookaheads
169 and both kinds of lookbehind were skipped).
170
171 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
172 inclusion of <windows.h> to before rather than after the definition of
173 INVALID_FILE_ATTRIBUTES (patch from David Byron).
174
175 17. Specifying a possessive quantifier with a specific limit for a Unicode
176 character property caused pcre_compile() to compile bad code, which led at
177 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
178 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
179 caused the error; without that there was no problem.
180
181 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
182
183 19. Added --enable-pcretest-libreadline.
184
185 20. In pcrecpp.cc, the variable 'count' was incremented twice in
186 RE::GlobalReplace(). As a result, the number of replacements returned was
187 double what it should be. I removed one of the increments, but Craig sent a
188 later patch that removed the other one (the right fix) and added unit tests
189 that check the return values (which was not done before).
190
191 21. Several CMake things:
192
193 (1) Arranged that, when cmake is used on Unix, the libraries end up with
194 the names libpcre and libpcreposix, not just pcre and pcreposix.
195
196 (2) The above change means that pcretest and pcregrep are now correctly
197 linked with the newly-built libraries, not previously installed ones.
198
199 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
200
201 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
202 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
203 UTF-8 newline character). The key issue is that the pattern starts .*;
204 this means that the match must be either at the beginning, or after a
205 newline. The bug was in the code for advancing after a failed match and
206 checking that the new position followed a newline. It was not taking
207 account of UTF-8 characters correctly.
208
209 23. PCRE was behaving differently from Perl in the way it recognized POSIX
210 character classes. PCRE was not treating the sequence [:...:] as a
211 character class unless the ... were all letters. Perl, however, seems to
212 allow any characters between [: and :], though of course it rejects as
213 unknown any "names" that contain non-letters, because all the known class
214 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
215 for example, whereas PCRE did not - it did not recognize a POSIX character
216 class. This seemed a bit dangerous, so the code has been changed to be
217 closer to Perl. The behaviour is not identical to Perl, because PCRE will
218 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
219 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
220 Perl does, and where it didn't before.
221
222 24. Rewrite so as to remove the single use of %n from pcregrep because in some
223 Windows environments %n is disabled by default.
224
225
226 Version 7.4 21-Sep-07
227 ---------------------
228
229 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
230 means that a class such as [\s] counted as "explicit reference to CR or
231 LF". That isn't really right - the whole point of the change was to try to
232 help when there was an actual mention of one of the two characters. So now
233 the change happens only if \r or \n (or a literal CR or LF) character is
234 encountered.
235
236 2. The 32-bit options word was also used for 6 internal flags, but the numbers
237 of both had grown to the point where there were only 3 bits left.
238 Fortunately, there was spare space in the data structure, and so I have
239 moved the internal flags into a new 16-bit field to free up more option
240 bits.
241
242 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
243 but did not set the internal JCHANGED flag - either of these is enough to
244 control the way the "get" function works - but the PCRE_INFO_JCHANGED
245 facility is supposed to tell if (?J) was ever used, so now (?J) at the
246 start sets both bits.
247
248 4. Added options (at build time, compile time, exec time) to change \R from
249 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
250
251 5. doc/pcresyntax.html was missing from the distribution.
252
253 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
254 compatibility, even though it is no longer used.
255
256 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
257 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
258 windows.h file is present (where different names are used). [This was
259 reversed later after testing - see 16 below.]
260
261 8. Changed all #include <config.h> to #include "config.h". There were also
262 some further <pcre.h> cases that I changed to "pcre.h".
263
264 9. When pcregrep was used with the --colour option, it missed the line ending
265 sequence off the lines that it output.
266
267 10. It was pointed out to me that arrays of string pointers cause lots of
268 relocations when a shared library is dynamically loaded. A technique of
269 using a single long string with a table of offsets can drastically reduce
270 these. I have refactored PCRE in four places to do this. The result is
271 dramatic:
272
273 Originally: 290
274 After changing UCP table: 187
275 After changing error message table: 43
276 After changing table of "verbs" 36
277 After changing table of Posix names 22
278
279 Thanks to the folks working on Gregex for glib for this insight.
280
281 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
282 unicode-properties was also set.
283
284 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
285
286 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
287 checked only for CRLF.
288
289 14. Added casts to pcretest.c to avoid compiler warnings.
290
291 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
292
293 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
294 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
295 entirely. This removes changes made in 7 above.
296
297 17. The CMake files have been updated, and there is now more information about
298 building with CMake in the NON-UNIX-USE document.
299
300
301 Version 7.3 28-Aug-07
302 ---------------------
303
304 1. In the rejigging of the build system that eventually resulted in 7.1, the
305 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
306 brackets there is not right, since it causes compilers to look for an
307 installed pcre.h, not the version that is in the source that is being
308 compiled (which of course may be different). I have changed it back to:
309
310 #include "pcre.h"
311
312 I have a vague recollection that the change was concerned with compiling in
313 different directories, but in the new build system, that is taken care of
314 by the VPATH setting the Makefile.
315
316 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
317 when the subject happened to end in the byte 0x85 (e.g. if the last
318 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
319 characters but of course it shouldn't be taken as a newline when it is part
320 of another character. The bug was that, for an unlimited repeat of . in
321 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
322 characters when looking for a newline.
323
324 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
325
326 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
327 in debug output.
328
329 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
330 long printing in the pcrecpp unittest when running under MinGW.
331
332 6. ESC_K was left out of the EBCDIC table.
333
334 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
335 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
336 limit also applies to "virtual nesting" when a pattern is recursive, and in
337 this case 1000 isn't so big. I have been able to remove this limit at the
338 expense of backing off one optimization in certain circumstances. Normally,
339 when pcre_exec() would call its internal match() function recursively and
340 immediately return the result unconditionally, it uses a "tail recursion"
341 feature to save stack. However, when a subpattern that can match an empty
342 string has an unlimited repetition quantifier, it no longer makes this
343 optimization. That gives it a stack frame in which to save the data for
344 checking that an empty string has been matched. Previously this was taken
345 from the 1000-entry workspace that had been reserved. So now there is no
346 explicit limit, but more stack is used.
347
348 8. Applied Daniel's patches to solve problems with the import/export magic
349 syntax that is required for Windows, and which was going wrong for the
350 pcreposix and pcrecpp parts of the library. These were overlooked when this
351 problem was solved for the main library.
352
353 9. There were some crude static tests to avoid integer overflow when computing
354 the size of patterns that contain repeated groups with explicit upper
355 limits. As the maximum quantifier is 65535, the maximum group length was
356 set at 30,000 so that the product of these two numbers did not overflow a
357 32-bit integer. However, it turns out that people want to use groups that
358 are longer than 30,000 bytes (though not repeat them that many times).
359 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
360 made it possible to implement the integer overflow checks in a much more
361 dynamic way, which I have now done. The artificial limitation on group
362 length has been removed - we now have only the limit on the total length of
363 the compiled pattern, which depends on the LINK_SIZE setting.
364
365 10. Fixed a bug in the documentation for get/copy named substring when
366 duplicate names are permitted. If none of the named substrings are set, the
367 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
368 empty string.
369
370 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
371 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
372 because the ] is interpreted as the first data character and the
373 terminating ] is not found. PCRE has been made compatible with Perl in this
374 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
375 cause memory overwriting.
376
377 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
378 string has been matched (to stop an infinite loop). It was not recognizing
379 a conditional subpattern that could match an empty string if that
380 subpattern was within another subpattern. For example, it looped when
381 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
382 condition was not nested. This bug has been fixed.
383
384 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
385 past the start of the subject in the presence of bytes with the top bit
386 set, for example "\x8aBCD".
387
388 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
389 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
390
391 14. Optimized (?!) to (*FAIL).
392
393 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
394 This restricts code points to be within the range 0 to 0x10FFFF, excluding
395 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
396 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
397 does: it's just the validity check that is more restrictive.
398
399 16. Inserted checks for integer overflows during escape sequence (backslash)
400 processing, and also fixed erroneous offset values for syntax errors during
401 backslash processing.
402
403 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
404 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
405
406 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
407 caused an overrun.
408
409 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
410 something other than just ASCII characters) inside a group that had an
411 unlimited repeat caused a loop at compile time (while checking to see
412 whether the group could match an empty string).
413
414 20. Debugging a pattern containing \p or \P could cause a crash. For example,
415 [\P{Any}] did so. (Error in the code for printing property names.)
416
417 21. An orphan \E inside a character class could cause a crash.
418
419 22. A repeated capturing bracket such as (A)? could cause a wild memory
420 reference during compilation.
421
422 23. There are several functions in pcre_compile() that scan along a compiled
423 expression for various reasons (e.g. to see if it's fixed length for look
424 behind). There were bugs in these functions when a repeated \p or \P was
425 present in the pattern. These operators have additional parameters compared
426 with \d, etc, and these were not being taken into account when moving along
427 the compiled data. Specifically:
428
429 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
430 length.
431
432 (b) An item such as \pL+ within a repeated group could cause crashes or
433 loops.
434
435 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
436 "reference to non-existent subpattern" error.
437
438 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
439
440 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
441 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
442
443 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
444
445 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
446 character were causing crashes (broken optimization).
447
448 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
449 \p or \P) caused a compile-time loop.
450
451 28. More problems have arisen in unanchored patterns when CRLF is a valid line
452 break. For example, the unstudied pattern [\r\n]A does not match the string
453 "\r\nA" because change 7.0/46 below moves the current point on by two
454 characters after failing to match at the start. However, the pattern \nA
455 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
456 the same is true. There doesn't seem any very clean way out of this, but
457 what I have chosen to do makes the common cases work: PCRE now takes note
458 of whether there can be an explicit match for \r or \n anywhere in the
459 pattern, and if so, 7.0/46 no longer applies. As part of this change,
460 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
461 pattern has explicit CR or LF references.
462
463 29. Added (*CR) etc for changing newline setting at start of pattern.
464
465
466 Version 7.2 19-Jun-07
467 ---------------------
468
469 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
470 which is apparently normally available under Windows.
471
472 2. Re-jig the pcregrep tests with different newline settings in an attempt
473 to make them independent of the local environment's newline setting.
474
475 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
476
477 4. Some of the "internals" tests were previously cut out when the link size
478 was not 2, because the output contained actual offsets. The recent new
479 "Z" feature of pcretest means that these can be cut out, making the tests
480 usable with all link sizes.
481
482 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
483 stack recursion. This gives a massive performance boost under BSD, but just
484 a small improvement under Linux. However, it saves one field in the frame
485 in all cases.
486
487 6. Added more features from the forthcoming Perl 5.10:
488
489 (a) (?-n) (where n is a string of digits) is a relative subroutine or
490 recursion call. It refers to the nth most recently opened parentheses.
491
492 (b) (?+n) is also a relative subroutine call; it refers to the nth next
493 to be opened parentheses.
494
495 (c) Conditions that refer to capturing parentheses can be specified
496 relatively, for example, (?(-2)... or (?(+3)...
497
498 (d) \K resets the start of the current match so that everything before
499 is not part of it.
500
501 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
502
503 (f) \g{name} is another synonym - part of Perl 5.10's unification of
504 reference syntax.
505
506 (g) (?| introduces a group in which the numbering of parentheses in each
507 alternative starts with the same number.
508
509 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
510
511 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
512 PCRE_INFO_JCHANGED.
513
514 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
515 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
516 for detecting groups that can match an empty string.
517
518 9. A pattern with a very large number of alternatives (more than several
519 hundred) was running out of internal workspace during the pre-compile
520 phase, where pcre_compile() figures out how much memory will be needed. A
521 bit of new cunning has reduced the workspace needed for groups with
522 alternatives. The 1000-alternative test pattern now uses 12 bytes of
523 workspace instead of running out of the 4096 that are available.
524
525 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
526
527 11. Applied patch from Google to remove an optimization that didn't quite work.
528 The report of the bug said:
529
530 pcrecpp::RE("a*").FullMatch("aaa") matches, while
531 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
532 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
533
534 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
535 it matched the wrong number of bytes.
536
537
538 Version 7.1 24-Apr-07
539 ---------------------
540
541 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
542 that is more "standard", making use of automake and other Autotools. There
543 is some re-arrangement of the files and adjustment of comments consequent
544 on this.
545
546 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
547 for recursive directory scanning broke on some systems because the files
548 are not scanned in any specific order and on different systems the order
549 was different. A call to "sort" has been inserted into RunGrepTest for the
550 approprate test as a short-term fix. In the longer term there may be an
551 alternative.
552
553 3. I had an email from Eric Raymond about problems translating some of PCRE's
554 man pages to HTML (despite the fact that I distribute HTML pages, some
555 people do their own conversions for various reasons). The problems
556 concerned the use of low-level troff macros .br and .in. I have therefore
557 removed all such uses from the man pages (some were redundant, some could
558 be replaced by .nf/.fi pairs). The 132html script that I use to generate
559 HTML has been updated to handle .nf/.fi and to complain if it encounters
560 .br or .in.
561
562 4. Updated comments in configure.ac that get placed in config.h.in and also
563 arranged for config.h to be included in the distribution, with the name
564 config.h.generic, for the benefit of those who have to compile without
565 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
566
567 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
568 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
569 makevp.bat for the current PCRE, using the additional files
570 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
571
572 6. A Windows user reported a minor discrepancy with test 2, which turned out
573 to be caused by a trailing space on an input line that had got lost in his
574 copy. The trailing space was an accident, so I've just removed it.
575
576 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
577 that is needed.
578
579 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
580 as "const" (a) because they are and (b) because it helps the PHP
581 maintainers who have recently made a script to detect big data structures
582 in the php code that should be moved to the .rodata section. I remembered
583 to update Builducptable as well, so it won't revert if ucptable.h is ever
584 re-created.
585
586 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
587 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
588 order to be able to cut out the UTF-8 tables in the latter when UTF-8
589 support is not required. This saves 1.5-2K of code, which is important in
590 some applications.
591
592 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
593 so as not to refer to the tables, even though these functions will never be
594 called when UTF-8 support is disabled. Otherwise there are problems with a
595 shared library.
596
597 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
598
599 (a) It was defining its arguments as char * instead of void *.
600
601 (b) It was assuming that all moves were upwards in memory; this was true
602 a long time ago when I wrote it, but is no longer the case.
603
604 The emulated memove() is provided for those environments that have neither
605 memmove() nor bcopy(). I didn't think anyone used it these days, but that
606 is clearly not the case, as these two bugs were recently reported.
607
608 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
609 and Detrail to create the HTML documentation, the .txt form of the man
610 pages, and it removes trailing spaces from listed files. It also creates
611 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
612 case, it wraps all the #defines with #ifndefs. This script should be run
613 before "make dist".
614
615 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
616 with Unicode property support.
617
618 (a) For a maximizing quantifier, if the two different cases of the
619 character were of different lengths in their UTF-8 codings (there are
620 some cases like this - I found 11), and the matching function had to
621 back up over a mixture of the two cases, it incorrectly assumed they
622 were both the same length.
623
624 (b) When PCRE was configured to use the heap rather than the stack for
625 recursion during matching, it was not correctly preserving the data for
626 the other case of a UTF-8 character when checking ahead for a match
627 while processing a minimizing repeat. If the check also involved
628 matching a wide character, but failed, corruption could cause an
629 erroneous result when trying to check for a repeat of the original
630 character.
631
632 13. Some tidying changes to the testing mechanism:
633
634 (a) The RunTest script now detects the internal link size and whether there
635 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
636 values substituted by "configure". (The RunGrepTest script already did
637 this for UTF-8.) The configure.ac script no longer substitutes the
638 relevant variables.
639
640 (b) The debugging options /B and /D in pcretest show the compiled bytecode
641 with length and offset values. This means that the output is different
642 for different internal link sizes. Test 2 is skipped for link sizes
643 other than 2 because of this, bypassing the problem. Unfortunately,
644 there was also a test in test 3 (the locale tests) that used /B and
645 failed for link sizes other than 2. Rather than cut the whole test out,
646 I have added a new /Z option to pcretest that replaces the length and
647 offset values with spaces. This is now used to make test 3 independent
648 of link size. (Test 2 will be tidied up later.)
649
650 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
651 segmentation fault instead of returning the appropriate error message.
652
653 15. In multiline mode when the newline sequence was set to "any", the pattern
654 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
655 This doesn't seem right; it now treats the CRLF combination as the line
656 ending, and so does not match in that case. It's only a pattern such as ^$
657 that would hit this one: something like ^ABC$ would have failed after \r
658 and then tried again after \r\n.
659
660 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
661 in an attempt to make files that differ only in their line terminators
662 compare equal. This works on Linux.
663
664 17. Under certain error circumstances pcregrep might try to free random memory
665 as it exited. This is now fixed, thanks to valgrind.
666
667 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
668 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
669 was because its rules for how to advance for /g after matching an empty
670 string at the end of a line did not allow for this case. They now check for
671 it specially.
672
673 20. pcretest is supposed to handle patterns and data of any length, by
674 extending its buffers when necessary. It was getting this wrong when the
675 buffer for a data line had to be extended.
676
677 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
678 CRLF as a newline sequence.
679
680 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
681 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
682 I have nevertheless tidied it up.
683
684 23. Added some casts to kill warnings from HP-UX ia64 compiler.
685
686 24. Added a man page for pcre-config.
687
688
689 Version 7.0 19-Dec-06
690 ---------------------
691
692 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
693 moving to gcc 4.1.1.
694
695 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
696 sys/time.h, which is documented as needed for this function. It doesn't
697 seem to matter on Linux, but it showed up on some releases of OS X.
698
699 3. It seems that there are systems where bytes whose values are greater than
700 127 match isprint() in the "C" locale. The "C" locale should be the
701 default when a C program starts up. In most systems, only ASCII printing
702 characters match isprint(). This difference caused the output from pcretest
703 to vary, making some of the tests fail. I have changed pcretest so that:
704
705 (a) When it is outputting text in the compiled version of a pattern, bytes
706 other than 32-126 are always shown as hex escapes.
707
708 (b) When it is outputting text that is a matched part of a subject string,
709 it does the same, unless a different locale has been set for the match
710 (using the /L modifier). In this case, it uses isprint() to decide.
711
712 4. Fixed a major bug that caused incorrect computation of the amount of memory
713 required for a compiled pattern when options that changed within the
714 pattern affected the logic of the preliminary scan that determines the
715 length. The relevant options are -x, and -i in UTF-8 mode. The result was
716 that the computed length was too small. The symptoms of this bug were
717 either the PCRE error "internal error: code overflow" from pcre_compile(),
718 or a glibc crash with a message such as "pcretest: free(): invalid next
719 size (fast)". Examples of patterns that provoked this bug (shown in
720 pcretest format) are:
721
722 /(?-x: )/x
723 /(?x)(?-x: \s*#\s*)/
724 /((?i)[\x{c0}])/8
725 /(?i:[\x{c0}])/8
726
727 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
728 is now done differently.
729
730 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
731 wrapper classes; (b) implement a new function in the C++ scanner that is
732 more efficient than the old way of doing things because it avoids levels of
733 recursion in the regex matching; (c) add a paragraph to the documentation
734 for the FullMatch() function.
735
736 6. The escape sequence \n was being treated as whatever was defined as
737 "newline". Not only was this contrary to the documentation, which states
738 that \n is character 10 (hex 0A), but it also went horribly wrong when
739 "newline" was defined as CRLF. This has been fixed.
740
741 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
742 was being set to -1 for the "end of line" case (supposedly a value that no
743 character can have). Though this value is never used (the check for end of
744 line is "zero bytes in current character"), it caused compiler complaints.
745 I've changed it to 0xffffffff.
746
747 8. In pcre_version.c, the version string was being built by a sequence of
748 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
749 string (as it is for production releases) called a macro with an empty
750 argument. The C standard says the result of this is undefined. The gcc
751 compiler treats it as an empty string (which was what was wanted) but it is
752 reported that Visual C gives an error. The source has been hacked around to
753 avoid this problem.
754
755 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
756 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
757 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
758 of them did).
759
760 10. Originally, pcretest opened its input and output without "b"; then I was
761 told that "b" was needed in some environments, so it was added for release
762 5.0 to both the input and output. (It makes no difference on Unix-like
763 systems.) Later I was told that it is wrong for the input on Windows. I've
764 now abstracted the modes into two macros, to make it easier to fiddle with
765 them, and removed "b" from the input mode under Windows.
766
767 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
768
769 12. Added -help and --help to pcretest as an official way of being reminded
770 of the options.
771
772 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
773 and pcrecpp.cc because they annoy compilers at high warning levels.
774
775 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
776
777 15. Fixed an occurrence of == in configure.ac that should have been = (shell
778 scripts are not C programs :-) and which was not noticed because it works
779 on Linux.
780
781 16. pcretest is supposed to handle any length of pattern and data line (as one
782 line or as a continued sequence of lines) by extending its input buffer if
783 necessary. This feature was broken for very long pattern lines, leading to
784 a string of junk being passed to pcre_compile() if the pattern was longer
785 than about 50K.
786
787 17. I have done a major re-factoring of the way pcre_compile() computes the
788 amount of memory needed for a compiled pattern. Previously, there was code
789 that made a preliminary scan of the pattern in order to do this. That was
790 OK when PCRE was new, but as the facilities have expanded, it has become
791 harder and harder to keep it in step with the real compile phase, and there
792 have been a number of bugs (see for example, 4 above). I have now found a
793 cunning way of running the real compile function in a "fake" mode that
794 enables it to compute how much memory it would need, while actually only
795 ever using a few hundred bytes of working memory and without too many
796 tests of the mode. This should make future maintenance and development
797 easier. A side effect of this work is that the limit of 200 on the nesting
798 depth of parentheses has been removed (though this was never a serious
799 limitation, I suspect). However, there is a downside: pcre_compile() now
800 runs more slowly than before (30% or more, depending on the pattern). I
801 hope this isn't a big issue. There is no effect on runtime performance.
802
803 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
804 newline (only possible for the last line of a file) and it was a
805 pattern that set a locale (followed by /Lsomething), pcretest crashed.
806
807 19. Added additional timing features to pcretest. (1) The -tm option now times
808 matching only, not compiling. (2) Both -t and -tm can be followed, as a
809 separate command line item, by a number that specifies the number of
810 repeats to use when timing. The default is 50000; this gives better
811 precision, but takes uncomfortably long for very large patterns.
812
813 20. Extended pcre_study() to be more clever in cases where a branch of a
814 subpattern has no definite first character. For example, (a*|b*)[cd] would
815 previously give no result from pcre_study(). Now it recognizes that the
816 first character must be a, b, c, or d.
817
818 21. There was an incorrect error "recursive call could loop indefinitely" if
819 a subpattern (or the entire pattern) that was being tested for matching an
820 empty string contained only one non-empty item after a nested subpattern.
821 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
822 incorrectly, because the \d was being skipped in the check.
823
824 22. The pcretest program now has a new pattern option /B and a command line
825 option -b, which is equivalent to adding /B to every pattern. This causes
826 it to show the compiled bytecode, without the additional information that
827 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
828 is the same as /B/I).
829
830 23. A new optimization is now able automatically to treat some sequences such
831 as a*b as a*+b. More specifically, if something simple (such as a character
832 or a simple class like \d) has an unlimited quantifier, and is followed by
833 something that cannot possibly match the quantified thing, the quantifier
834 is automatically "possessified".
835
836 24. A recursive reference to a subpattern whose number was greater than 39
837 went wrong under certain circumstances in UTF-8 mode. This bug could also
838 have affected the operation of pcre_study().
839
840 25. Realized that a little bit of performance could be had by replacing
841 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
842
843 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
844
845 27. Possessive quantifiers such as a++ were previously implemented by turning
846 them into atomic groups such as ($>a+). Now they have their own opcodes,
847 which improves performance. This includes the automatically created ones
848 from 23 above.
849
850 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
851 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
852 the first matched character to be a colon. This applied both to named and
853 numbered groups.
854
855 29. The ucpinternal.h header file was missing its idempotency #ifdef.
856
857 30. I was sent a "project" file called libpcre.a.dev which I understand makes
858 building PCRE on Windows easier, so I have included it in the distribution.
859
860 31. There is now a check in pcretest against a ridiculously large number being
861 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
862 loop, the loop is abandoned.
863
864 32. Forward references to subpatterns in conditions such as (?(2)...) where
865 subpattern 2 is defined later cause pcre_compile() to search forwards in
866 the pattern for the relevant set of parentheses. This search went wrong
867 when there were unescaped parentheses in a character class, parentheses
868 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
869
870 33. "Subroutine" calls and backreferences were previously restricted to
871 referencing subpatterns earlier in the regex. This restriction has now
872 been removed.
873
874 34. Added a number of extra features that are going to be in Perl 5.10. On the
875 whole, these are just syntactic alternatives for features that PCRE had
876 previously implemented using the Python syntax or my own invention. The
877 other formats are all retained for compatibility.
878
879 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
880 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
881 also .NET compatible.
882
883 (b) A recursion or subroutine call to a named group can now be defined as
884 (?&name) as well as (?P>name).
885
886 (c) A backreference to a named group can now be defined as \k<name> or
887 \k'name' as well as (?P=name). The new forms, as well as being in Perl
888 5.10, are also .NET compatible.
889
890 (d) A conditional reference to a named group can now use the syntax
891 (?(<name>) or (?('name') as well as (?(name).
892
893 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
894 groups (named and numbered) that are never evaluated inline, but can be
895 called as "subroutines" from elsewhere. In effect, the DEFINE condition
896 is always false. There may be only one alternative in such a group.
897
898 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
899 as the simple (?(R). The condition is true only if the most recent
900 recursion is that of the given number or name. It does not search out
901 through the entire recursion stack.
902
903 (g) The escape \gN or \g{N} has been added, where N is a positive or
904 negative number, specifying an absolute or relative reference.
905
906 35. Tidied to get rid of some further signed/unsigned compiler warnings and
907 some "unreachable code" warnings.
908
909 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
910 things, this adds five new scripts.
911
912 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
913 There were also incompatibilities regarding the handling of \Q..\E inside
914 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
915 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
916
917 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
918 matches an empty string, and forcibly breaks the loop. There were bugs in
919 this code in non-simple cases. For a pattern such as ^(a()*)* matched
920 against aaaa the result was just "a" rather than "aaaa", for example. Two
921 separate and independent bugs (that affected different cases) have been
922 fixed.
923
924 39. Refactored the code to abolish the use of different opcodes for small
925 capturing bracket numbers. This is a tidy that I avoided doing when I
926 removed the limit on the number of capturing brackets for 3.5 back in 2001.
927 The new approach is not only tidier, it makes it possible to reduce the
928 memory needed to fix the previous bug (38).
929
930 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
931 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
932 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
933 mode.
934
935 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
936 report.
937
938 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
939 copy construction and assignment in the C++ wrapper.
940
941 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
942 couple of bugs that could have given wrong results in the "--newline=crlf"
943 case.
944
945 44. Added a number of casts and did some reorganization of signed/unsigned int
946 variables following suggestions from Dair Grant. Also renamed the variable
947 "this" as "item" because it is a C++ keyword.
948
949 45. Arranged for dftables to add
950
951 #include "pcre_internal.h"
952
953 to pcre_chartables.c because without it, gcc 4.x may remove the array
954 definition from the final binary if PCRE is built into a static library and
955 dead code stripping is activated.
956
957 46. For an unanchored pattern, if a match attempt fails at the start of a
958 newline sequence, and the newline setting is CRLF or ANY, and the next two
959 characters are CRLF, advance by two characters instead of one.
960
961
962 Version 6.7 04-Jul-06
963 ---------------------
964
965 1. In order to handle tests when input lines are enormously long, pcretest has
966 been re-factored so that it automatically extends its buffers when
967 necessary. The code is crude, but this _is_ just a test program. The
968 default size has been increased from 32K to 50K.
969
970 2. The code in pcre_study() was using the value of the re argument before
971 testing it for NULL. (Of course, in any sensible call of the function, it
972 won't be NULL.)
973
974 3. The memmove() emulation function in pcre_internal.h, which is used on
975 systems that lack both memmove() and bcopy() - that is, hardly ever -
976 was missing a "static" storage class specifier.
977
978 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
979 containing an extended class (one that cannot be represented by a bitmap
980 because it contains high-valued characters or Unicode property items, e.g.
981 [\pZ]). Almost always one would set UTF-8 mode when processing such a
982 pattern, but PCRE should not loop if you do not (it no longer does).
983 [Detail: two cases were found: (a) a repeated subpattern containing an
984 extended class; (b) a recursive reference to a subpattern that followed a
985 previous extended class. It wasn't skipping over the extended class
986 correctly when UTF-8 mode was not set.]
987
988 5. A negated single-character class was not being recognized as fixed-length
989 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
990 compile error "lookbehind assertion is not fixed length".
991
992 6. The RunPerlTest auxiliary script was showing an unexpected difference
993 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
994 write a Perl script that can interpret lines of an input file either as
995 byte characters or as UTF-8, which is what "perltest" was being required to
996 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
997 can't do is switch easily at run time between having the "use utf8;" pragma
998 or not. In the end, I fudged it by using the RunPerlTest script to insert
999 "use utf8;" explicitly for the UTF-8 tests.
1000
1001 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1002 the end of the subject string, contrary to the documentation and to what
1003 Perl does. This was true of both matching functions. Now it matches only at
1004 the start of the subject and immediately after *internal* newlines.
1005
1006 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1007 a pointer to an int instead of a pointer to an unsigned long int. This
1008 caused problems on 64-bit systems.
1009
1010 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1011 instance of the 'standard' template library not being so standard".
1012
1013 10. There was no check on the number of named subpatterns nor the maximum
1014 length of a subpattern name. The product of these values is used to compute
1015 the size of the memory block for a compiled pattern. By supplying a very
1016 long subpattern name and a large number of named subpatterns, the size
1017 computation could be caused to overflow. This is now prevented by limiting
1018 the length of names to 32 characters, and the number of named subpatterns
1019 to 10,000.
1020
1021 11. Subpatterns that are repeated with specific counts have to be replicated in
1022 the compiled pattern. The size of memory for this was computed from the
1023 length of the subpattern and the repeat count. The latter is limited to
1024 65535, but there was no limit on the former, meaning that integer overflow
1025 could in principle occur. The compiled length of a repeated subpattern is
1026 now limited to 30,000 bytes in order to prevent this.
1027
1028 12. Added the optional facility to have named substrings with the same name.
1029
1030 13. Added the ability to use a named substring as a condition, using the
1031 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1032 are numbers (not recommended). Forward references are permitted.
1033
1034 14. Added forward references in named backreferences (if you see what I mean).
1035
1036 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1037 pattern could run off the end of the subject. For example, the pattern
1038 "(?s)(.{1,5})"8 did this with the subject "ab".
1039
1040 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1041 PCRE_CASELESS was set when matching characters that were quantified with ?
1042 or *.
1043
1044 17. A character class other than a single negated character that had a minimum
1045 but no maximum quantifier - for example [ab]{6,} - was not handled
1046 correctly by pce_dfa_exec(). It would match only one character.
1047
1048 18. A valid (though odd) pattern that looked like a POSIX character
1049 class but used an invalid character after [ (for example [[,abc,]]) caused
1050 pcre_compile() to give the error "Failed: internal error: code overflow" or
1051 in some cases to crash with a glibc free() error. This could even happen if
1052 the pattern terminated after [[ but there just happened to be a sequence of
1053 letters, a binary zero, and a closing ] in the memory that followed.
1054
1055 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1056 over the years. Originally (before any Unicode support), just the bottom 8
1057 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1058 output from "man perlunicode" includes this:
1059
1060 The regular expression compiler produces polymorphic opcodes. That
1061 is, the pattern adapts to the data and automatically switches to
1062 the Unicode character scheme when presented with Unicode data--or
1063 instead uses a traditional byte scheme when presented with byte
1064 data.
1065
1066 Sadly, a wide octal escape does not cause a switch, and in a string with
1067 no other multibyte characters, these octal escapes are treated as before.
1068 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1069 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1070 Unicode string.
1071
1072 I have not perpetrated such confusion in PCRE. Up till now, it took just
1073 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1074 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1075 translate to the appropriate multibyte character.
1076
1077 29. Applied some refactoring to reduce the number of warnings from Microsoft
1078 and Borland compilers. This has included removing the fudge introduced
1079 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1080 a warning about an unused variable.
1081
1082 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1083 characters since release 4.0, because Perl (from release 5.004) does not.
1084 [Or at least, is documented not to: some releases seem to be in conflict
1085 with the documentation.] However, when a pattern was studied with
1086 pcre_study() and all its branches started with \s, PCRE still included VT
1087 as a possible starting character. Of course, this did no harm; it just
1088 caused an unnecessary match attempt.
1089
1090 22. Removed a now-redundant internal flag bit that recorded the fact that case
1091 dependency changed within the pattern. This was once needed for "required
1092 byte" processing, but is no longer used. This recovers a now-scarce options
1093 bit. Also moved the least significant internal flag bit to the most-
1094 significant bit of the word, which was not previously used (hangover from
1095 the days when it was an int rather than a uint) to free up another bit for
1096 the future.
1097
1098 23. Added support for CRLF line endings as well as CR and LF. As well as the
1099 default being selectable at build time, it can now be changed at runtime
1100 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1101 specify that it is scanning data with non-default line endings.
1102
1103 24. Changed the definition of CXXLINK to make it agree with the definition of
1104 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1105
1106 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1107 recursions. This makes a big different to stack usage for some patterns.
1108
1109 26. If a subpattern containing a named recursion or subroutine reference such
1110 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1111 the space required for the compiled pattern went wrong and gave too small a
1112 value. Depending on the environment, this could lead to "Failed: internal
1113 error: code overflow at offset 49" or "glibc detected double free or
1114 corruption" errors.
1115
1116 27. Applied patches from Google (a) to support the new newline modes and (b) to
1117 advance over multibyte UTF-8 characters in GlobalReplace.
1118
1119 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1120 difference for some implementation of PCRE in some Windows version.
1121
1122 29. Added some extra testing facilities to pcretest:
1123
1124 \q<number> in a data line sets the "match limit" value
1125 \Q<number> in a data line sets the "match recursion limt" value
1126 -S <number> sets the stack size, where <number> is in megabytes
1127
1128 The -S option isn't available for Windows.
1129
1130
1131 Version 6.6 06-Feb-06
1132 ---------------------
1133
1134 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1135 in pcreposix.h. I have copied the definition from pcre.h.
1136
1137 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1138 because pcre.h is no longer a built file.
1139
1140 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1141 not normally included in the compiled code.
1142
1143
1144 Version 6.5 01-Feb-06
1145 ---------------------
1146
1147 1. When using the partial match feature with pcre_dfa_exec(), it was not
1148 anchoring the second and subsequent partial matches at the new starting
1149 point. This could lead to incorrect results. For example, with the pattern
1150 /1234/, partially matching against "123" and then "a4" gave a match.
1151
1152 2. Changes to pcregrep:
1153
1154 (a) All non-match returns from pcre_exec() were being treated as failures
1155 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1156 error message is output. Some extra information is given for the
1157 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1158 probably the only errors that are likely to be caused by users (by
1159 specifying a regex that has nested indefinite repeats, for instance).
1160 If there are more than 20 of these errors, pcregrep is abandoned.
1161
1162 (b) A binary zero was treated as data while matching, but terminated the
1163 output line if it was written out. This has been fixed: binary zeroes
1164 are now no different to any other data bytes.
1165
1166 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1167 used to set a locale for matching. The --locale=xxxx long option has
1168 been added (no short equivalent) to specify a locale explicitly on the
1169 pcregrep command, overriding the environment variables.
1170
1171 (d) When -B was used with -n, some line numbers in the output were one less
1172 than they should have been.
1173
1174 (e) Added the -o (--only-matching) option.
1175
1176 (f) If -A or -C was used with -c (count only), some lines of context were
1177 accidentally printed for the final match.
1178
1179 (g) Added the -H (--with-filename) option.
1180
1181 (h) The combination of options -rh failed to suppress file names for files
1182 that were found from directory arguments.
1183
1184 (i) Added the -D (--devices) and -d (--directories) options.
1185
1186 (j) Added the -F (--fixed-strings) option.
1187
1188 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1189
1190 (l) Added the --colo(u)r option.
1191
1192 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1193 is not present by default.
1194
1195 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1196 items such as (?R) or (?1), when the recursion could match a number of
1197 alternatives. If it matched one of the alternatives, but subsequently,
1198 outside the recursion, there was a failure, the code tried to back up into
1199 the recursion. However, because of the way PCRE is implemented, this is not
1200 possible, and the result was an incorrect result from the match.
1201
1202 In order to prevent this happening, the specification of recursion has
1203 been changed so that all such subpatterns are automatically treated as
1204 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1205
1206 4. I had overlooked the fact that, in some locales, there are characters for
1207 which isalpha() is true but neither isupper() nor islower() are true. In
1208 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1209 and ordfeminine) are like this. This affected the treatment of \w and \W
1210 when they appeared in character classes, but not when they appeared outside
1211 a character class. The bit map for "word" characters is now created
1212 separately from the results of isalnum() instead of just taking it from the
1213 upper, lower, and digit maps. (Plus the underscore character, of course.)
1214
1215 5. The above bug also affected the handling of POSIX character classes such as
1216 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1217 permanent tables. Instead, the bit maps for such a class were previously
1218 created as the appropriate unions of the upper, lower, and digit bitmaps.
1219 Now they are created by subtraction from the [[:word:]] class, which has
1220 its own bitmap.
1221
1222 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1223 It is created by subtracting the vertical space characters (\x09, \x0a,
1224 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1225 subtraction was done in the overall bitmap for a character class, meaning
1226 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1227 be recognized. This bug has been fixed.
1228
1229 7. Patches from the folks at Google:
1230
1231 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1232 real life, but is still worth protecting against".
1233
1234 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1235 regular expressions".
1236
1237 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1238 have it.
1239
1240 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1241 "configure" and the latter not, in order to fix a problem somebody had
1242 with compiling the Arg class on HP-UX.
1243
1244 (e) Improve the error-handling of the C++ wrapper a little bit.
1245
1246 (f) New tests for checking recursion limiting.
1247
1248 8. The pcre_memmove() function, which is used only if the environment does not
1249 have a standard memmove() function (and is therefore rarely compiled),
1250 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1251 returning a result (though PCRE never actually uses the result).
1252
1253 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1254 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1255 returned instead of calling malloc() with an overflowing number that would
1256 most likely cause subsequent chaos.
1257
1258 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1259
1260 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1261 with this option is matched, the nmatch and pmatch options of regexec() are
1262 ignored.
1263
1264 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1265 provided in case anyone wants to the the POSIX interface with UTF-8
1266 strings.
1267
1268 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1269 C++ linking (needed for some HP-UX environments).
1270
1271 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1272 (unused parameter) and in the pcre_printint() function (omitted "default"
1273 switch label when the default is to do nothing).
1274
1275 15. Added some code to make it possible, when PCRE is compiled as a C++
1276 library, to replace subject pointers for pcre_exec() with a smart pointer
1277 class, thus making it possible to process discontinuous strings.
1278
1279 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1280 much the same function. They were added by different people who were trying
1281 to make PCRE easy to compile on non-Unix systems. It has been suggested
1282 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1283 for compiling on Windows systems. I have therefore replaced it with
1284 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1285 defaults to "extern" for C or "extern C" for C++, which works fine on
1286 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1287 SCOPE with something explicit in config.h. In addition:
1288
1289 (a) pcreposix.h still had just "extern" instead of either of these macros;
1290 I have replaced it with PCRE_DATA_SCOPE.
1291
1292 (b) Functions such as _pcre_xclass(), which are internal to the library,
1293 but external in the C sense, all had PCRE_EXPORT in their definitions.
1294 This is apparently wrong for the Windows case, so I have removed it.
1295 (It makes no difference on Unix-like systems.)
1296
1297 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1298 of recursive calls to match(). This is different to MATCH_LIMIT because
1299 that limits the total number of calls to match(), not all of which increase
1300 the depth of recursion. Limiting the recursion depth limits the amount of
1301 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1302 when PCRE is compiled, and changed at run time. A patch from Google adds
1303 this functionality to the C++ interface.
1304
1305 18. Changes to the handling of Unicode character properties:
1306
1307 (a) Updated the table to Unicode 4.1.0.
1308
1309 (b) Recognize characters that are not in the table as "Cn" (undefined).
1310
1311 (c) I revised the way the table is implemented to a much improved format
1312 which includes recognition of ranges. It now supports the ranges that
1313 are defined in UnicodeData.txt, and it also amalgamates other
1314 characters into ranges. This has reduced the number of entries in the
1315 table from around 16,000 to around 3,000, thus reducing its size
1316 considerably. I realized I did not need to use a tree structure after
1317 all - a binary chop search is just as efficient. Having reduced the
1318 number of entries, I extended their size from 6 bytes to 8 bytes to
1319 allow for more data.
1320
1321 (d) Added support for Unicode script names via properties such as \p{Han}.
1322
1323 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1324 matching that character.
1325
1326 20. When matching a repeated Unicode property with a minimum greater than zero,
1327 (for example \pL{2,}), PCRE could look past the end of the subject if it
1328 reached it while seeking the minimum number of characters. This could
1329 happen only if some of the characters were more than one byte long, because
1330 there is a check for at least the minimum number of bytes.
1331
1332 21. Refactored the implementation of \p and \P so as to be more general, to
1333 allow for more different types of property in future. This has changed the
1334 compiled form incompatibly. Anybody with saved compiled patterns that use
1335 \p or \P will have to recompile them.
1336
1337 22. Added "Any" and "L&" to the supported property types.
1338
1339 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1340 but give a compile time error if the value is greater than 0xff.
1341
1342 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1343 accidentally not being installed or uninstalled.
1344
1345 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1346 made were to insert the current release number. This seemed silly, because
1347 it made things harder for people building PCRE on systems that don't run
1348 "configure". I have turned pcre.h into a distributed file, no longer built
1349 by "configure", with the version identification directly included. There is
1350 no longer a pcre.h.in file.
1351
1352 However, this change necessitated a change to the pcre-config script as
1353 well. It is built from pcre-config.in, and one of the substitutions was the
1354 release number. I have updated configure.ac so that ./configure now finds
1355 the release number by grepping pcre.h.
1356
1357 26. Added the ability to run the tests under valgrind.
1358
1359
1360 Version 6.4 05-Sep-05
1361 ---------------------
1362
1363 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1364 "--" to be printed when multiple files were scanned, even when none of the
1365 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1366 consider it to be a bug, and have restored the previous behaviour.
1367
1368 2. A couple of code tidies to get rid of compiler warnings.
1369
1370 3. The pcretest program used to cheat by referring to symbols in the library
1371 whose names begin with _pcre_. These are internal symbols that are not
1372 really supposed to be visible externally, and in some environments it is
1373 possible to suppress them. The cheating is now confined to including
1374 certain files from the library's source, which is a bit cleaner.
1375
1376 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1377 file's purpose clearer.
1378
1379 5. Reorganized pcre_ucp_findchar().
1380
1381
1382 Version 6.3 15-Aug-05
1383 ---------------------
1384
1385 1. The file libpcre.pc.in did not have general read permission in the tarball.
1386
1387 2. There were some problems when building without C++ support:
1388
1389 (a) If C++ support was not built, "make install" and "make test" still
1390 tried to test it.
1391
1392 (b) There were problems when the value of CXX was explicitly set. Some
1393 changes have been made to try to fix these, and ...
1394
1395 (c) --disable-cpp can now be used to explicitly disable C++ support.
1396
1397 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1398 backslash in a target when C++ was disabled. This confuses some
1399 versions of "make", apparently. Using an intermediate variable solves
1400 this. (Same for CPP_LOBJ.)
1401
1402 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1403 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1404 necessary on certain architectures.
1405
1406 4. Added a setting of -export-symbols-regex to the link command to remove
1407 those symbols that are exported in the C sense, but actually are local
1408 within the library, and not documented. Their names all begin with
1409 "_pcre_". This is not a perfect job, because (a) we have to except some
1410 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1411 available (and never for static libraries). I have made a note to try to
1412 find a way round (a) in the future.
1413
1414
1415 Version 6.2 01-Aug-05
1416 ---------------------
1417
1418 1. There was no test for integer overflow of quantifier values. A construction
1419 such as {1111111111111111} would give undefined results. What is worse, if
1420 a minimum quantifier for a parenthesized subpattern overflowed and became
1421 negative, the calculation of the memory size went wrong. This could have
1422 led to memory overwriting.
1423
1424 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1425
1426 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1427 operating environments where this matters.
1428
1429 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1430 PCRE options from within the C++ wrapper.
1431
1432 5. Named capturing subpatterns were not being correctly counted when a pattern
1433 was compiled. This caused two problems: (a) If there were more than 100
1434 such subpatterns, the calculation of the memory needed for the whole
1435 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1436 back references of the form \12, where the number was greater than 9, were
1437 not recognized as back references, even though there were sufficient
1438 previous subpatterns.
1439
1440 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1441 versions of gcc, e.g. 2.95.4.
1442
1443
1444 Version 6.1 21-Jun-05
1445 ---------------------
1446
1447 1. There was one reference to the variable "posix" in pcretest.c that was not
1448 surrounded by "#if !defined NOPOSIX".
1449
1450 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1451 the cross-check on the old pcre_info() function, for the benefit of the
1452 cut-down version of PCRE that is currently imported into Exim.
1453
1454 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1455 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1456 patterns that start (?i) but I don't think that matters. The use of (?i) is
1457 just an example; this all applies to the other options as well.
1458
1459 4. Since libtool seems to echo the compile commands it is issuing, the output
1460 from "make" can be reduced a bit by putting "@" in front of each libtool
1461 compile command.
1462
1463 5. Patch from the folks at Google for configure.in to be a bit more thorough
1464 in checking for a suitable C++ installation before trying to compile the
1465 C++ stuff. This should fix a reported problem when a compiler was present,
1466 but no suitable headers.
1467
1468 6. The man pages all had just "PCRE" as their title. I have changed them to
1469 be the relevant file name. I have also arranged that these names are
1470 retained in the file doc/pcre.txt, which is a concatenation in text format
1471 of all the man pages except the little individual ones for each function.
1472
1473 7. The NON-UNIX-USE file had not been updated for the different set of source
1474 files that come with release 6. I also added a few comments about the C++
1475 wrapper.
1476
1477
1478 Version 6.0 07-Jun-05
1479 ---------------------
1480
1481 1. Some minor internal re-organization to help with my DFA experiments.
1482
1483 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1484 didn't matter for the library itself when fully configured, but did matter
1485 when compiling without UCP support, or within Exim, where the ucp files are
1486 not imported.
1487
1488 3. Refactoring of the library code to split up the various functions into
1489 different source modules. The addition of the new DFA matching code (see
1490 below) to a single monolithic source would have made it really too
1491 unwieldy, quite apart from causing all the code to be include in a
1492 statically linked application, when only some functions are used. This is
1493 relevant even without the DFA addition now that patterns can be compiled in
1494 one application and matched in another.
1495
1496 The downside of splitting up is that there have to be some external
1497 functions and data tables that are used internally in different modules of
1498 the library but which are not part of the API. These have all had their
1499 names changed to start with "_pcre_" so that they are unlikely to clash
1500 with other external names.
1501
1502 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1503 a different (DFA) algorithm. Although it is slower than the original
1504 function, it does have some advantages for certain types of matching
1505 problem.
1506
1507 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1508 including restarting after a partial match.
1509
1510 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1511 defined when compiling for Windows was sent to me. I have put it into the
1512 code, though I have no means of testing or verifying it.
1513
1514 7. Added the pcre_refcount() auxiliary function.
1515
1516 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1517 match before or at the first newline in the subject string. In pcretest,
1518 the /f option on a pattern can be used to set this.
1519
1520 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1521 would behave wrongly. This has been present in PCRE since release 4.0.
1522
1523 10. A number of changes to the pcregrep command:
1524
1525 (a) Refactored how -x works; insert ^(...)$ instead of setting
1526 PCRE_ANCHORED and checking the length, in preparation for adding
1527 something similar for -w.
1528
1529 (b) Added the -w (match as a word) option.
1530
1531 (c) Refactored the way lines are read and buffered so as to have more
1532 than one at a time available.
1533
1534 (d) Implemented a pcregrep test script.
1535
1536 (e) Added the -M (multiline match) option. This allows patterns to match
1537 over several lines of the subject. The buffering ensures that at least
1538 8K, or the rest of the document (whichever is the shorter) is available
1539 for matching (and similarly the previous 8K for lookbehind assertions).
1540
1541 (f) Changed the --help output so that it now says
1542
1543 -w, --word-regex(p)
1544
1545 instead of two lines, one with "regex" and the other with "regexp"
1546 because that confused at least one person since the short forms are the
1547 same. (This required a bit of code, as the output is generated
1548 automatically from a table. It wasn't just a text change.)
1549
1550 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1551 option but starts with a hyphen. Could be a pattern or a path name
1552 starting with a hyphen, for instance.
1553
1554 (h) "-" can be given as a file name to represent stdin.
1555
1556 (i) When file names are being printed, "(standard input)" is used for
1557 the standard input, for compatibility with GNU grep. Previously
1558 "<stdin>" was used.
1559
1560 (j) The option --label=xxx can be used to supply a name to be used for
1561 stdin when file names are being printed. There is no short form.
1562
1563 (k) Re-factored the options decoding logic because we are going to add
1564 two more options that take data. Such options can now be given in four
1565 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1566
1567 (l) Added the -A, -B, and -C options for requesting that lines of context
1568 around matches be printed.
1569
1570 (m) Added the -L option to print the names of files that do not contain
1571 any matching lines, that is, the complement of -l.
1572
1573 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1574 continue to scan other files.
1575
1576 (o) The -s option was incorrectly implemented. For compatibility with other
1577 greps, it now suppresses the error message for a non-existent or non-
1578 accessible file (but not the return code). There is a new option called
1579 -q that suppresses the output of matching lines, which was what -s was
1580 previously doing.
1581
1582 (p) Added --include and --exclude options to specify files for inclusion
1583 and exclusion when recursing.
1584
1585 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1586 Hopefully, it now does.
1587
1588 12. Missing cast in pcre_study().
1589
1590 13. Added an "uninstall" target to the makefile.
1591
1592 14. Replaced "extern" in the function prototypes in Makefile.in with
1593 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1594 world, but is set differently for Windows.
1595
1596 15. Added a second compiling function called pcre_compile2(). The only
1597 difference is that it has an extra argument, which is a pointer to an
1598 integer error code. When there is a compile-time failure, this is set
1599 non-zero, in addition to the error test pointer being set to point to an
1600 error message. The new argument may be NULL if no error number is required
1601 (but then you may as well call pcre_compile(), which is now just a
1602 wrapper). This facility is provided because some applications need a
1603 numeric error indication, but it has also enabled me to tidy up the way
1604 compile-time errors are handled in the POSIX wrapper.
1605
1606 16. Added VPATH=.libs to the makefile; this should help when building with one
1607 prefix path and installing with another. (Or so I'm told by someone who
1608 knows more about this stuff than I do.)
1609
1610 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1611 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1612 match everything, including newlines. This is not POSIX-compatible, but
1613 somebody wanted the feature. From pcretest it can be activated by using
1614 both the P and the s flags.
1615
1616 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1617
1618 19. libpcre.pc was being incorrectly installed as executable.
1619
1620 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1621 it now also looks for '\r' so that it will work unmodified on Windows.
1622
1623 21. Added Google's contributed C++ wrapper to the distribution.
1624
1625 22. Added some untidy missing memory free() calls in pcretest, to keep
1626 Electric Fence happy when testing.
1627
1628
1629
1630 Version 5.0 13-Sep-04
1631 ---------------------
1632
1633 1. Internal change: literal characters are no longer packed up into items
1634 containing multiple characters in a single byte-string. Each character
1635 is now matched using a separate opcode. However, there may be more than one
1636 byte in the character in UTF-8 mode.
1637
1638 2. The pcre_callout_block structure has two new fields: pattern_position and
1639 next_item_length. These contain the offset in the pattern to the next match
1640 item, and its length, respectively.
1641
1642 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1643 insertion of callouts before each pattern item. Added the /C option to
1644 pcretest to make use of this.
1645
1646 4. On the advice of a Windows user, the lines
1647
1648 #if defined(_WIN32) || defined(WIN32)
1649 _setmode( _fileno( stdout ), 0x8000 );
1650 #endif /* defined(_WIN32) || defined(WIN32) */
1651
1652 have been added to the source of pcretest. This apparently does useful
1653 magic in relation to line terminators.
1654
1655 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1656 for the benefit of those environments where the "b" makes a difference.
1657
1658 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1659 to know about it. I have put a hack into configure.in that adds in code
1660 to set GCC=yes if CC=icc. This seems to end up at a point in the
1661 generated configure script that is early enough to affect the setting of
1662 compiler options, which is what is needed, but I have no means of testing
1663 whether it really works. (The user who reported this had patched the
1664 generated configure script, which of course I cannot do.)
1665
1666 LATER: After change 22 below (new libtool files), the configure script
1667 seems to know about icc (and also ecc). Therefore, I have commented out
1668 this hack in configure.in.
1669
1670 7. Added support for pkg-config (2 patches were sent in).
1671
1672 8. Negated POSIX character classes that used a combination of internal tables
1673 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1674 [[:^ascii]]. Typically, they would match almost any characters. The other
1675 POSIX classes were not broken in this way.
1676
1677 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1678 to find the match, as PCRE was deluded into thinking that the match had to
1679 start at the start point or following a newline. The same bug applied to
1680 patterns with negative forward assertions or any backward assertions
1681 preceding ".*" at the start, unless the pattern required a fixed first
1682 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1683
1684 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1685 starting at the last subject character, bytes beyond the end of the subject
1686 string were read.
1687
1688 11. Renamed the variable "class" as "classbits" to make life easier for C++
1689 users. (Previously there was a macro definition, but it apparently wasn't
1690 enough.)
1691
1692 12. Added the new field "tables" to the extra data so that tables can be passed
1693 in at exec time, or the internal tables can be re-selected. This allows
1694 a compiled regex to be saved and re-used at a later time by a different
1695 program that might have everything at different addresses.
1696
1697 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1698 -R library as well as a -L library.
1699
1700 14. The debugging options of pcretest (-d on the command line or D on a
1701 pattern) showed incorrect output for anything following an extended class
1702 that contained multibyte characters and which was followed by a quantifier.
1703
1704 15. Added optional support for general category Unicode character properties
1705 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1706 support. It adds about 90K to the size of the library. The meanings of the
1707 inbuilt class escapes such as \d and \s have NOT been changed.
1708
1709 16. Updated pcredemo.c to include calls to free() to release the memory for the
1710 compiled pattern.
1711
1712 17. The generated file chartables.c was being created in the source directory
1713 instead of in the building directory. This caused the build to fail if the
1714 source directory was different from the building directory, and was
1715 read-only.
1716
1717 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1718 file. No doubt somebody will tell me if they don't make sense... Also added
1719 Dan Mooney's comments about building on OpenVMS.
1720
1721 19. Added support for partial matching via the PCRE_PARTIAL option for
1722 pcre_exec() and the \P data escape in pcretest.
1723
1724 20. Extended pcretest with 3 new pattern features:
1725
1726 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1727 write the compiled pattern to the file whose name is "rest-of-line".
1728 This is a straight binary dump of the data, with the saved pointer to
1729 the character tables forced to be NULL. The study data, if any, is
1730 written too. After writing, pcretest reads a new pattern.
1731
1732 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1733 compiled pattern from the given file. There must not be any
1734 occurrences of "<" in the file name (pretty unlikely); if there are,
1735 pcretest will instead treat the initial "<" as a pattern delimiter.
1736 After reading in the pattern, pcretest goes on to read data lines as
1737 usual.
1738
1739 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1740 and 16-bit fields in a compiled pattern, to simulate a pattern that
1741 was compiled on a host of opposite endianness.
1742
1743 21. The pcre-exec() function can now cope with patterns that were compiled on
1744 hosts of opposite endianness, with this restriction:
1745
1746 As for any compiled expression that is saved and used later, the tables
1747 pointer field cannot be preserved; the extra_data field in the arguments
1748 to pcre_exec() should be used to pass in a tables address if a value
1749 other than the default internal tables were used at compile time.
1750
1751 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1752 now diagnosed as an error. Previously, most of the time, a negative number
1753 would have been treated as zero, but if in addition "ovector" was passed as
1754 NULL, a crash could occur.
1755
1756 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1757 new versions from the libtool 1.5 distribution (the last one is a copy of
1758 a file called libtool.m4). This seems to have fixed the need to patch
1759 "configure" to support Darwin 1.3 (which I used to do). However, I still
1760 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1761 workstation).
1762
1763 24. Changed the PCRE licence to be the more standard "BSD" licence.
1764
1765
1766 Version 4.5 01-Dec-03
1767 ---------------------
1768
1769 1. There has been some re-arrangement of the code for the match() function so
1770 that it can be compiled in a version that does not call itself recursively.
1771 Instead, it keeps those local variables that need separate instances for
1772 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1773 needs to "recurse". Keeping track of where control must go is done by means
1774 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1775 hide most of the details from the main code, and operates only if
1776 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1777 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1778 operating.
1779
1780 To make it easier for callers to provide specially tailored get/free
1781 functions for this usage, two new functions, pcre_stack_malloc, and
1782 pcre_stack_free, are used. They are always called in strict stacking order,
1783 and the size of block requested is always the same.
1784
1785 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1786 PCRE has been compiled to use the stack or the heap for recursion. The
1787 -C option of pcretest uses this to show which version is compiled.
1788
1789 A new data escape \S, is added to pcretest; it causes the amounts of store
1790 obtained and freed by both kinds of malloc/free at match time to be added
1791 to the output.
1792
1793 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1794 what's available on my current Linux desktop machine.
1795
1796 3. When matching a UTF-8 string, the test for a valid string at the start has
1797 been extended. If start_offset is not zero, PCRE now checks that it points
1798 to a byte that is the start of a UTF-8 character. If not, it returns
1799 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1800 this is necessary because there may be backward assertions in the pattern.
1801 When matching the same subject several times, it may save resources to use
1802 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1803
1804 4. The code for checking the validity of UTF-8 strings has been tightened so
1805 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1806 containing "overlong sequences".
1807
1808 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1809 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1810 should have been "&", but it just so happened that all the cases this let
1811 through by mistake were picked up later in the function.
1812
1813 6. I had used a variable called "isblank" - this is a C99 function, causing
1814 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1815
1816 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1817 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1818 so the version is shown; (c) stop "make test" echoing "./RunTest".
1819
1820 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1821
1822 9. The prototype for memmove() for systems that don't have it was using
1823 size_t, but the inclusion of the header that defines size_t was later. I've
1824 moved the #includes for the C headers earlier to avoid this.
1825
1826 10. Added some adjustments to the code to make it easier to compiler on certain
1827 special systems:
1828
1829 (a) Some "const" qualifiers were missing.
1830 (b) Added the macro EXPORT before all exported functions; by default this
1831 is defined to be empty.
1832 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1833 that it reads its output file name as an argument instead of writing
1834 to the standard output and assuming this can be redirected.
1835
1836 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1837 class containing characters with values greater than 255, PCRE compilation
1838 went into a loop.
1839
1840 12. A recursive reference to a subpattern that was within another subpattern
1841 that had a minimum quantifier of zero caused PCRE to crash. For example,
1842 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1843 recursion. If the recursively-called subpattern itself had a zero repeat,
1844 that was OK.
1845
1846 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1847 buffer into which it was copied (for escape processing) was still set at
1848 1024, so long lines caused crashes.
1849
1850 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1851 "internal error: code overflow...". This applied to any character class
1852 that was followed by a possessive quantifier.
1853
1854 15. Modified the Makefile to add libpcre.la as a prerequisite for
1855 libpcreposix.la because I was told this is needed for a parallel build to
1856 work.
1857
1858 16. If a pattern that contained .* following optional items at the start was
1859 studied, the wrong optimizing data was generated, leading to matching
1860 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1861 matching string must start with a or b or c. The correct conclusion for
1862 this pattern is that a match can start with any character.
1863
1864
1865 Version 4.4 13-Aug-03
1866 ---------------------
1867
1868 1. In UTF-8 mode, a character class containing characters with values between
1869 127 and 255 was not handled correctly if the compiled pattern was studied.
1870 In fixing this, I have also improved the studying algorithm for such
1871 classes (slightly).
1872
1873 2. Three internal functions had redundant arguments passed to them. Removal
1874 might give a very teeny performance improvement.
1875
1876 3. Documentation bug: the value of the capture_top field in a callout is *one
1877 more than* the number of the hightest numbered captured substring.
1878
1879 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1880 in incorrectly linking with a previously installed version. They now link
1881 explicitly with libpcre.la.
1882
1883 5. configure.in no longer needs to recognize Cygwin specially.
1884
1885 6. A problem in pcre.in for Windows platforms is fixed.
1886
1887 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1888 pcretest, it used to include the size of the study block as part of its
1889 output. Unfortunately, the structure contains a field that has a different
1890 size on different hardware architectures. This meant that the tests that
1891 showed this size failed. As the block is currently always of a fixed size,
1892 this information isn't actually particularly useful in pcretest output, so
1893 I have just removed it.
1894
1895 8. Three pre-processor statements accidentally did not start in column 1.
1896 Sadly, there are *still* compilers around that complain, even though
1897 standard C has not required this for well over a decade. Sigh.
1898
1899 9. In pcretest, the code for checking callouts passed small integers in the
1900 callout_data field, which is a void * field. However, some picky compilers
1901 complained about the casts involved for this on 64-bit systems. Now
1902 pcretest passes the address of the small integer instead, which should get
1903 rid of the warnings.
1904
1905 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1906 both compile and run time, and gives an error if an invalid UTF-8 sequence
1907 is found. There is a option for disabling this check in cases where the
1908 string is known to be correct and/or the maximum performance is wanted.
1909
1910 11. In response to a bug report, I changed one line in Makefile.in from
1911
1912 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1913 to
1914 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1915
1916 to look similar to other lines, but I have no way of telling whether this
1917 is the right thing to do, as I do not use Windows. No doubt I'll get told
1918 if it's wrong...
1919
1920
1921 Version 4.3 21-May-03
1922 ---------------------
1923
1924 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1925 Makefile.
1926
1927 2. Some refactoring to improve the quality of the code:
1928
1929 (i) The utf8_table... variables are now declared "const".
1930
1931 (ii) The code for \cx, which used the "case flipping" table to upper case
1932 lower case letters, now just substracts 32. This is ASCII-specific,
1933 but the whole concept of \cx is ASCII-specific, so it seems
1934 reasonable.
1935
1936 (iii) PCRE was using its character types table to recognize decimal and
1937 hexadecimal digits in the pattern. This is silly, because it handles
1938 only 0-9, a-f, and A-F, but the character types table is locale-
1939 specific, which means strange things might happen. A private
1940 table is now used for this - though it costs 256 bytes, a table is
1941 much faster than multiple explicit tests. Of course, the standard
1942 character types table is still used for matching digits in subject
1943 strings against \d.
1944
1945 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1946 ending in _t are). So I've renamed it as ESC_tee.
1947
1948 3. The first argument for regexec() in the POSIX wrapper should have been
1949 defined as "const".
1950
1951 4. Changed pcretest to use malloc() for its buffers so that they can be
1952 Electric Fenced for debugging.
1953
1954 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1955 to read one or more bytes before the start of the subject string. Often this
1956 had no effect on PCRE's behaviour, but in some circumstances it could
1957 provoke a segmentation fault.
1958
1959 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1960 to try to read one or more bytes before the start of the subject string.
1961
1962 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1963 UTF-8 support could misbehave in various ways if the subject string
1964 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1965 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1966 back over UTF-8 characters.)
1967
1968
1969 Version 4.2 14-Apr-03
1970 ---------------------
1971
1972 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1973
1974 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1975 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1976 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1977 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1978 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1979 and BUILD_EXEEXT
1980 Note: automatic setting of the BUILD variables is not yet working
1981 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1982 compile-time but not at link-time
1983 [LINK]: use for linking executables only
1984 make different versions for Windows and non-Windows
1985 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1986 libraries
1987 [LINK_FOR_BUILD]: new variable
1988 [OBJEXT]: use throughout
1989 [EXEEXT]: use throughout
1990 <winshared>: new target
1991 <wininstall>: new target
1992 <dftables.o>: use native compiler
1993 <dftables>: use native linker
1994 <install>: handle Windows platform correctly
1995 <clean>: ditto
1996 <check>: ditto
1997 copy DLL to top builddir before testing
1998
1999 As part of these changes, -no-undefined was removed again. This was reported
2000 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2001 in any case.
2002
2003 3. Some tidies to get rid of compiler warnings:
2004
2005 . In the match_data structure, match_limit was an unsigned long int, whereas
2006 match_call_count was an int. I've made them both unsigned long ints.
2007
2008 . In pcretest the fact that a const uschar * doesn't automatically cast to
2009 a void * provoked a warning.
2010
2011 . Turning on some more compiler warnings threw up some "shadow" variables
2012 and a few more missing casts.
2013
2014 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2015 option, a class that contained a single character with a value between 128
2016 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2017
2018 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2019 option, a class that contained several characters, but with at least one
2020 whose value was between 128 and 255 caused PCRE to crash.
2021
2022
2023 Version 4.1 12-Mar-03
2024 ---------------------
2025
2026 1. Compiling with gcc -pedantic found a couple of places where casts were
2027 needed, and a string in dftables.c that was longer than standard compilers are
2028 required to support.
2029
2030 2. Compiling with Sun's compiler found a few more places where the code could
2031 be tidied up in order to avoid warnings.
2032
2033 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2034 first of these names is deprecated in the latest Autoconf in favour of the name
2035 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2036 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2037 analogy I have changed it to CFLAGS_FOR_BUILD.
2038
2039 4. Added -no-undefined to the linking command in the Makefile, because this is
2040 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2041 linking step for the pcreposix library.
2042
2043 5. PCRE was failing to diagnose the case of two named groups with the same
2044 name.
2045
2046 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2047 literal character that is needed in the subject for a match, and scans along to
2048 ensure that it is present before embarking on the full matching process. This
2049 saves time in cases of nested unlimited repeats that are never going to match.
2050 Problem: the scan can take a lot of time if the subject is very long (e.g.
2051 megabytes), thus penalizing straightforward matches. It is now done only if the
2052 amount of subject to be scanned is less than 1000 bytes.
2053
2054 7. A lesser problem with the same optimization is that it was recording the
2055 first character of an anchored pattern as "needed", thus provoking a search
2056 right along the subject, even when the first match of the pattern was going to
2057 fail. The "needed" character is now not set for anchored patterns, unless it
2058 follows something in the pattern that is of non-fixed length. Thus, it still
2059 fulfils its original purpose of finding quick non-matches in cases of nested
2060 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2061
2062
2063 Version 4.0 17-Feb-03
2064 ---------------------
2065
2066 1. If a comment in an extended regex that started immediately after a meta-item
2067 extended to the end of string, PCRE compiled incorrect data. This could lead to
2068 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2069
2070 2. Moved to autoconf 2.53 and libtool 1.4.2.
2071
2072 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2073 the special perltest8 script is no longer needed - all the tests can be run
2074 from a single perltest script.
2075
2076 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2077 by \s. It has now been removed in PCRE. This means it isn't recognized as
2078 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2079 class [:space:] *does* include VT, thereby creating a mess.
2080
2081 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2082 space and tab.
2083
2084 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2085 its new features into the main test script, reducing the number of scripts.
2086
2087 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2088 were backward compatible, and made the (?i) apply to the whole pattern, as if
2089 /i were given. Now it behaves more logically, and applies the option setting
2090 only to what follows. PCRE has been changed to follow suit. However, if it
2091 finds options settings right at the start of the pattern, it extracts them into
2092 the global options, as before. Thus, they show up in the info data.
2093
2094 8. Added support for the \Q...\E escape sequence. Characters in between are
2095 treated as literals. This is slightly different from Perl in that $ and @ are
2096 also handled as literals inside the quotes. In Perl, they will cause variable
2097 interpolation. Note the following examples:
2098
2099 Pattern PCRE matches Perl matches
2100
2101 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2102 \Qabc\$xyz\E abc\$xyz abc\$xyz
2103 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2104
2105 For compatibility with Perl, \Q...\E sequences are recognized inside character
2106 classes as well as outside them.
2107
2108 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2109 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2110 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2111 signed/unsigned warnings.
2112
2113 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2114 option for pcretest, so I've replaced it by a simple function that does just
2115 that job.
2116
2117 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2118 "pcregrep -".
2119
2120 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2121 Java package. This provides some syntactic sugar for simple cases of what my
2122 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2123 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2124 item, you can use this simplified notation. Note that only makes sense with
2125 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2126 greediness, whatever the setting of the PCRE_UNGREEDY option.
2127
2128 13. A change of greediness default within a pattern was not taking effect at
2129 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2130 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2131 was abstracted outside.
2132
2133 14. PCRE now supports the \G assertion. It is true when the current matching
2134 position is at the start point of the match. This differs from \A when the
2135 starting offset is non-zero. Used with the /g option of pcretest (or similar
2136 code), it works in the same way as it does for Perl's /g option. If all
2137 alternatives of a regex begin with \G, the expression is anchored to the start
2138 match position, and the "anchored" flag is set in the compiled expression.
2139
2140 15. Some bugs concerning the handling of certain option changes within patterns
2141 have been fixed. These applied to options other than (?ims). For example,
2142 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2143 been the other way round. Some of this was related to change 7 above.
2144
2145 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2146 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2147 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2148 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2149
2150 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2151 mode. Unlike ".", it always matches newline, whatever the setting of
2152 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2153 assertions. Perl allows it, but it doesn't (in general) work because it can't
2154 calculate the length of the lookbehind. At least, that's the case for Perl
2155 5.8.0 - I've been told they are going to document that it doesn't work in
2156 future.
2157
2158 18. Added an error diagnosis for escapes that PCRE does not support: these are
2159 \L, \l, \N, \P, \p, \U, \u, and \X.
2160
2161 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2162 reading past the end of the pattern in cases such as /[abcd/.
2163
2164 20. PCRE was getting more memory than necessary for patterns with classes that
2165 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2166
2167 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2168 compiling PCRE for use with Virtual Pascal.
2169
2170 22. Small fix to the Makefile to make it work properly if the build is done
2171 outside the source tree.
2172
2173 23. Added a new extension: a condition to go with recursion. If a conditional
2174 subpattern starts with (?(R) the "true" branch is used if recursion has
2175 happened, whereas the "false" branch is used only at the top level.
2176
2177 24. When there was a very long string of literal characters (over 255 bytes
2178 without UTF support, over 250 bytes with UTF support), the computation of how
2179 much memory was required could be incorrect, leading to segfaults or other
2180 strange effects.
2181
2182 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2183 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2184 there was a subsequent back reference to those brackets. This meant that, for
2185 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2186 possible to check for precisely this case. All we can do is abandon the
2187 optimization if .* occurs inside capturing brackets when there are any back
2188 references whatsoever. (See below for a better fix that came later.)
2189
2190 26. The handling of the optimization for finding the first character of a
2191 non-anchored pattern, and for finding a character that is required later in the
2192 match were failing in some cases. This didn't break the matching; it just
2193 failed to optimize when it could. The way this is done has been re-implemented.
2194
2195 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2196
2197 28. Added a new feature that provides some of the functionality that Perl
2198 provides with (?{...}). The facility is termed a "callout". The way it is done
2199 in PCRE is for the caller to provide an optional function, by setting
2200 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2201 global variable. By default it is unset, which disables all calling out. To get
2202 the function called, the regex must include (?C) at appropriate points. This
2203 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2204 This provides a means of identifying different callout points. When PCRE
2205 reaches such a point in the regex, if pcre_callout has been set, the external
2206 function is called. It is provided with data in a structure called
2207 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2208 matching continues; if it returns a non-zero value, the match at the current
2209 point fails. However, backtracking will occur if possible. [This was changed
2210 later and other features added - see item 49 below.]
2211
2212 29. pcretest is upgraded to test the callout functionality. It provides a
2213 callout function that displays information. By default, it shows the start of
2214 the match and the current position in the text. There are some new data escapes
2215 to vary what happens:
2216
2217 \C+ in addition, show current contents of captured substrings
2218 \C- do not supply a callout function
2219 \C!n return 1 when callout number n is reached
2220 \C!n!m return 1 when callout number n is reached for the mth time
2221
2222 30. If pcregrep was called with the -l option and just a single file name, it
2223 output "<stdin>" if a match was found, instead of the file name.
2224
2225 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2226 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2227 pcre_exec(). This saves a malloc/free per call. The default value of
2228 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2229 when configuring.
2230
2231 32. The default maximum size of a compiled pattern is 64K. There have been a
2232 few cases of people hitting this limit. The code now uses macros to handle the
2233 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2234 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2235 configuring. Tests 2 and 5 work only with 2-byte links because they output
2236 debugging information about compiled patterns.
2237
2238 33. Internal code re-arrangements:
2239
2240 (a) Moved the debugging function for printing out a compiled regex into
2241 its own source file (printint.c) and used #include to pull it into
2242 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2243 separate copies.
2244
2245 (b) Defined the list of op-code names for debugging as a macro in
2246 internal.h so that it is next to the definition of the opcodes.
2247
2248 (c) Defined a table of op-code lengths for simpler skipping along compiled
2249 code. This is again a macro in internal.h so that it is next to the
2250 definition of the opcodes.
2251
2252 34. Added support for recursive calls to individual subpatterns, along the
2253 lines of Robin Houston's patch (but implemented somewhat differently).
2254
2255 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2256 allow it to read and process whole directories in Win32. This code was
2257 contributed by Lionel Fourquaux; it has not been tested by me.
2258
2259 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2260 used to name a group. Names consist of alphanumerics and underscores, and must
2261 be unique. Back references use the syntax (?P=name) and recursive calls use
2262 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2263 numbers. The function pcre_fullinfo() can be used after compilation to extract
2264 a name/number map. There are three relevant calls:
2265
2266 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2267 PCRE_INFO_NAMECOUNT yields the number of entries
2268 PCRE_INFO_NAMETABLE yields a pointer to the map.
2269
2270 The map is a vector of fixed-size entries. The size of each entry depends on
2271 the length of the longest name used. The first two bytes of each entry are the
2272 group number, most significant byte first. There follows the corresponding
2273 name, zero terminated. The names are in alphabetical order.
2274
2275 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2276 case instead of 255. Making it the same both with and without UTF-8 support
2277 means that the same test output works with both.
2278
2279 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2280 calling malloc() with a zero argument.
2281
2282 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2283 optimization. I've improved things by keeping a bitmap of backreferences with
2284 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2285 fact referenced, the optimization can be applied. It is unlikely that a
2286 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2287 the match to follow \n) will appear inside brackets with a number greater than
2288 31, but if it does, any back reference > 31 suppresses the optimization.
2289
2290 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2291 of disabling numbered capturing parentheses. Any opening parenthesis that is
2292 not followed by ? behaves as if it were followed by ?: but named parentheses
2293 can still be used for capturing (and they will acquire numbers in the usual
2294 way).
2295
2296 41. Redesigned the return codes from the match() function into yes/no/error so
2297 that errors can be passed back from deep inside the nested calls. A malloc
2298 failure while inside a recursive subpattern call now causes the
2299 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2300
2301 42. It is now possible to set a limit on the number of times the match()
2302 function is called in a call to pcre_exec(). This facility makes it possible to
2303 limit the amount of recursion and backtracking, though not in a directly
2304 obvious way, because the match() function is used in a number of different
2305 circumstances. The count starts from zero for each position in the subject
2306 string (for non-anchored patterns). The default limit is, for compatibility, a
2307 large number, namely 10 000 000. You can change this in two ways:
2308
2309 (a) When configuring PCRE before making, you can use --with-match-limit=n
2310 to set a default value for the compiled library.
2311
2312 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2313 a different value is set. See 45 below.
2314
2315 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2316
2317 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2318 of things that can be changed at compile time. The first argument specifies
2319 what is wanted and the second points to where the information is to be placed.
2320 The current list of available information is:
2321
2322 PCRE_CONFIG_UTF8
2323
2324 The output is an integer that is set to one if UTF-8 support is available;
2325 otherwise it is set to zero.
2326
2327 PCRE_CONFIG_NEWLINE
2328
2329 The output is an integer that it set to the value of the code that is used for
2330 newline. It is either LF (10) or CR (13).
2331
2332 PCRE_CONFIG_LINK_SIZE
2333
2334 The output is an integer that contains the number of bytes used for internal
2335 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2336
2337 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2338
2339 The output is an integer that contains the threshold above which the POSIX
2340 interface uses malloc() for output vectors. See item 31 above.
2341
2342 PCRE_CONFIG_MATCH_LIMIT
2343
2344 The output is an unsigned integer that contains the default limit of the number
2345 of match() calls in a pcre_exec() execution. See 42 above.
2346
2347 44. pcretest has been upgraded by the addition of the -C option. This causes it
2348 to extract all the available output from the new pcre_config() function, and to
2349 output it. The program then exits immediately.
2350
2351 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2352 order to support additional features. One way would have been to define
2353 pcre_exec2() (for example) with extra arguments, but this would not have been
2354 extensible, and would also have required all calls to the original function to
2355 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2356 is used for passing in "extra" data from pcre_study().
2357
2358 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2359 contains the following fields:
2360
2361 flags a bitmap indicating which of the following fields are set
2362 study_data opaque data from pcre_study()
2363 match_limit a way of specifying a limit on match() calls for a specific
2364 call to pcre_exec()
2365 callout_data data for callouts (see 49 below)
2366
2367 The flag bits are also defined in pcre.h, and are
2368
2369 PCRE_EXTRA_STUDY_DATA
2370 PCRE_EXTRA_MATCH_LIMIT
2371 PCRE_EXTRA_CALLOUT_DATA
2372
2373 The pcre_study() function now returns one of these new pcre_extra blocks, with
2374 the actual study data pointed to by the study_data field, and the
2375 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2376 before. That is, this change is entirely upwards-compatible and requires no
2377 change to existing code.
2378
2379 If you want to pass in additional data to pcre_exec(), you can either place it
2380 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2381 block.
2382
2383 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2384 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2385 times with different match limits, until it finds the minimum value needed for
2386 pcre_exec() to complete. The value is then output. This can be instructive; for
2387 most simple matches the number is quite small, but for pathological cases it
2388 gets very large very quickly.
2389
2390 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2391 returns the size of the data block pointed to by the study_data field in a
2392 pcre_extra block, that is, the value that was passed as the argument to
2393 pcre_malloc() when PCRE was getting memory in which to place the information
2394 created by pcre_study(). The fourth argument should point to a size_t variable.
2395 pcretest has been extended so that this information is shown after a successful
2396 pcre_study() call when information about the compiled regex is being displayed.
2397
2398 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2399 because what follows is always an absolute path. (Later: it turns out that this
2400 is more than cosmetic for MinGW, because it doesn't like empty path
2401 components.)
2402
2403 49. Some changes have been made to the callout feature (see 28 above):
2404
2405 (i) A callout function now has three choices for what it returns:
2406
2407 0 => success, carry on matching
2408 > 0 => failure at this point, but backtrack if possible
2409 < 0 => serious error, return this value from pcre_exec()
2410
2411 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2412 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2413 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2414 use by callout functions. It will never be used by PCRE itself.
2415
2416 (ii) The pcre_extra structure (see 45 above) has a void * field called
2417 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2418 pcre_callout_block structure has a field of the same name. The contents of
2419 the field passed in the pcre_extra structure are passed to the callout
2420 function in the corresponding field in the callout block. This makes it
2421 easier to use the same callout-containing regex from multiple threads. For
2422 testing, the pcretest program has a new data escape
2423
2424 \C*n pass the number n (may be negative) as callout_data
2425
2426 If the callout function in pcretest receives a non-zero value as
2427 callout_data, it returns that value.
2428
2429 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2430 there were some redundant $(CFLAGS) in commands that are now specified as
2431 $(LINK), which already includes $(CFLAGS).
2432
2433 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2434 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2435 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2436 one-byte characters throughout. Note that case-insensitive matching applies
2437 only to characters whose values are less than 256. PCRE doesn't support the
2438 notion of cases for higher-valued characters.
2439
2440 (i) A character class whose characters are all within 0-255 is handled as
2441 a bit map, and the map is inverted for negative classes. Previously, a
2442 character > 255 always failed to match such a class; however it should
2443 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2444
2445 (ii) A negated character class with a single character < 255 is coded as
2446 "not this character" (OP_NOT). This wasn't working properly when the test
2447 character was multibyte, either singly or repeated.
2448
2449 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2450 mode, for example: \x{100}{2,3}.
2451
2452 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2453 singly or repeated) now correctly test multibyte characters. However,
2454 PCRE doesn't recognize any characters with values greater than 255 as
2455 digits, spaces, or word characters. Such characters always match \D, \S,
2456 and \W, and never match \d, \s, or \w.
2457
2458 (v) Classes may now contain characters and character ranges with values
2459 greater than 255. For example: [ab\x{100}-\x{400}].
2460
2461 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2462 PCRE in UTF-8 mode.
2463
2464 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2465 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2466 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2467 value.)
2468
2469 53. The single man page has become too large. I have therefore split it up into
2470 a number of separate man pages. These also give rise to individual HTML pages;
2471 these are now put in a separate directory, and there is an index.html page that
2472 lists them all. Some hyperlinking between the pages has been installed.
2473
2474 54. Added convenience functions for handling named capturing parentheses.
2475
2476 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2477 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2478 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2479 are faulted.
2480
2481 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2482 calling configure. These values are used when compiling the dftables.c program
2483 which is run to generate the source of the default character tables. They
2484 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2485 you will need to set these values.
2486
2487 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2488
2489
2490 Version 3.9 02-Jan-02
2491 ---------------------
2492
2493 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2494
2495 2. If --disable-static was given, the building process failed when trying to
2496 build pcretest and pcregrep. (For some reason it was using libtool to compile
2497 them, which is not right, as they aren't part of the library.)
2498
2499
2500 Version 3.8 18-Dec-01
2501 ---------------------
2502
2503 1. The experimental UTF-8 code was completely screwed up. It was packing the
2504 bytes in the wrong order. How dumb can you get?
2505
2506
2507 Version 3.7 29-Oct-01
2508 ---------------------
2509
2510 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2511 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2512 this didn't happen under Solaris 8, where I normally test things.
2513
2514 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2515 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2516 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2517 this fix an hour or so after the initial 3.7 release.)
2518
2519
2520 Version 3.6 23-Oct-01
2521 ---------------------
2522
2523 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2524 offsets passed as NULL with zero offset count.
2525
2526 2. The config.guess and config.sub files had not been updated when I moved to
2527 the latest autoconf.
2528
2529
2530 Version 3.5 15-Aug-01
2531 ---------------------
2532
2533 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2534 had been forgotten.
2535
2536 2. By using declared but undefined structures, we can avoid using "void"
2537 definitions in pcre.h while keeping the internal definitions of the structures
2538 private.
2539
2540 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2541 user point of view, this means that both static and shared libraries are built
2542 by default, but this can be individually controlled. More of the work of
2543 handling this static/shared cases is now inside libtool instead of PCRE's make
2544 file.
2545
2546 4. The pcretest utility is now installed along with pcregrep because it is
2547 useful for users (to test regexs) and by doing this, it automatically gets
2548 relinked by libtool. The documentation has been turned into a man page, so
2549 there are now .1, .txt, and .html versions in /doc.
2550
2551 5. Upgrades to pcregrep:
2552 (i) Added long-form option names like gnu grep.
2553 (ii) Added --help to list all options with an explanatory phrase.
2554 (iii) Added -r, --recursive to recurse into sub-directories.
2555 (iv) Added -f, --file to read patterns from a file.
2556
2557 6. pcre_exec() was referring to its "code" argument before testing that
2558 argument for NULL (and giving an error if it was NULL).
2559
2560 7. Upgraded Makefile.in to allow for compiling in a different directory from
2561 the source directory.
2562
2563 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2564 options bits, the pointer it was passed was to an int instead of to an unsigned
2565 long int. This mattered only on 64-bit systems.
2566
2567 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2568 generated) instead of pcre.in, which it its source. Also made the same change
2569 in several of the .c files.
2570
2571 10. A new release of gcc defines printf() as a macro, which broke pcretest
2572 because it had an ifdef in the middle of a string argument for printf(). Fixed
2573 by using separate calls to printf().
2574
2575 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2576 script, to force use of CR or LF instead of \n in the source. On non-Unix
2577 systems, the value can be set in config.h.
2578
2579 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2580 absolute limit. Changed the text of the error message to make this clear, and
2581 likewise updated the man page.
2582
2583 13. The limit of 99 on the number of capturing subpatterns has been removed.
2584 The new limit is 65535, which I hope will not be a "real" limit.
2585
2586
2587 Version 3.4 22-Aug-00
2588 ---------------------
2589
2590 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2591
2592 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2593
2594
2595 Version 3.3 01-Aug-00
2596 ---------------------
2597
2598 1. If an octal character was given, but the value was greater than \377, it
2599 was not getting masked to the least significant bits, as documented. This could
2600 lead to crashes in some systems.
2601
2602 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2603 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2604
2605 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2606 These just pass their arguments on to (pcre_free)(), but they are provided
2607 because some uses of PCRE bind it to non-C systems that can call its functions,
2608 but cannot call free() or pcre_free() directly.
2609
2610 4. Add "make test" as a synonym for "make check". Corrected some comments in
2611 the Makefile.
2612
2613 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2614 Makefile.
2615
2616 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2617 command called pgrep for grepping around the active processes.
2618
2619 7. Added the beginnings of support for UTF-8 character strings.
2620
2621 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2622 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2623 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2624 out for the ar command.)
2625
2626
2627 Version 3.2 12-May-00
2628 ---------------------
2629
2630 This is purely a bug fixing release.
2631
2632 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2633 of ZA. This was just one example of several cases that could provoke this bug,
2634 which was introduced by change 9 of version 2.00. The code for breaking
2635 infinite loops after an iteration that matches an empty string was't working
2636 correctly.
2637
2638 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2639 when matched against abbab (for example). After matching an empty string, it
2640 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2641 caused it to match further down the string than it should.
2642
2643 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2644 was there because it doesn't seem to be needed, and it causes trouble on some
2645 systems, as it is not a Standard C header. It has been removed.
2646
2647 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2648 were reported on the Macintosh. The changes were from
2649
2650 while ((c = *(++ptr)) != 0 && c != '\n');
2651 to
2652 while ((c = *(++ptr)) != 0 && c != '\n') ;
2653
2654 Totally extraordinary, but if that's what it takes...
2655
2656 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2657 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2658 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2659 assumes the way PCRE uses memmove() (always moving upwards).
2660
2661 6. PCRE is being used in one environment where strchr() is not available. There
2662 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2663 faster code anyway.
2664
2665
2666 Version 3.1 09-Feb-00
2667 ---------------------
2668
2669 The only change in this release is the fixing of some bugs in Makefile.in for
2670 the "install" target:
2671
2672 (1) It was failing to install pcreposix.h.
2673
2674 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2675
2676
2677 Version 3.0 01-Feb-00
2678 ---------------------
2679
2680 1. Add support for the /+ modifier to perltest (to output $` like it does in
2681 pcretest).
2682
2683 2. Add support for the /g modifier to perltest.
2684
2685 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2686 matches null strings.
2687
2688 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2689 pattern. Perl treats empty patterns specially - it reuses the most recent
2690 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2691 effect.
2692
2693 5. The POSIX interface was broken in that it was just handing over the POSIX
2694 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2695 required a bigger vector, with some working space on the end. This means that
2696 the POSIX wrapper now has to get and free some memory, and copy the results.
2697
2698 6. Added some simple autoconf support, placing the test data and the
2699 documentation in separate directories, re-organizing some of the
2700 information files, and making it build pcre-config (a GNU standard). Also added
2701 libtool support for building PCRE as a shared library, which is now the
2702 default.
2703
2704 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2705 09 are not valid octal constants. Single digits will be used for minor values
2706 less than 10.
2707
2708 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2709 existing programs that set these in the POSIX interface can use PCRE without
2710 modification.
2711
2712 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2713 return all that pcre_info() returns, plus additional data. The pcre_info()
2714 function is retained for compatibility, but is considered to be obsolete.
2715
2716 10. Added experimental recursion feature (?R) to handle one common case that
2717 Perl 5.6 will be able to do with (?p{...}).
2718
2719 11. Added support for POSIX character classes like [:alpha:], which Perl is
2720 adopting.
2721
2722
2723 Version 2.08 31-Aug-99
2724 ----------------------
2725
2726 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2727 trying to match at the startoffset position, but instead was moving forward to
2728 the next newline as if a previous match had failed.
2729
2730 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2731 and could get into a loop if a null string was matched other than at the start
2732 of the subject.
2733
2734 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2735 be distinguished at compile time, and for completeness also added PCRE_DATE.
2736
2737 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2738 in GnuWin32 environments.
2739
2740
2741 Version 2.07 29-Jul-99
2742 ----------------------
2743
2744 1. The documentation is now supplied in plain text form and HTML as well as in
2745 the form of man page sources.
2746
2747 2. C++ compilers don't like assigning (void *) values to other pointer types.
2748 In particular this affects malloc(). Although there is no problem in Standard
2749 C, I've put in casts to keep C++ compilers happy.
2750
2751 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2752 should be (const char *).
2753
2754 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2755 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2756 However, I haven't made this a standard facility. The documentation doesn't
2757 mention it, and the Makefile doesn't support it.
2758
2759 5. The Makefile now contains an "install" target, with editable destinations at
2760 the top of the file. The pcretest program is not installed.
2761
2762 6. pgrep -V now gives the PCRE version number and date.
2763
2764 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2765 causing the entire string to be ignored, instead of just the last character.
2766
2767 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2768 non-matching string, it can take a very, very long time, even for strings of
2769 quite modest length, because of the nested recursion. PCRE now does better in
2770 some of these cases. It does this by remembering the last required literal
2771 character in the pattern, and pre-searching the subject to ensure it is present
2772 before running the real match. In other words, it applies a heuristic to detect
2773 some types of certain failure quickly, and in the above example, if presented
2774 with a string that has no trailing " it gives "no match" very quickly.
2775
2776 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2777 other alternatives are tried instead.
2778
2779
2780 Version 2.06 09-Jun-99
2781 ----------------------
2782
2783 1. Change pcretest's output for amount of store used to show just the code
2784 space, because the remainder (the data block) varies in size between 32-bit and
2785 64-bit systems.
2786
2787 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2788 start matching at. This allows lookbehinds to work when searching for multiple
2789 occurrences in a string.
2790
2791 3. Added additional options to pcretest for testing multiple occurrences:
2792
2793 /+ outputs the rest of the string that follows a match
2794 /g loops for multiple occurrences, using the new startoffset argument
2795 /G loops for multiple occurrences by passing an incremented pointer
2796
2797 4. PCRE wasn't doing the "first character" optimization for patterns starting
2798 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2799 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2800 the letter 'x'. On long subject strings, this gives a significant speed-up.
2801
2802
2803 Version 2.05 21-Apr-99
2804 ----------------------
2805
2806 1. Changed the type of magic_number from int to long int so that it works
2807 properly on 16-bit systems.
2808
2809 2. Fixed a bug which caused patterns starting with .* not to work correctly
2810 when the subject string contained newline characters. PCRE was assuming
2811 anchoring for such patterns in all cases, which is not correct because .* will
2812 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2813 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2814 must be retried after every newline in the subject.
2815
2816
2817 Version 2.04 18-Feb-99
2818 ----------------------
2819
2820 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2821 computation of the store needed to hold the pattern was incorrect (too large).
2822 If such patterns were nested a few deep, this could multiply and become a real
2823 problem.
2824
2825 2. Added /M option to pcretest to show the memory requirement of a specific
2826 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2827
2828 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2829 compiled in such a way that the backtracking after subsequent failure was
2830 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2831 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2832
2833
2834 Version 2.03 02-Feb-99
2835 ----------------------
2836
2837 1. Fixed typo and small mistake in man page.
2838
2839 2. Added 4th condition (GPL supersedes if conflict) and created separate
2840 LICENCE file containing the conditions.
2841
2842 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2843 Perl, that is the internal \ allows the delimiter to be included in the
2844 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2845 the final delimiter, add \ to the end of the pattern (to test the error).
2846
2847 4. Added the convenience functions for extracting substrings after a successful
2848 match. Updated pcretest to make it able to test these functions.
2849
2850
2851 Version 2.02 14-Jan-99
2852 ----------------------
2853
2854 1. Initialized the working variables associated with each extraction so that
2855 their saving and restoring doesn't refer to uninitialized store.
2856
2857 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2858 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2859 fix the problem.
2860
2861 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2862 calls, and wasn't printing the correct value for compiling calls. Increased the
2863 default value of LOOPREPEAT, and the number of significant figures in the
2864 times.
2865
2866 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2867
2868 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2869 a building problem on Windows NT with a FAT file system.
2870
2871
2872 Version 2.01 21-Oct-98
2873 ----------------------
2874
2875 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2876 to character tables built by pcre_maketables() in the current locale. If NULL
2877 is passed, the default tables are used.
2878
2879
2880 Version 2.00 24-Sep-98
2881 ----------------------
2882
2883 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2884 it any more.
2885
2886 2. Allow quantification of (?>) groups, and make it work correctly.
2887
2888 3. The first character computation wasn't working for (?>) groups.
2889
2890 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2891 end of the subject) and add 5.005's \z, which really does match only at the
2892 very end of the subject.
2893
2894 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2895
2896 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2897 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2898 localized options. All options to pcre_study() were also removed.
2899
2900 7. Add other new features from 5.005:
2901
2902 $(?<= positive lookbehind
2903 $(?<! negative lookbehind
2904 (?imsx-imsx) added the unsetting capability
2905 such a setting is global if at outer level; local otherwise
2906 (?imsx-imsx:) non-capturing groups with option setting
2907 (?(cond)re|re) conditional pattern matching
2908
2909 A backreference to itself in a repeated group matches the previous
2910 captured string.
2911
2912 8. General tidying up of studying (both automatic and via "study")
2913 consequential on the addition of new assertions.
2914
2915 9. As in 5.005, unlimited repeated groups that could match an empty substring
2916 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2917 runtime if any iteration does actually match an empty substring.
2918
2919 10. Include the RunTest script in the distribution.
2920
2921 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2922 discrepancies, some of which were old and were also with respect to 5.004. They
2923 have now been fixed.
2924
2925
2926 Version 1.09 28-Apr-98
2927 ----------------------
2928
2929 1. A negated single character class followed by a quantifier with a minimum
2930 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2931 program crashes, or just wrong answers. This did not apply to negated classes
2932 containing more than one character, or to minima other than one.
2933
2934
2935 Version 1.08 27-Mar-98
2936 ----------------------
2937
2938 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2939
2940 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2941 latter must appear before anything that relies on it in the pattern.
2942
2943
2944 Version 1.07 16-Feb-98
2945 ----------------------
2946
2947 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2948 repeat of a potentially empty string).
2949
2950
2951 Version 1.06 23-Jan-98
2952 ----------------------
2953
2954 1. Added Markus Oberhumer's little patches for C++.
2955
2956 2. Literal strings longer than 255 characters were broken.
2957
2958
2959 Version 1.05 23-Dec-97
2960 ----------------------
2961
2962 1. Negated character classes containing more than one character were failing if
2963 PCRE_CASELESS was set at run time.
2964
2965
2966 Version 1.04 19-Dec-97
2967 ----------------------
2968
2969 1. Corrected the man page, where some "const" qualifiers had been omitted.
2970
2971 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2972 input syntax.
2973
2974 3. Fixed memory leak which occurred when a regex with back references was
2975 matched with an offsets vector that wasn't big enough. The temporary memory
2976 that is used in this case wasn't being freed if the match failed.
2977
2978 4. Tidied pcretest to ensure it frees memory that it gets.
2979
2980 5. Temporary memory was being obtained in the case where the passed offsets
2981 vector was exactly big enough.
2982
2983 6. Corrected definition of offsetof() from change 5 below.
2984
2985 7. I had screwed up change 6 below and broken the rules for the use of
2986 setjmp(). Now fixed.
2987
2988
2989 Version 1.03 18-Dec-97
2990 ----------------------
2991
2992 1. A erroneous regex with a missing opening parenthesis was correctly
2993 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2994 on some systems.
2995
2996 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2997 it was reported that one broken compiler failed on the former because "code" is
2998 also an independent variable.
2999
3000 3. The erroneous regex a[]b caused an array overrun reference.
3001
3002 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3003 fail on data ending with that character. (It was going on too far, and checking
3004 the next character, typically a binary zero.) This was specific to the
3005 optimized code for single-character negative classes.
3006
3007 5. Added a contributed patch from the TIN world which does the following:
3008
3009 + Add an undef for memmove, in case the the system defines a macro for it.
3010
3011 + Add a definition of offsetof(), in case there isn't one. (I don't know
3012 the reason behind this - offsetof() is part of the ANSI standard - but
3013 it does no harm).
3014
3015 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3016 most of the places where whitespace preceded '#'. I have given up and
3017 allowed the remaining 2 cases to be at the margin.
3018
3019 + Rename some variables in pcre to eliminate shadowing. This seems very
3020 pedantic, but does no harm, of course.
3021
3022 6. Moved the call to setjmp() into its own function, to get rid of warnings
3023 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3024
3025 7. Constructs such as \d{8,} were compiling into the equivalent of
3026 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3027 outcome, but in this particular case used more store than had been allocated,
3028 which caused the bug to be discovered because it threw up an internal error.
3029
3030 8. The debugging code in both pcre and pcretest for outputting the compiled
3031 form of a regex was going wrong in the case of back references followed by
3032 curly-bracketed repeats.
3033
3034
3035 Version 1.02 12-Dec-97
3036 ----------------------
3037
3038 1. Typos in pcre.3 and comments in the source fixed.
3039
3040 2. Applied a contributed patch to get rid of places where it used to remove
3041 'const' from variables, and fixed some signed/unsigned and uninitialized
3042 variable warnings.
3043
3044 3. Added the "runtest" target to Makefile.
3045
3046 4. Set default compiler flag to -O2 rather than just -O.
3047
3048
3049 Version 1.01 19-Nov-97
3050 ----------------------
3051
3052 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3053 like /([ab]*)*/, that is, for classes with more than one character in them.
3054
3055 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3056 as /((?>a*))*/ (a PCRE_EXTRA facility).
3057
3058
3059 Version 1.00 18-Nov-97
3060 ----------------------
3061
3062 1. Added compile-time macros to support systems such as SunOS4 which don't have
3063 memmove() or strerror() but have other things that can be used instead.
3064
3065 2. Arranged that "make clean" removes the executables.
3066
3067
3068 Version 0.99 27-Oct-97
3069 ----------------------
3070
3071 1. Fixed bug in code for optimizing classes with only one character. It was
3072 initializing a 32-byte map regardless, which could cause it to run off the end
3073 of the memory it had got.
3074
3075 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3076
3077
3078 Version 0.98 22-Oct-97
3079 ----------------------
3080
3081 1. Fixed bug in code for handling temporary memory usage when there are more
3082 back references than supplied space in the ovector. This could cause segfaults.
3083
3084
3085 Version 0.97 21-Oct-97
3086 ----------------------
3087
3088 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3089
3090 2. Optimized negated single characters not to use a bit map.
3091
3092 3. Brought error texts together as macro definitions; clarified some of them;
3093 fixed one that was wrong - it said "range out of order" when it meant "invalid
3094 escape sequence".
3095
3096 4. Changed some char * arguments to const char *.
3097
3098 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3099
3100 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3101 pcretest.
3102
3103
3104 Version 0.96 16-Oct-97
3105 ----------------------
3106
3107 1. Added a simple "pgrep" utility to the distribution.
3108
3109 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3110 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3111 where "ddd" means "one or more decimal digits".
3112
3113 3. Fixed serious bug. If a pattern had a back reference, but the call to
3114 pcre_exec() didn't supply a large enough ovector to record the related
3115 identifying subpattern, the match always failed. PCRE now remembers the number
3116 of the largest back reference, and gets some temporary memory in which to save
3117 the offsets during matching if necessary, in order to ensure that
3118 backreferences always work.
3119
3120 4. Increased the compatibility with Perl in a number of ways:
3121
3122 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3123 to request this handling. The option can be set at compile or exec time.
3124
3125 (b) $ matches before a terminating newline by default; an option
3126 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3127 mode). The option can be set at compile or exec time.
3128
3129 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3130 the same as Perl's. If the decimal number it represents is less than 10
3131 or there aren't that many previous left capturing parentheses, an octal
3132 escape is read. Inside a character class, it's always an octal escape,
3133 even if it is a single digit.
3134
3135 (d) An escaped but undefined alphabetic character is taken as a literal,
3136 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3137 escapes.
3138
3139 (e) {0} is now permitted. (The previous item is removed from the compiled
3140 pattern).
3141
3142 5. Changed all the names of code files so that the basic parts are no longer
3143 than 10 characters, and abolished the teeny "globals.c" file.
3144
3145 6. Changed the handling of character classes; they are now done with a 32-byte
3146 bit map always.
3147
3148 7. Added the -d and /D options to pcretest to make it possible to look at the
3149 internals of compilation without having to recompile pcre.
3150
3151
3152 Version 0.95 23-Sep-97
3153 ----------------------
3154
3155 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3156 \x20 at the start of a run of normal characters. These were being treated as
3157 real characters, instead of the source characters being re-checked.
3158
3159
3160 Version 0.94 18-Sep-97
3161 ----------------------
3162
3163 1. The functions are now thread-safe, with the caveat that the global variables
3164 containing pointers to malloc() and free() or alternative functions are the
3165 same for all threads.
3166
3167 2. Get pcre_study() to generate a bitmap of initial characters for non-
3168 anchored patterns when this is possible, and use it if passed to pcre_exec().
3169
3170
3171 Version 0.93 15-Sep-97
3172 ----------------------
3173
3174 1. /(b)|(:+)/ was computing an incorrect first character.
3175
3176 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3177 but not actually doing anything yet.
3178
3179 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3180 as Perl does (e.g. [-az] or [az-]).
3181
3182 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3183 all possible positions.
3184
3185 5. Split up into different modules to avoid including unneeded functions in a
3186 compiled binary. However, compile and exec are still in one module. The "study"
3187 function is split off.
3188
3189 6. The character tables are now in a separate module whose source is generated
3190 by an auxiliary program - but can then be edited by hand if required. There are
3191 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3192 toupper() in the code.
3193
3194 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3195 make them global. Abolish the function for setting them, as the caller can now
3196 set them directly.
3197
3198
3199 Version 0.92 11-Sep-97
3200 ----------------------
3201
3202 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3203 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3204
3205 2. Caseless matching was not working in character classes if the characters in
3206 the pattern were in upper case.
3207
3208 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3209
3210 4. Make PCRE_ANCHORED public and accept as a compile option.
3211
3212 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3213 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3214 pass them.
3215
3216 6. Give an error if bad option bits passed at compile or run time.
3217
3218 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3219 pcretest to cause it to pass that flag.
3220
3221 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3222 options, and the first character, if set.
3223
3224 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3225
3226
3227 Version 0.91 10-Sep-97
3228 ----------------------
3229
3230 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3231 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3232
3233 2. PCRE was looping on encountering an indefinitely repeated back reference to
3234 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3235 Perl does - treats the match as successful.
3236
3237 ****

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