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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 407 - (show annotations) (download)
Mon Mar 23 15:29:18 2009 UTC (5 years, 5 months ago) by ph10
File size: 164739 byte(s)
Upgrade CMake support to disable tests, pcregrep, and configuration listing.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.9 xx-xxx-09
5 ---------------------
6
7 1. When building with support for bzlib/zlib (pcregrep) and/or readline
8 (pcretest), all targets were linked against these libraries. This included
9 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
10 libraries. This caused unwanted dependencies to be created. This problem
11 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
12 pcretest is linked with readline.
13
14 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
15 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
16 moved outside it again, because FALSE and TRUE are already defined in AIX,
17 but BOOL is not.
18
19 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
20 PCRE_MATCH_LIMIT_RETURSION values as ints, when they should be long ints.
21
22 4. The pcregrep documentation said spaces were inserted as well as colons (or
23 hyphens) following file names and line numbers when outputting matching
24 lines. This is not true; no spaces are inserted. I have also clarified the
25 wording for the --colour (or --color) option.
26
27 5. In pcregrep, when --colour was used with -o, the list of matching strings
28 was not coloured; this is different to GNU grep, so I have changed it to be
29 the same.
30
31 6. When --colo(u)r was used in pcregrep, only the first matching substring in
32 each matching line was coloured. Now it goes on to look for further matches
33 of any of the test patterns, which is the same behaviour as GNU grep.
34
35 7. A pattern that could match an empty string could cause pcregrep to loop; it
36 doesn't make sense to accept an empty string match in pcregrep, so I have
37 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
38 seems to be how GNU grep behaves.
39
40 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
41 start or after a newline", because the conditional assertion was not being
42 correctly handled. The rule now is that both the assertion and what follows
43 in the first alternative must satisfy the test.
44
45 9. If auto-callout was enabled in a pattern with a conditional group whose
46 condition was an assertion, PCRE could crash during matching, both with
47 pcre_exec() and pcre_dfa_exec().
48
49 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
50 used for matching.
51
52 11. Unicode property support in character classes was not working for
53 characters (bytes) greater than 127 when not in UTF-8 mode.
54
55 12. Added the -M command line option to pcretest.
56
57 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
58
59 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
60
61 16. Added comments and documentation about mis-use of no_arg in the C++
62 wrapper.
63
64 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
65 from Martin Jerabek that uses macro names for all relevant character and
66 string constants.
67
68 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
69 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
70 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
71 these, but not everybody uses configure.
72
73 19. A conditional group that had only one branch was not being correctly
74 recognized as an item that could match an empty string. This meant that an
75 enclosing group might also not be so recognized, causing infinite looping
76 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
77 with the subject "ab", where knowledge that the repeated group can match
78 nothing is needed in order to break the loop.
79
80 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
81 exec(), but without supplying a callout function, matching went wrong.
82
83 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
84 leak if the size of the offset vector was greater than 30. When the vector
85 is smaller, the saved offsets during recursion go onto a local stack
86 vector, but for larger vectors malloc() is used. It was failing to free
87 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
88 error, in fact).
89
90 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
91 heapframe that is used only when UTF-8 support is enabled. This caused no
92 problem, but was untidy.
93
94 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
95 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
96 included within another project.
97
98 24. Steven Van Ingelgem's patches to add more options to the CMake support,
99 slightly modified by me:
100
101 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
102 not building pcregrep.
103
104 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
105 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
106
107
108 Version 7.8 05-Sep-08
109 ---------------------
110
111 1. Replaced UCP searching code with optimized version as implemented for Ad
112 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
113 stage table and inline lookup instead of a function, giving speed ups of 2
114 to 5 times on some simple patterns that I tested. Permission was given to
115 distribute the MultiStage2.py script that generates the tables (it's not in
116 the tarball, but is in the Subversion repository).
117
118 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
119 scripts.
120
121 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
122 a group with a zero qualifier. The result of the study could be incorrect,
123 or the function might crash, depending on the pattern.
124
125 4. Caseless matching was not working for non-ASCII characters in back
126 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
127 It now works when Unicode Property Support is available.
128
129 5. In pcretest, an escape such as \x{de} in the data was always generating
130 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
131 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
132 truncation.
133
134 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
135
136 7. Added two (int) casts to pcregrep when printing the difference of two
137 pointers, in case they are 64-bit values.
138
139 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
140 test 2 if it fails.
141
142 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
143 and a #define of that name to empty if it is not externally set. This is to
144 allow users of MSVC to set it if necessary.
145
146 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
147 the convenience functions in the pcre_get.c source file.
148
149 11. An option change at the start of a pattern that had top-level alternatives
150 could cause overwriting and/or a crash. This command provoked a crash in
151 some environments:
152
153 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
154
155 This potential security problem was recorded as CVE-2008-2371.
156
157 12. For a pattern where the match had to start at the beginning or immediately
158 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
159 pcre_dfa_exec() could read past the end of the passed subject if there was
160 no match. To help with detecting such bugs (e.g. with valgrind), I modified
161 pcretest so that it places the subject at the end of its malloc-ed buffer.
162
163 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
164 exec() might read past the end of the data buffer in UTF-8 mode.
165
166 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
167 the data contained the byte 0x85 as part of a UTF-8 character within its
168 first line. This applied both to normal and DFA matching.
169
170 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
171 /^[^d]*?$/8 failed to match "abc".
172
173 16. Added a missing copyright notice to pcrecpp_internal.h.
174
175 17. Make it more clear in the documentation that values returned from
176 pcre_exec() in ovector are byte offsets, not character counts.
177
178 18. Tidied a few places to stop certain compilers from issuing warnings.
179
180 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
181 supplied by Stefan Weber. I made a further small update for 7.8 because
182 there is a change of source arrangements: the pcre_searchfuncs.c module is
183 replaced by pcre_ucd.c.
184
185
186 Version 7.7 07-May-08
187 ---------------------
188
189 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
190 a string to a long long, pretend we don't even have a long long." This is
191 done by checking for the strtoq, strtoll, and _strtoi64 functions.
192
193 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
194 pre-7.6 versions, which defined a global no_arg variable instead of putting
195 it in the RE class. (See also #8 below.)
196
197 3. Remove a line of dead code, identified by coverity and reported by Nuno
198 Lopes.
199
200 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
201
202 (1) The include/exclude patterns were being applied to the whole pathnames
203 of files, instead of just to the final components.
204
205 (2) If there was more than one level of directory, the subdirectories were
206 skipped unless they satisfied the include/exclude conditions. This is
207 inconsistent with GNU grep (and could even be seen as contrary to the
208 pcregrep specification - which I improved to make it absolutely clear).
209 The action now is always to scan all levels of directory, and just
210 apply the include/exclude patterns to regular files.
211
212 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
213 --exclude_dir in the tests to avoid scanning .svn directories.
214
215 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
216 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
217 doesn't support NULs in patterns.
218
219 7. Added some missing "const"s to declarations of static tables in
220 pcre_compile.c and pcre_dfa_exec.c.
221
222 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
223 caused by fix #2 above. (Subsequently also a second patch to fix the
224 first patch. And a third patch - this was a messy problem.)
225
226 9. Applied Craig's patch to remove the use of push_back().
227
228 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
229 matching function regexec().
230
231 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
232 which, however, unlike Perl's \g{...}, are subroutine calls, not back
233 references. PCRE supports relative numbers with this syntax (I don't think
234 Oniguruma does).
235
236 12. Previously, a group with a zero repeat such as (...){0} was completely
237 omitted from the compiled regex. However, this means that if the group
238 was called as a subroutine from elsewhere in the pattern, things went wrong
239 (an internal error was given). Such groups are now left in the compiled
240 pattern, with a new opcode that causes them to be skipped at execution
241 time.
242
243 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
244 to the way PCRE behaves:
245
246 (a) A lone ] character is dis-allowed (Perl treats it as data).
247
248 (b) A back reference to an unmatched subpattern matches an empty string
249 (Perl fails the current match path).
250
251 (c) A data ] in a character class must be notated as \] because if the
252 first data character in a class is ], it defines an empty class. (In
253 Perl it is not possible to have an empty class.) The empty class []
254 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
255 The negative empty class [^] matches any one character, independently
256 of the DOTALL setting.
257
258 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
259 non-existent subpattern following a character class starting with ']' and
260 containing () gave an internal compiling error instead of "reference to
261 non-existent subpattern". Fortunately, when the pattern did exist, the
262 compiled code was correct. (When scanning forwards to check for the
263 existencd of the subpattern, it was treating the data ']' as terminating
264 the class, so got the count wrong. When actually compiling, the reference
265 was subsequently set up correctly.)
266
267 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
268 it was being rejected as not supported by pcre_dfa_exec(), even though
269 other assertions are supported. I have made pcre_dfa_exec() support
270 (*FAIL).
271
272 16. The implementation of 13c above involved the invention of a new opcode,
273 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
274 cannot be changed at match time, I realized I could make a small
275 improvement to matching performance by compiling OP_ALLANY instead of
276 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
277 on the OP_ANY path.
278
279 17. Compiling pcretest on Windows with readline support failed without the
280 following two fixes: (1) Make the unistd.h include conditional on
281 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
282
283 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
284 ncurses library to be included for pcretest when ReadLine support is
285 requested, but also to allow for it to be overridden. This patch came from
286 Daniel Bergström.
287
288 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
289 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
290 any errors with the current Unicode tables. Thanks to Peter Kankowski for
291 spotting this.
292
293
294 Version 7.6 28-Jan-08
295 ---------------------
296
297 1. A character class containing a very large number of characters with
298 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
299 overflow.
300
301 2. Patch to cut out the "long long" test in pcrecpp_unittest when
302 HAVE_LONG_LONG is not defined.
303
304 3. Applied Christian Ehrlicher's patch to update the CMake build files to
305 bring them up to date and include new features. This patch includes:
306
307 - Fixed PH's badly added libz and libbz2 support.
308 - Fixed a problem with static linking.
309 - Added pcredemo. [But later removed - see 7 below.]
310 - Fixed dftables problem and added an option.
311 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
312 HAVE_LONG_LONG.
313 - Added readline support for pcretest.
314 - Added an listing of the option settings after cmake has run.
315
316 4. A user submitted a patch to Makefile that makes it easy to create
317 "pcre.dll" under mingw when using Configure/Make. I added stuff to
318 Makefile.am that cause it to include this special target, without
319 affecting anything else. Note that the same mingw target plus all
320 the other distribution libraries and programs are now supported
321 when configuring with CMake (see 6 below) instead of with
322 Configure/Make.
323
324 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
325 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
326 exported in the Windows port". It has not yet been confirmed that the patch
327 solves the problem, but it does no harm.
328
329 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
330 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
331 with CMake, and also correct the comment about stack recursion.
332
333 7. Remove the automatic building of pcredemo from the ./configure system and
334 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
335 of a program that users should build themselves after PCRE is installed, so
336 building it automatically is not really right. What is more, it gave
337 trouble in some build environments.
338
339 8. Further tidies to CMakeLists.txt from Sheri and Christian.
340
341
342 Version 7.5 10-Jan-08
343 ---------------------
344
345 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
346 values in parens when parsing an RE using the C++ wrapper."
347
348 2. Negative specials like \S did not work in character classes in UTF-8 mode.
349 Characters greater than 255 were excluded from the class instead of being
350 included.
351
352 3. The same bug as (2) above applied to negated POSIX classes such as
353 [:^space:].
354
355 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
356 defined or documented. It seems to have been a typo for PCRE_STATIC, so
357 I have changed it.
358
359 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
360 first named subpattern) and a construct such as (?&a) would reference the
361 first named subpattern whose name started with "a" (in other words, the
362 length check was missing). Both these problems are fixed. "Subpattern name
363 expected" is now given for (?&) (a zero-length name), and this patch also
364 makes it give the same error for \k'' (previously it complained that that
365 was a reference to a non-existent subpattern).
366
367 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
368 this is right because (?- can be followed by option settings as well as by
369 digits. I have, however, made the messages clearer.
370
371 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
372 than the number used in the conditional) now cause a compile-time error.
373 This is actually not compatible with Perl, which accepts such patterns, but
374 treats the conditional as always being FALSE (as PCRE used to), but it
375 seems to me that giving a diagnostic is better.
376
377 8. Change "alphameric" to the more common word "alphanumeric" in comments
378 and messages.
379
380 9. Fix two occurrences of "backslash" in comments that should have been
381 "backspace".
382
383 10. Remove two redundant lines of code that can never be obeyed (their function
384 was moved elsewhere).
385
386 11. The program that makes PCRE's Unicode character property table had a bug
387 which caused it to generate incorrect table entries for sequences of
388 characters that have the same character type, but are in different scripts.
389 It amalgamated them into a single range, with the script of the first of
390 them. In other words, some characters were in the wrong script. There were
391 thirteen such cases, affecting characters in the following ranges:
392
393 U+002b0 - U+002c1
394 U+0060c - U+0060d
395 U+0061e - U+00612
396 U+0064b - U+0065e
397 U+0074d - U+0076d
398 U+01800 - U+01805
399 U+01d00 - U+01d77
400 U+01d9b - U+01dbf
401 U+0200b - U+0200f
402 U+030fc - U+030fe
403 U+03260 - U+0327f
404 U+0fb46 - U+0fbb1
405 U+10450 - U+1049d
406
407 12. The -o option (show only the matching part of a line) for pcregrep was not
408 compatible with GNU grep in that, if there was more than one match in a
409 line, it showed only the first of them. It now behaves in the same way as
410 GNU grep.
411
412 13. If the -o and -v options were combined for pcregrep, it printed a blank
413 line for every non-matching line. GNU grep prints nothing, and pcregrep now
414 does the same. The return code can be used to tell if there were any
415 non-matching lines.
416
417 14. Added --file-offsets and --line-offsets to pcregrep.
418
419 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
420 infinitely looping recursion. The bug was that positive lookaheads were not
421 being skipped when checking for a possible empty match (negative lookaheads
422 and both kinds of lookbehind were skipped).
423
424 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
425 inclusion of <windows.h> to before rather than after the definition of
426 INVALID_FILE_ATTRIBUTES (patch from David Byron).
427
428 17. Specifying a possessive quantifier with a specific limit for a Unicode
429 character property caused pcre_compile() to compile bad code, which led at
430 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
431 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
432 caused the error; without that there was no problem.
433
434 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
435
436 19. Added --enable-pcretest-libreadline.
437
438 20. In pcrecpp.cc, the variable 'count' was incremented twice in
439 RE::GlobalReplace(). As a result, the number of replacements returned was
440 double what it should be. I removed one of the increments, but Craig sent a
441 later patch that removed the other one (the right fix) and added unit tests
442 that check the return values (which was not done before).
443
444 21. Several CMake things:
445
446 (1) Arranged that, when cmake is used on Unix, the libraries end up with
447 the names libpcre and libpcreposix, not just pcre and pcreposix.
448
449 (2) The above change means that pcretest and pcregrep are now correctly
450 linked with the newly-built libraries, not previously installed ones.
451
452 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
453
454 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
455 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
456 UTF-8 newline character). The key issue is that the pattern starts .*;
457 this means that the match must be either at the beginning, or after a
458 newline. The bug was in the code for advancing after a failed match and
459 checking that the new position followed a newline. It was not taking
460 account of UTF-8 characters correctly.
461
462 23. PCRE was behaving differently from Perl in the way it recognized POSIX
463 character classes. PCRE was not treating the sequence [:...:] as a
464 character class unless the ... were all letters. Perl, however, seems to
465 allow any characters between [: and :], though of course it rejects as
466 unknown any "names" that contain non-letters, because all the known class
467 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
468 for example, whereas PCRE did not - it did not recognize a POSIX character
469 class. This seemed a bit dangerous, so the code has been changed to be
470 closer to Perl. The behaviour is not identical to Perl, because PCRE will
471 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
472 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
473 Perl does, and where it didn't before.
474
475 24. Rewrite so as to remove the single use of %n from pcregrep because in some
476 Windows environments %n is disabled by default.
477
478
479 Version 7.4 21-Sep-07
480 ---------------------
481
482 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
483 means that a class such as [\s] counted as "explicit reference to CR or
484 LF". That isn't really right - the whole point of the change was to try to
485 help when there was an actual mention of one of the two characters. So now
486 the change happens only if \r or \n (or a literal CR or LF) character is
487 encountered.
488
489 2. The 32-bit options word was also used for 6 internal flags, but the numbers
490 of both had grown to the point where there were only 3 bits left.
491 Fortunately, there was spare space in the data structure, and so I have
492 moved the internal flags into a new 16-bit field to free up more option
493 bits.
494
495 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
496 but did not set the internal JCHANGED flag - either of these is enough to
497 control the way the "get" function works - but the PCRE_INFO_JCHANGED
498 facility is supposed to tell if (?J) was ever used, so now (?J) at the
499 start sets both bits.
500
501 4. Added options (at build time, compile time, exec time) to change \R from
502 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
503
504 5. doc/pcresyntax.html was missing from the distribution.
505
506 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
507 compatibility, even though it is no longer used.
508
509 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
510 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
511 windows.h file is present (where different names are used). [This was
512 reversed later after testing - see 16 below.]
513
514 8. Changed all #include <config.h> to #include "config.h". There were also
515 some further <pcre.h> cases that I changed to "pcre.h".
516
517 9. When pcregrep was used with the --colour option, it missed the line ending
518 sequence off the lines that it output.
519
520 10. It was pointed out to me that arrays of string pointers cause lots of
521 relocations when a shared library is dynamically loaded. A technique of
522 using a single long string with a table of offsets can drastically reduce
523 these. I have refactored PCRE in four places to do this. The result is
524 dramatic:
525
526 Originally: 290
527 After changing UCP table: 187
528 After changing error message table: 43
529 After changing table of "verbs" 36
530 After changing table of Posix names 22
531
532 Thanks to the folks working on Gregex for glib for this insight.
533
534 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
535 unicode-properties was also set.
536
537 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
538
539 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
540 checked only for CRLF.
541
542 14. Added casts to pcretest.c to avoid compiler warnings.
543
544 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
545
546 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
547 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
548 entirely. This removes changes made in 7 above.
549
550 17. The CMake files have been updated, and there is now more information about
551 building with CMake in the NON-UNIX-USE document.
552
553
554 Version 7.3 28-Aug-07
555 ---------------------
556
557 1. In the rejigging of the build system that eventually resulted in 7.1, the
558 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
559 brackets there is not right, since it causes compilers to look for an
560 installed pcre.h, not the version that is in the source that is being
561 compiled (which of course may be different). I have changed it back to:
562
563 #include "pcre.h"
564
565 I have a vague recollection that the change was concerned with compiling in
566 different directories, but in the new build system, that is taken care of
567 by the VPATH setting the Makefile.
568
569 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
570 when the subject happened to end in the byte 0x85 (e.g. if the last
571 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
572 characters but of course it shouldn't be taken as a newline when it is part
573 of another character. The bug was that, for an unlimited repeat of . in
574 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
575 characters when looking for a newline.
576
577 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
578
579 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
580 in debug output.
581
582 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
583 long printing in the pcrecpp unittest when running under MinGW.
584
585 6. ESC_K was left out of the EBCDIC table.
586
587 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
588 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
589 limit also applies to "virtual nesting" when a pattern is recursive, and in
590 this case 1000 isn't so big. I have been able to remove this limit at the
591 expense of backing off one optimization in certain circumstances. Normally,
592 when pcre_exec() would call its internal match() function recursively and
593 immediately return the result unconditionally, it uses a "tail recursion"
594 feature to save stack. However, when a subpattern that can match an empty
595 string has an unlimited repetition quantifier, it no longer makes this
596 optimization. That gives it a stack frame in which to save the data for
597 checking that an empty string has been matched. Previously this was taken
598 from the 1000-entry workspace that had been reserved. So now there is no
599 explicit limit, but more stack is used.
600
601 8. Applied Daniel's patches to solve problems with the import/export magic
602 syntax that is required for Windows, and which was going wrong for the
603 pcreposix and pcrecpp parts of the library. These were overlooked when this
604 problem was solved for the main library.
605
606 9. There were some crude static tests to avoid integer overflow when computing
607 the size of patterns that contain repeated groups with explicit upper
608 limits. As the maximum quantifier is 65535, the maximum group length was
609 set at 30,000 so that the product of these two numbers did not overflow a
610 32-bit integer. However, it turns out that people want to use groups that
611 are longer than 30,000 bytes (though not repeat them that many times).
612 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
613 made it possible to implement the integer overflow checks in a much more
614 dynamic way, which I have now done. The artificial limitation on group
615 length has been removed - we now have only the limit on the total length of
616 the compiled pattern, which depends on the LINK_SIZE setting.
617
618 10. Fixed a bug in the documentation for get/copy named substring when
619 duplicate names are permitted. If none of the named substrings are set, the
620 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
621 empty string.
622
623 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
624 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
625 because the ] is interpreted as the first data character and the
626 terminating ] is not found. PCRE has been made compatible with Perl in this
627 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
628 cause memory overwriting.
629
630 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
631 string has been matched (to stop an infinite loop). It was not recognizing
632 a conditional subpattern that could match an empty string if that
633 subpattern was within another subpattern. For example, it looped when
634 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
635 condition was not nested. This bug has been fixed.
636
637 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
638 past the start of the subject in the presence of bytes with the top bit
639 set, for example "\x8aBCD".
640
641 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
642 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
643
644 14. Optimized (?!) to (*FAIL).
645
646 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
647 This restricts code points to be within the range 0 to 0x10FFFF, excluding
648 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
649 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
650 does: it's just the validity check that is more restrictive.
651
652 16. Inserted checks for integer overflows during escape sequence (backslash)
653 processing, and also fixed erroneous offset values for syntax errors during
654 backslash processing.
655
656 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
657 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
658
659 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
660 caused an overrun.
661
662 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
663 something other than just ASCII characters) inside a group that had an
664 unlimited repeat caused a loop at compile time (while checking to see
665 whether the group could match an empty string).
666
667 20. Debugging a pattern containing \p or \P could cause a crash. For example,
668 [\P{Any}] did so. (Error in the code for printing property names.)
669
670 21. An orphan \E inside a character class could cause a crash.
671
672 22. A repeated capturing bracket such as (A)? could cause a wild memory
673 reference during compilation.
674
675 23. There are several functions in pcre_compile() that scan along a compiled
676 expression for various reasons (e.g. to see if it's fixed length for look
677 behind). There were bugs in these functions when a repeated \p or \P was
678 present in the pattern. These operators have additional parameters compared
679 with \d, etc, and these were not being taken into account when moving along
680 the compiled data. Specifically:
681
682 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
683 length.
684
685 (b) An item such as \pL+ within a repeated group could cause crashes or
686 loops.
687
688 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
689 "reference to non-existent subpattern" error.
690
691 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
692
693 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
694 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
695
696 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
697
698 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
699 character were causing crashes (broken optimization).
700
701 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
702 \p or \P) caused a compile-time loop.
703
704 28. More problems have arisen in unanchored patterns when CRLF is a valid line
705 break. For example, the unstudied pattern [\r\n]A does not match the string
706 "\r\nA" because change 7.0/46 below moves the current point on by two
707 characters after failing to match at the start. However, the pattern \nA
708 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
709 the same is true. There doesn't seem any very clean way out of this, but
710 what I have chosen to do makes the common cases work: PCRE now takes note
711 of whether there can be an explicit match for \r or \n anywhere in the
712 pattern, and if so, 7.0/46 no longer applies. As part of this change,
713 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
714 pattern has explicit CR or LF references.
715
716 29. Added (*CR) etc for changing newline setting at start of pattern.
717
718
719 Version 7.2 19-Jun-07
720 ---------------------
721
722 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
723 which is apparently normally available under Windows.
724
725 2. Re-jig the pcregrep tests with different newline settings in an attempt
726 to make them independent of the local environment's newline setting.
727
728 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
729
730 4. Some of the "internals" tests were previously cut out when the link size
731 was not 2, because the output contained actual offsets. The recent new
732 "Z" feature of pcretest means that these can be cut out, making the tests
733 usable with all link sizes.
734
735 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
736 stack recursion. This gives a massive performance boost under BSD, but just
737 a small improvement under Linux. However, it saves one field in the frame
738 in all cases.
739
740 6. Added more features from the forthcoming Perl 5.10:
741
742 (a) (?-n) (where n is a string of digits) is a relative subroutine or
743 recursion call. It refers to the nth most recently opened parentheses.
744
745 (b) (?+n) is also a relative subroutine call; it refers to the nth next
746 to be opened parentheses.
747
748 (c) Conditions that refer to capturing parentheses can be specified
749 relatively, for example, (?(-2)... or (?(+3)...
750
751 (d) \K resets the start of the current match so that everything before
752 is not part of it.
753
754 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
755
756 (f) \g{name} is another synonym - part of Perl 5.10's unification of
757 reference syntax.
758
759 (g) (?| introduces a group in which the numbering of parentheses in each
760 alternative starts with the same number.
761
762 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
763
764 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
765 PCRE_INFO_JCHANGED.
766
767 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
768 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
769 for detecting groups that can match an empty string.
770
771 9. A pattern with a very large number of alternatives (more than several
772 hundred) was running out of internal workspace during the pre-compile
773 phase, where pcre_compile() figures out how much memory will be needed. A
774 bit of new cunning has reduced the workspace needed for groups with
775 alternatives. The 1000-alternative test pattern now uses 12 bytes of
776 workspace instead of running out of the 4096 that are available.
777
778 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
779
780 11. Applied patch from Google to remove an optimization that didn't quite work.
781 The report of the bug said:
782
783 pcrecpp::RE("a*").FullMatch("aaa") matches, while
784 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
785 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
786
787 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
788 it matched the wrong number of bytes.
789
790
791 Version 7.1 24-Apr-07
792 ---------------------
793
794 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
795 that is more "standard", making use of automake and other Autotools. There
796 is some re-arrangement of the files and adjustment of comments consequent
797 on this.
798
799 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
800 for recursive directory scanning broke on some systems because the files
801 are not scanned in any specific order and on different systems the order
802 was different. A call to "sort" has been inserted into RunGrepTest for the
803 approprate test as a short-term fix. In the longer term there may be an
804 alternative.
805
806 3. I had an email from Eric Raymond about problems translating some of PCRE's
807 man pages to HTML (despite the fact that I distribute HTML pages, some
808 people do their own conversions for various reasons). The problems
809 concerned the use of low-level troff macros .br and .in. I have therefore
810 removed all such uses from the man pages (some were redundant, some could
811 be replaced by .nf/.fi pairs). The 132html script that I use to generate
812 HTML has been updated to handle .nf/.fi and to complain if it encounters
813 .br or .in.
814
815 4. Updated comments in configure.ac that get placed in config.h.in and also
816 arranged for config.h to be included in the distribution, with the name
817 config.h.generic, for the benefit of those who have to compile without
818 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
819
820 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
821 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
822 makevp.bat for the current PCRE, using the additional files
823 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
824
825 6. A Windows user reported a minor discrepancy with test 2, which turned out
826 to be caused by a trailing space on an input line that had got lost in his
827 copy. The trailing space was an accident, so I've just removed it.
828
829 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
830 that is needed.
831
832 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
833 as "const" (a) because they are and (b) because it helps the PHP
834 maintainers who have recently made a script to detect big data structures
835 in the php code that should be moved to the .rodata section. I remembered
836 to update Builducptable as well, so it won't revert if ucptable.h is ever
837 re-created.
838
839 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
840 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
841 order to be able to cut out the UTF-8 tables in the latter when UTF-8
842 support is not required. This saves 1.5-2K of code, which is important in
843 some applications.
844
845 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
846 so as not to refer to the tables, even though these functions will never be
847 called when UTF-8 support is disabled. Otherwise there are problems with a
848 shared library.
849
850 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
851
852 (a) It was defining its arguments as char * instead of void *.
853
854 (b) It was assuming that all moves were upwards in memory; this was true
855 a long time ago when I wrote it, but is no longer the case.
856
857 The emulated memove() is provided for those environments that have neither
858 memmove() nor bcopy(). I didn't think anyone used it these days, but that
859 is clearly not the case, as these two bugs were recently reported.
860
861 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
862 and Detrail to create the HTML documentation, the .txt form of the man
863 pages, and it removes trailing spaces from listed files. It also creates
864 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
865 case, it wraps all the #defines with #ifndefs. This script should be run
866 before "make dist".
867
868 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
869 with Unicode property support.
870
871 (a) For a maximizing quantifier, if the two different cases of the
872 character were of different lengths in their UTF-8 codings (there are
873 some cases like this - I found 11), and the matching function had to
874 back up over a mixture of the two cases, it incorrectly assumed they
875 were both the same length.
876
877 (b) When PCRE was configured to use the heap rather than the stack for
878 recursion during matching, it was not correctly preserving the data for
879 the other case of a UTF-8 character when checking ahead for a match
880 while processing a minimizing repeat. If the check also involved
881 matching a wide character, but failed, corruption could cause an
882 erroneous result when trying to check for a repeat of the original
883 character.
884
885 13. Some tidying changes to the testing mechanism:
886
887 (a) The RunTest script now detects the internal link size and whether there
888 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
889 values substituted by "configure". (The RunGrepTest script already did
890 this for UTF-8.) The configure.ac script no longer substitutes the
891 relevant variables.
892
893 (b) The debugging options /B and /D in pcretest show the compiled bytecode
894 with length and offset values. This means that the output is different
895 for different internal link sizes. Test 2 is skipped for link sizes
896 other than 2 because of this, bypassing the problem. Unfortunately,
897 there was also a test in test 3 (the locale tests) that used /B and
898 failed for link sizes other than 2. Rather than cut the whole test out,
899 I have added a new /Z option to pcretest that replaces the length and
900 offset values with spaces. This is now used to make test 3 independent
901 of link size. (Test 2 will be tidied up later.)
902
903 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
904 segmentation fault instead of returning the appropriate error message.
905
906 15. In multiline mode when the newline sequence was set to "any", the pattern
907 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
908 This doesn't seem right; it now treats the CRLF combination as the line
909 ending, and so does not match in that case. It's only a pattern such as ^$
910 that would hit this one: something like ^ABC$ would have failed after \r
911 and then tried again after \r\n.
912
913 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
914 in an attempt to make files that differ only in their line terminators
915 compare equal. This works on Linux.
916
917 17. Under certain error circumstances pcregrep might try to free random memory
918 as it exited. This is now fixed, thanks to valgrind.
919
920 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
921 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
922 was because its rules for how to advance for /g after matching an empty
923 string at the end of a line did not allow for this case. They now check for
924 it specially.
925
926 20. pcretest is supposed to handle patterns and data of any length, by
927 extending its buffers when necessary. It was getting this wrong when the
928 buffer for a data line had to be extended.
929
930 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
931 CRLF as a newline sequence.
932
933 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
934 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
935 I have nevertheless tidied it up.
936
937 23. Added some casts to kill warnings from HP-UX ia64 compiler.
938
939 24. Added a man page for pcre-config.
940
941
942 Version 7.0 19-Dec-06
943 ---------------------
944
945 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
946 moving to gcc 4.1.1.
947
948 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
949 sys/time.h, which is documented as needed for this function. It doesn't
950 seem to matter on Linux, but it showed up on some releases of OS X.
951
952 3. It seems that there are systems where bytes whose values are greater than
953 127 match isprint() in the "C" locale. The "C" locale should be the
954 default when a C program starts up. In most systems, only ASCII printing
955 characters match isprint(). This difference caused the output from pcretest
956 to vary, making some of the tests fail. I have changed pcretest so that:
957
958 (a) When it is outputting text in the compiled version of a pattern, bytes
959 other than 32-126 are always shown as hex escapes.
960
961 (b) When it is outputting text that is a matched part of a subject string,
962 it does the same, unless a different locale has been set for the match
963 (using the /L modifier). In this case, it uses isprint() to decide.
964
965 4. Fixed a major bug that caused incorrect computation of the amount of memory
966 required for a compiled pattern when options that changed within the
967 pattern affected the logic of the preliminary scan that determines the
968 length. The relevant options are -x, and -i in UTF-8 mode. The result was
969 that the computed length was too small. The symptoms of this bug were
970 either the PCRE error "internal error: code overflow" from pcre_compile(),
971 or a glibc crash with a message such as "pcretest: free(): invalid next
972 size (fast)". Examples of patterns that provoked this bug (shown in
973 pcretest format) are:
974
975 /(?-x: )/x
976 /(?x)(?-x: \s*#\s*)/
977 /((?i)[\x{c0}])/8
978 /(?i:[\x{c0}])/8
979
980 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
981 is now done differently.
982
983 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
984 wrapper classes; (b) implement a new function in the C++ scanner that is
985 more efficient than the old way of doing things because it avoids levels of
986 recursion in the regex matching; (c) add a paragraph to the documentation
987 for the FullMatch() function.
988
989 6. The escape sequence \n was being treated as whatever was defined as
990 "newline". Not only was this contrary to the documentation, which states
991 that \n is character 10 (hex 0A), but it also went horribly wrong when
992 "newline" was defined as CRLF. This has been fixed.
993
994 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
995 was being set to -1 for the "end of line" case (supposedly a value that no
996 character can have). Though this value is never used (the check for end of
997 line is "zero bytes in current character"), it caused compiler complaints.
998 I've changed it to 0xffffffff.
999
1000 8. In pcre_version.c, the version string was being built by a sequence of
1001 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1002 string (as it is for production releases) called a macro with an empty
1003 argument. The C standard says the result of this is undefined. The gcc
1004 compiler treats it as an empty string (which was what was wanted) but it is
1005 reported that Visual C gives an error. The source has been hacked around to
1006 avoid this problem.
1007
1008 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1009 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1010 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1011 of them did).
1012
1013 10. Originally, pcretest opened its input and output without "b"; then I was
1014 told that "b" was needed in some environments, so it was added for release
1015 5.0 to both the input and output. (It makes no difference on Unix-like
1016 systems.) Later I was told that it is wrong for the input on Windows. I've
1017 now abstracted the modes into two macros, to make it easier to fiddle with
1018 them, and removed "b" from the input mode under Windows.
1019
1020 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1021
1022 12. Added -help and --help to pcretest as an official way of being reminded
1023 of the options.
1024
1025 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1026 and pcrecpp.cc because they annoy compilers at high warning levels.
1027
1028 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1029
1030 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1031 scripts are not C programs :-) and which was not noticed because it works
1032 on Linux.
1033
1034 16. pcretest is supposed to handle any length of pattern and data line (as one
1035 line or as a continued sequence of lines) by extending its input buffer if
1036 necessary. This feature was broken for very long pattern lines, leading to
1037 a string of junk being passed to pcre_compile() if the pattern was longer
1038 than about 50K.
1039
1040 17. I have done a major re-factoring of the way pcre_compile() computes the
1041 amount of memory needed for a compiled pattern. Previously, there was code
1042 that made a preliminary scan of the pattern in order to do this. That was
1043 OK when PCRE was new, but as the facilities have expanded, it has become
1044 harder and harder to keep it in step with the real compile phase, and there
1045 have been a number of bugs (see for example, 4 above). I have now found a
1046 cunning way of running the real compile function in a "fake" mode that
1047 enables it to compute how much memory it would need, while actually only
1048 ever using a few hundred bytes of working memory and without too many
1049 tests of the mode. This should make future maintenance and development
1050 easier. A side effect of this work is that the limit of 200 on the nesting
1051 depth of parentheses has been removed (though this was never a serious
1052 limitation, I suspect). However, there is a downside: pcre_compile() now
1053 runs more slowly than before (30% or more, depending on the pattern). I
1054 hope this isn't a big issue. There is no effect on runtime performance.
1055
1056 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1057 newline (only possible for the last line of a file) and it was a
1058 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1059
1060 19. Added additional timing features to pcretest. (1) The -tm option now times
1061 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1062 separate command line item, by a number that specifies the number of
1063 repeats to use when timing. The default is 50000; this gives better
1064 precision, but takes uncomfortably long for very large patterns.
1065
1066 20. Extended pcre_study() to be more clever in cases where a branch of a
1067 subpattern has no definite first character. For example, (a*|b*)[cd] would
1068 previously give no result from pcre_study(). Now it recognizes that the
1069 first character must be a, b, c, or d.
1070
1071 21. There was an incorrect error "recursive call could loop indefinitely" if
1072 a subpattern (or the entire pattern) that was being tested for matching an
1073 empty string contained only one non-empty item after a nested subpattern.
1074 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1075 incorrectly, because the \d was being skipped in the check.
1076
1077 22. The pcretest program now has a new pattern option /B and a command line
1078 option -b, which is equivalent to adding /B to every pattern. This causes
1079 it to show the compiled bytecode, without the additional information that
1080 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1081 is the same as /B/I).
1082
1083 23. A new optimization is now able automatically to treat some sequences such
1084 as a*b as a*+b. More specifically, if something simple (such as a character
1085 or a simple class like \d) has an unlimited quantifier, and is followed by
1086 something that cannot possibly match the quantified thing, the quantifier
1087 is automatically "possessified".
1088
1089 24. A recursive reference to a subpattern whose number was greater than 39
1090 went wrong under certain circumstances in UTF-8 mode. This bug could also
1091 have affected the operation of pcre_study().
1092
1093 25. Realized that a little bit of performance could be had by replacing
1094 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1095
1096 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1097
1098 27. Possessive quantifiers such as a++ were previously implemented by turning
1099 them into atomic groups such as ($>a+). Now they have their own opcodes,
1100 which improves performance. This includes the automatically created ones
1101 from 23 above.
1102
1103 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1104 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1105 the first matched character to be a colon. This applied both to named and
1106 numbered groups.
1107
1108 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1109
1110 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1111 building PCRE on Windows easier, so I have included it in the distribution.
1112
1113 31. There is now a check in pcretest against a ridiculously large number being
1114 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1115 loop, the loop is abandoned.
1116
1117 32. Forward references to subpatterns in conditions such as (?(2)...) where
1118 subpattern 2 is defined later cause pcre_compile() to search forwards in
1119 the pattern for the relevant set of parentheses. This search went wrong
1120 when there were unescaped parentheses in a character class, parentheses
1121 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1122
1123 33. "Subroutine" calls and backreferences were previously restricted to
1124 referencing subpatterns earlier in the regex. This restriction has now
1125 been removed.
1126
1127 34. Added a number of extra features that are going to be in Perl 5.10. On the
1128 whole, these are just syntactic alternatives for features that PCRE had
1129 previously implemented using the Python syntax or my own invention. The
1130 other formats are all retained for compatibility.
1131
1132 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1133 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1134 also .NET compatible.
1135
1136 (b) A recursion or subroutine call to a named group can now be defined as
1137 (?&name) as well as (?P>name).
1138
1139 (c) A backreference to a named group can now be defined as \k<name> or
1140 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1141 5.10, are also .NET compatible.
1142
1143 (d) A conditional reference to a named group can now use the syntax
1144 (?(<name>) or (?('name') as well as (?(name).
1145
1146 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1147 groups (named and numbered) that are never evaluated inline, but can be
1148 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1149 is always false. There may be only one alternative in such a group.
1150
1151 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1152 as the simple (?(R). The condition is true only if the most recent
1153 recursion is that of the given number or name. It does not search out
1154 through the entire recursion stack.
1155
1156 (g) The escape \gN or \g{N} has been added, where N is a positive or
1157 negative number, specifying an absolute or relative reference.
1158
1159 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1160 some "unreachable code" warnings.
1161
1162 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1163 things, this adds five new scripts.
1164
1165 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1166 There were also incompatibilities regarding the handling of \Q..\E inside
1167 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1168 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1169
1170 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1171 matches an empty string, and forcibly breaks the loop. There were bugs in
1172 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1173 against aaaa the result was just "a" rather than "aaaa", for example. Two
1174 separate and independent bugs (that affected different cases) have been
1175 fixed.
1176
1177 39. Refactored the code to abolish the use of different opcodes for small
1178 capturing bracket numbers. This is a tidy that I avoided doing when I
1179 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1180 The new approach is not only tidier, it makes it possible to reduce the
1181 memory needed to fix the previous bug (38).
1182
1183 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1184 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1185 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1186 mode.
1187
1188 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1189 report.
1190
1191 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1192 copy construction and assignment in the C++ wrapper.
1193
1194 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1195 couple of bugs that could have given wrong results in the "--newline=crlf"
1196 case.
1197
1198 44. Added a number of casts and did some reorganization of signed/unsigned int
1199 variables following suggestions from Dair Grant. Also renamed the variable
1200 "this" as "item" because it is a C++ keyword.
1201
1202 45. Arranged for dftables to add
1203
1204 #include "pcre_internal.h"
1205
1206 to pcre_chartables.c because without it, gcc 4.x may remove the array
1207 definition from the final binary if PCRE is built into a static library and
1208 dead code stripping is activated.
1209
1210 46. For an unanchored pattern, if a match attempt fails at the start of a
1211 newline sequence, and the newline setting is CRLF or ANY, and the next two
1212 characters are CRLF, advance by two characters instead of one.
1213
1214
1215 Version 6.7 04-Jul-06
1216 ---------------------
1217
1218 1. In order to handle tests when input lines are enormously long, pcretest has
1219 been re-factored so that it automatically extends its buffers when
1220 necessary. The code is crude, but this _is_ just a test program. The
1221 default size has been increased from 32K to 50K.
1222
1223 2. The code in pcre_study() was using the value of the re argument before
1224 testing it for NULL. (Of course, in any sensible call of the function, it
1225 won't be NULL.)
1226
1227 3. The memmove() emulation function in pcre_internal.h, which is used on
1228 systems that lack both memmove() and bcopy() - that is, hardly ever -
1229 was missing a "static" storage class specifier.
1230
1231 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1232 containing an extended class (one that cannot be represented by a bitmap
1233 because it contains high-valued characters or Unicode property items, e.g.
1234 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1235 pattern, but PCRE should not loop if you do not (it no longer does).
1236 [Detail: two cases were found: (a) a repeated subpattern containing an
1237 extended class; (b) a recursive reference to a subpattern that followed a
1238 previous extended class. It wasn't skipping over the extended class
1239 correctly when UTF-8 mode was not set.]
1240
1241 5. A negated single-character class was not being recognized as fixed-length
1242 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1243 compile error "lookbehind assertion is not fixed length".
1244
1245 6. The RunPerlTest auxiliary script was showing an unexpected difference
1246 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1247 write a Perl script that can interpret lines of an input file either as
1248 byte characters or as UTF-8, which is what "perltest" was being required to
1249 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1250 can't do is switch easily at run time between having the "use utf8;" pragma
1251 or not. In the end, I fudged it by using the RunPerlTest script to insert
1252 "use utf8;" explicitly for the UTF-8 tests.
1253
1254 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1255 the end of the subject string, contrary to the documentation and to what
1256 Perl does. This was true of both matching functions. Now it matches only at
1257 the start of the subject and immediately after *internal* newlines.
1258
1259 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1260 a pointer to an int instead of a pointer to an unsigned long int. This
1261 caused problems on 64-bit systems.
1262
1263 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1264 instance of the 'standard' template library not being so standard".
1265
1266 10. There was no check on the number of named subpatterns nor the maximum
1267 length of a subpattern name. The product of these values is used to compute
1268 the size of the memory block for a compiled pattern. By supplying a very
1269 long subpattern name and a large number of named subpatterns, the size
1270 computation could be caused to overflow. This is now prevented by limiting
1271 the length of names to 32 characters, and the number of named subpatterns
1272 to 10,000.
1273
1274 11. Subpatterns that are repeated with specific counts have to be replicated in
1275 the compiled pattern. The size of memory for this was computed from the
1276 length of the subpattern and the repeat count. The latter is limited to
1277 65535, but there was no limit on the former, meaning that integer overflow
1278 could in principle occur. The compiled length of a repeated subpattern is
1279 now limited to 30,000 bytes in order to prevent this.
1280
1281 12. Added the optional facility to have named substrings with the same name.
1282
1283 13. Added the ability to use a named substring as a condition, using the
1284 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1285 are numbers (not recommended). Forward references are permitted.
1286
1287 14. Added forward references in named backreferences (if you see what I mean).
1288
1289 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1290 pattern could run off the end of the subject. For example, the pattern
1291 "(?s)(.{1,5})"8 did this with the subject "ab".
1292
1293 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1294 PCRE_CASELESS was set when matching characters that were quantified with ?
1295 or *.
1296
1297 17. A character class other than a single negated character that had a minimum
1298 but no maximum quantifier - for example [ab]{6,} - was not handled
1299 correctly by pce_dfa_exec(). It would match only one character.
1300
1301 18. A valid (though odd) pattern that looked like a POSIX character
1302 class but used an invalid character after [ (for example [[,abc,]]) caused
1303 pcre_compile() to give the error "Failed: internal error: code overflow" or
1304 in some cases to crash with a glibc free() error. This could even happen if
1305 the pattern terminated after [[ but there just happened to be a sequence of
1306 letters, a binary zero, and a closing ] in the memory that followed.
1307
1308 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1309 over the years. Originally (before any Unicode support), just the bottom 8
1310 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1311 output from "man perlunicode" includes this:
1312
1313 The regular expression compiler produces polymorphic opcodes. That
1314 is, the pattern adapts to the data and automatically switches to
1315 the Unicode character scheme when presented with Unicode data--or
1316 instead uses a traditional byte scheme when presented with byte
1317 data.
1318
1319 Sadly, a wide octal escape does not cause a switch, and in a string with
1320 no other multibyte characters, these octal escapes are treated as before.
1321 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1322 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1323 Unicode string.
1324
1325 I have not perpetrated such confusion in PCRE. Up till now, it took just
1326 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1327 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1328 translate to the appropriate multibyte character.
1329
1330 29. Applied some refactoring to reduce the number of warnings from Microsoft
1331 and Borland compilers. This has included removing the fudge introduced
1332 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1333 a warning about an unused variable.
1334
1335 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1336 characters since release 4.0, because Perl (from release 5.004) does not.
1337 [Or at least, is documented not to: some releases seem to be in conflict
1338 with the documentation.] However, when a pattern was studied with
1339 pcre_study() and all its branches started with \s, PCRE still included VT
1340 as a possible starting character. Of course, this did no harm; it just
1341 caused an unnecessary match attempt.
1342
1343 22. Removed a now-redundant internal flag bit that recorded the fact that case
1344 dependency changed within the pattern. This was once needed for "required
1345 byte" processing, but is no longer used. This recovers a now-scarce options
1346 bit. Also moved the least significant internal flag bit to the most-
1347 significant bit of the word, which was not previously used (hangover from
1348 the days when it was an int rather than a uint) to free up another bit for
1349 the future.
1350
1351 23. Added support for CRLF line endings as well as CR and LF. As well as the
1352 default being selectable at build time, it can now be changed at runtime
1353 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1354 specify that it is scanning data with non-default line endings.
1355
1356 24. Changed the definition of CXXLINK to make it agree with the definition of
1357 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1358
1359 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1360 recursions. This makes a big different to stack usage for some patterns.
1361
1362 26. If a subpattern containing a named recursion or subroutine reference such
1363 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1364 the space required for the compiled pattern went wrong and gave too small a
1365 value. Depending on the environment, this could lead to "Failed: internal
1366 error: code overflow at offset 49" or "glibc detected double free or
1367 corruption" errors.
1368
1369 27. Applied patches from Google (a) to support the new newline modes and (b) to
1370 advance over multibyte UTF-8 characters in GlobalReplace.
1371
1372 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1373 difference for some implementation of PCRE in some Windows version.
1374
1375 29. Added some extra testing facilities to pcretest:
1376
1377 \q<number> in a data line sets the "match limit" value
1378 \Q<number> in a data line sets the "match recursion limt" value
1379 -S <number> sets the stack size, where <number> is in megabytes
1380
1381 The -S option isn't available for Windows.
1382
1383
1384 Version 6.6 06-Feb-06
1385 ---------------------
1386
1387 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1388 in pcreposix.h. I have copied the definition from pcre.h.
1389
1390 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1391 because pcre.h is no longer a built file.
1392
1393 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1394 not normally included in the compiled code.
1395
1396
1397 Version 6.5 01-Feb-06
1398 ---------------------
1399
1400 1. When using the partial match feature with pcre_dfa_exec(), it was not
1401 anchoring the second and subsequent partial matches at the new starting
1402 point. This could lead to incorrect results. For example, with the pattern
1403 /1234/, partially matching against "123" and then "a4" gave a match.
1404
1405 2. Changes to pcregrep:
1406
1407 (a) All non-match returns from pcre_exec() were being treated as failures
1408 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1409 error message is output. Some extra information is given for the
1410 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1411 probably the only errors that are likely to be caused by users (by
1412 specifying a regex that has nested indefinite repeats, for instance).
1413 If there are more than 20 of these errors, pcregrep is abandoned.
1414
1415 (b) A binary zero was treated as data while matching, but terminated the
1416 output line if it was written out. This has been fixed: binary zeroes
1417 are now no different to any other data bytes.
1418
1419 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1420 used to set a locale for matching. The --locale=xxxx long option has
1421 been added (no short equivalent) to specify a locale explicitly on the
1422 pcregrep command, overriding the environment variables.
1423
1424 (d) When -B was used with -n, some line numbers in the output were one less
1425 than they should have been.
1426
1427 (e) Added the -o (--only-matching) option.
1428
1429 (f) If -A or -C was used with -c (count only), some lines of context were
1430 accidentally printed for the final match.
1431
1432 (g) Added the -H (--with-filename) option.
1433
1434 (h) The combination of options -rh failed to suppress file names for files
1435 that were found from directory arguments.
1436
1437 (i) Added the -D (--devices) and -d (--directories) options.
1438
1439 (j) Added the -F (--fixed-strings) option.
1440
1441 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1442
1443 (l) Added the --colo(u)r option.
1444
1445 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1446 is not present by default.
1447
1448 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1449 items such as (?R) or (?1), when the recursion could match a number of
1450 alternatives. If it matched one of the alternatives, but subsequently,
1451 outside the recursion, there was a failure, the code tried to back up into
1452 the recursion. However, because of the way PCRE is implemented, this is not
1453 possible, and the result was an incorrect result from the match.
1454
1455 In order to prevent this happening, the specification of recursion has
1456 been changed so that all such subpatterns are automatically treated as
1457 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1458
1459 4. I had overlooked the fact that, in some locales, there are characters for
1460 which isalpha() is true but neither isupper() nor islower() are true. In
1461 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1462 and ordfeminine) are like this. This affected the treatment of \w and \W
1463 when they appeared in character classes, but not when they appeared outside
1464 a character class. The bit map for "word" characters is now created
1465 separately from the results of isalnum() instead of just taking it from the
1466 upper, lower, and digit maps. (Plus the underscore character, of course.)
1467
1468 5. The above bug also affected the handling of POSIX character classes such as
1469 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1470 permanent tables. Instead, the bit maps for such a class were previously
1471 created as the appropriate unions of the upper, lower, and digit bitmaps.
1472 Now they are created by subtraction from the [[:word:]] class, which has
1473 its own bitmap.
1474
1475 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1476 It is created by subtracting the vertical space characters (\x09, \x0a,
1477 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1478 subtraction was done in the overall bitmap for a character class, meaning
1479 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1480 be recognized. This bug has been fixed.
1481
1482 7. Patches from the folks at Google:
1483
1484 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1485 real life, but is still worth protecting against".
1486
1487 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1488 regular expressions".
1489
1490 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1491 have it.
1492
1493 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1494 "configure" and the latter not, in order to fix a problem somebody had
1495 with compiling the Arg class on HP-UX.
1496
1497 (e) Improve the error-handling of the C++ wrapper a little bit.
1498
1499 (f) New tests for checking recursion limiting.
1500
1501 8. The pcre_memmove() function, which is used only if the environment does not
1502 have a standard memmove() function (and is therefore rarely compiled),
1503 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1504 returning a result (though PCRE never actually uses the result).
1505
1506 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1507 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1508 returned instead of calling malloc() with an overflowing number that would
1509 most likely cause subsequent chaos.
1510
1511 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1512
1513 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1514 with this option is matched, the nmatch and pmatch options of regexec() are
1515 ignored.
1516
1517 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1518 provided in case anyone wants to the the POSIX interface with UTF-8
1519 strings.
1520
1521 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1522 C++ linking (needed for some HP-UX environments).
1523
1524 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1525 (unused parameter) and in the pcre_printint() function (omitted "default"
1526 switch label when the default is to do nothing).
1527
1528 15. Added some code to make it possible, when PCRE is compiled as a C++
1529 library, to replace subject pointers for pcre_exec() with a smart pointer
1530 class, thus making it possible to process discontinuous strings.
1531
1532 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1533 much the same function. They were added by different people who were trying
1534 to make PCRE easy to compile on non-Unix systems. It has been suggested
1535 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1536 for compiling on Windows systems. I have therefore replaced it with
1537 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1538 defaults to "extern" for C or "extern C" for C++, which works fine on
1539 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1540 SCOPE with something explicit in config.h. In addition:
1541
1542 (a) pcreposix.h still had just "extern" instead of either of these macros;
1543 I have replaced it with PCRE_DATA_SCOPE.
1544
1545 (b) Functions such as _pcre_xclass(), which are internal to the library,
1546 but external in the C sense, all had PCRE_EXPORT in their definitions.
1547 This is apparently wrong for the Windows case, so I have removed it.
1548 (It makes no difference on Unix-like systems.)
1549
1550 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1551 of recursive calls to match(). This is different to MATCH_LIMIT because
1552 that limits the total number of calls to match(), not all of which increase
1553 the depth of recursion. Limiting the recursion depth limits the amount of
1554 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1555 when PCRE is compiled, and changed at run time. A patch from Google adds
1556 this functionality to the C++ interface.
1557
1558 18. Changes to the handling of Unicode character properties:
1559
1560 (a) Updated the table to Unicode 4.1.0.
1561
1562 (b) Recognize characters that are not in the table as "Cn" (undefined).
1563
1564 (c) I revised the way the table is implemented to a much improved format
1565 which includes recognition of ranges. It now supports the ranges that
1566 are defined in UnicodeData.txt, and it also amalgamates other
1567 characters into ranges. This has reduced the number of entries in the
1568 table from around 16,000 to around 3,000, thus reducing its size
1569 considerably. I realized I did not need to use a tree structure after
1570 all - a binary chop search is just as efficient. Having reduced the
1571 number of entries, I extended their size from 6 bytes to 8 bytes to
1572 allow for more data.
1573
1574 (d) Added support for Unicode script names via properties such as \p{Han}.
1575
1576 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1577 matching that character.
1578
1579 20. When matching a repeated Unicode property with a minimum greater than zero,
1580 (for example \pL{2,}), PCRE could look past the end of the subject if it
1581 reached it while seeking the minimum number of characters. This could
1582 happen only if some of the characters were more than one byte long, because
1583 there is a check for at least the minimum number of bytes.
1584
1585 21. Refactored the implementation of \p and \P so as to be more general, to
1586 allow for more different types of property in future. This has changed the
1587 compiled form incompatibly. Anybody with saved compiled patterns that use
1588 \p or \P will have to recompile them.
1589
1590 22. Added "Any" and "L&" to the supported property types.
1591
1592 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1593 but give a compile time error if the value is greater than 0xff.
1594
1595 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1596 accidentally not being installed or uninstalled.
1597
1598 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1599 made were to insert the current release number. This seemed silly, because
1600 it made things harder for people building PCRE on systems that don't run
1601 "configure". I have turned pcre.h into a distributed file, no longer built
1602 by "configure", with the version identification directly included. There is
1603 no longer a pcre.h.in file.
1604
1605 However, this change necessitated a change to the pcre-config script as
1606 well. It is built from pcre-config.in, and one of the substitutions was the
1607 release number. I have updated configure.ac so that ./configure now finds
1608 the release number by grepping pcre.h.
1609
1610 26. Added the ability to run the tests under valgrind.
1611
1612
1613 Version 6.4 05-Sep-05
1614 ---------------------
1615
1616 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1617 "--" to be printed when multiple files were scanned, even when none of the
1618 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1619 consider it to be a bug, and have restored the previous behaviour.
1620
1621 2. A couple of code tidies to get rid of compiler warnings.
1622
1623 3. The pcretest program used to cheat by referring to symbols in the library
1624 whose names begin with _pcre_. These are internal symbols that are not
1625 really supposed to be visible externally, and in some environments it is
1626 possible to suppress them. The cheating is now confined to including
1627 certain files from the library's source, which is a bit cleaner.
1628
1629 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1630 file's purpose clearer.
1631
1632 5. Reorganized pcre_ucp_findchar().
1633
1634
1635 Version 6.3 15-Aug-05
1636 ---------------------
1637
1638 1. The file libpcre.pc.in did not have general read permission in the tarball.
1639
1640 2. There were some problems when building without C++ support:
1641
1642 (a) If C++ support was not built, "make install" and "make test" still
1643 tried to test it.
1644
1645 (b) There were problems when the value of CXX was explicitly set. Some
1646 changes have been made to try to fix these, and ...
1647
1648 (c) --disable-cpp can now be used to explicitly disable C++ support.
1649
1650 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1651 backslash in a target when C++ was disabled. This confuses some
1652 versions of "make", apparently. Using an intermediate variable solves
1653 this. (Same for CPP_LOBJ.)
1654
1655 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1656 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1657 necessary on certain architectures.
1658
1659 4. Added a setting of -export-symbols-regex to the link command to remove
1660 those symbols that are exported in the C sense, but actually are local
1661 within the library, and not documented. Their names all begin with
1662 "_pcre_". This is not a perfect job, because (a) we have to except some
1663 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1664 available (and never for static libraries). I have made a note to try to
1665 find a way round (a) in the future.
1666
1667
1668 Version 6.2 01-Aug-05
1669 ---------------------
1670
1671 1. There was no test for integer overflow of quantifier values. A construction
1672 such as {1111111111111111} would give undefined results. What is worse, if
1673 a minimum quantifier for a parenthesized subpattern overflowed and became
1674 negative, the calculation of the memory size went wrong. This could have
1675 led to memory overwriting.
1676
1677 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1678
1679 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1680 operating environments where this matters.
1681
1682 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1683 PCRE options from within the C++ wrapper.
1684
1685 5. Named capturing subpatterns were not being correctly counted when a pattern
1686 was compiled. This caused two problems: (a) If there were more than 100
1687 such subpatterns, the calculation of the memory needed for the whole
1688 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1689 back references of the form \12, where the number was greater than 9, were
1690 not recognized as back references, even though there were sufficient
1691 previous subpatterns.
1692
1693 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1694 versions of gcc, e.g. 2.95.4.
1695
1696
1697 Version 6.1 21-Jun-05
1698 ---------------------
1699
1700 1. There was one reference to the variable "posix" in pcretest.c that was not
1701 surrounded by "#if !defined NOPOSIX".
1702
1703 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1704 the cross-check on the old pcre_info() function, for the benefit of the
1705 cut-down version of PCRE that is currently imported into Exim.
1706
1707 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1708 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1709 patterns that start (?i) but I don't think that matters. The use of (?i) is
1710 just an example; this all applies to the other options as well.
1711
1712 4. Since libtool seems to echo the compile commands it is issuing, the output
1713 from "make" can be reduced a bit by putting "@" in front of each libtool
1714 compile command.
1715
1716 5. Patch from the folks at Google for configure.in to be a bit more thorough
1717 in checking for a suitable C++ installation before trying to compile the
1718 C++ stuff. This should fix a reported problem when a compiler was present,
1719 but no suitable headers.
1720
1721 6. The man pages all had just "PCRE" as their title. I have changed them to
1722 be the relevant file name. I have also arranged that these names are
1723 retained in the file doc/pcre.txt, which is a concatenation in text format
1724 of all the man pages except the little individual ones for each function.
1725
1726 7. The NON-UNIX-USE file had not been updated for the different set of source
1727 files that come with release 6. I also added a few comments about the C++
1728 wrapper.
1729
1730
1731 Version 6.0 07-Jun-05
1732 ---------------------
1733
1734 1. Some minor internal re-organization to help with my DFA experiments.
1735
1736 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1737 didn't matter for the library itself when fully configured, but did matter
1738 when compiling without UCP support, or within Exim, where the ucp files are
1739 not imported.
1740
1741 3. Refactoring of the library code to split up the various functions into
1742 different source modules. The addition of the new DFA matching code (see
1743 below) to a single monolithic source would have made it really too
1744 unwieldy, quite apart from causing all the code to be include in a
1745 statically linked application, when only some functions are used. This is
1746 relevant even without the DFA addition now that patterns can be compiled in
1747 one application and matched in another.
1748
1749 The downside of splitting up is that there have to be some external
1750 functions and data tables that are used internally in different modules of
1751 the library but which are not part of the API. These have all had their
1752 names changed to start with "_pcre_" so that they are unlikely to clash
1753 with other external names.
1754
1755 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1756 a different (DFA) algorithm. Although it is slower than the original
1757 function, it does have some advantages for certain types of matching
1758 problem.
1759
1760 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1761 including restarting after a partial match.
1762
1763 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1764 defined when compiling for Windows was sent to me. I have put it into the
1765 code, though I have no means of testing or verifying it.
1766
1767 7. Added the pcre_refcount() auxiliary function.
1768
1769 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1770 match before or at the first newline in the subject string. In pcretest,
1771 the /f option on a pattern can be used to set this.
1772
1773 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1774 would behave wrongly. This has been present in PCRE since release 4.0.
1775
1776 10. A number of changes to the pcregrep command:
1777
1778 (a) Refactored how -x works; insert ^(...)$ instead of setting
1779 PCRE_ANCHORED and checking the length, in preparation for adding
1780 something similar for -w.
1781
1782 (b) Added the -w (match as a word) option.
1783
1784 (c) Refactored the way lines are read and buffered so as to have more
1785 than one at a time available.
1786
1787 (d) Implemented a pcregrep test script.
1788
1789 (e) Added the -M (multiline match) option. This allows patterns to match
1790 over several lines of the subject. The buffering ensures that at least
1791 8K, or the rest of the document (whichever is the shorter) is available
1792 for matching (and similarly the previous 8K for lookbehind assertions).
1793
1794 (f) Changed the --help output so that it now says
1795
1796 -w, --word-regex(p)
1797
1798 instead of two lines, one with "regex" and the other with "regexp"
1799 because that confused at least one person since the short forms are the
1800 same. (This required a bit of code, as the output is generated
1801 automatically from a table. It wasn't just a text change.)
1802
1803 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1804 option but starts with a hyphen. Could be a pattern or a path name
1805 starting with a hyphen, for instance.
1806
1807 (h) "-" can be given as a file name to represent stdin.
1808
1809 (i) When file names are being printed, "(standard input)" is used for
1810 the standard input, for compatibility with GNU grep. Previously
1811 "<stdin>" was used.
1812
1813 (j) The option --label=xxx can be used to supply a name to be used for
1814 stdin when file names are being printed. There is no short form.
1815
1816 (k) Re-factored the options decoding logic because we are going to add
1817 two more options that take data. Such options can now be given in four
1818 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1819
1820 (l) Added the -A, -B, and -C options for requesting that lines of context
1821 around matches be printed.
1822
1823 (m) Added the -L option to print the names of files that do not contain
1824 any matching lines, that is, the complement of -l.
1825
1826 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1827 continue to scan other files.
1828
1829 (o) The -s option was incorrectly implemented. For compatibility with other
1830 greps, it now suppresses the error message for a non-existent or non-
1831 accessible file (but not the return code). There is a new option called
1832 -q that suppresses the output of matching lines, which was what -s was
1833 previously doing.
1834
1835 (p) Added --include and --exclude options to specify files for inclusion
1836 and exclusion when recursing.
1837
1838 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1839 Hopefully, it now does.
1840
1841 12. Missing cast in pcre_study().
1842
1843 13. Added an "uninstall" target to the makefile.
1844
1845 14. Replaced "extern" in the function prototypes in Makefile.in with
1846 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1847 world, but is set differently for Windows.
1848
1849 15. Added a second compiling function called pcre_compile2(). The only
1850 difference is that it has an extra argument, which is a pointer to an
1851 integer error code. When there is a compile-time failure, this is set
1852 non-zero, in addition to the error test pointer being set to point to an
1853 error message. The new argument may be NULL if no error number is required
1854 (but then you may as well call pcre_compile(), which is now just a
1855 wrapper). This facility is provided because some applications need a
1856 numeric error indication, but it has also enabled me to tidy up the way
1857 compile-time errors are handled in the POSIX wrapper.
1858
1859 16. Added VPATH=.libs to the makefile; this should help when building with one
1860 prefix path and installing with another. (Or so I'm told by someone who
1861 knows more about this stuff than I do.)
1862
1863 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1864 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1865 match everything, including newlines. This is not POSIX-compatible, but
1866 somebody wanted the feature. From pcretest it can be activated by using
1867 both the P and the s flags.
1868
1869 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1870
1871 19. libpcre.pc was being incorrectly installed as executable.
1872
1873 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1874 it now also looks for '\r' so that it will work unmodified on Windows.
1875
1876 21. Added Google's contributed C++ wrapper to the distribution.
1877
1878 22. Added some untidy missing memory free() calls in pcretest, to keep
1879 Electric Fence happy when testing.
1880
1881
1882
1883 Version 5.0 13-Sep-04
1884 ---------------------
1885
1886 1. Internal change: literal characters are no longer packed up into items
1887 containing multiple characters in a single byte-string. Each character
1888 is now matched using a separate opcode. However, there may be more than one
1889 byte in the character in UTF-8 mode.
1890
1891 2. The pcre_callout_block structure has two new fields: pattern_position and
1892 next_item_length. These contain the offset in the pattern to the next match
1893 item, and its length, respectively.
1894
1895 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1896 insertion of callouts before each pattern item. Added the /C option to
1897 pcretest to make use of this.
1898
1899 4. On the advice of a Windows user, the lines
1900
1901 #if defined(_WIN32) || defined(WIN32)
1902 _setmode( _fileno( stdout ), 0x8000 );
1903 #endif /* defined(_WIN32) || defined(WIN32) */
1904
1905 have been added to the source of pcretest. This apparently does useful
1906 magic in relation to line terminators.
1907
1908 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1909 for the benefit of those environments where the "b" makes a difference.
1910
1911 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1912 to know about it. I have put a hack into configure.in that adds in code
1913 to set GCC=yes if CC=icc. This seems to end up at a point in the
1914 generated configure script that is early enough to affect the setting of
1915 compiler options, which is what is needed, but I have no means of testing
1916 whether it really works. (The user who reported this had patched the
1917 generated configure script, which of course I cannot do.)
1918
1919 LATER: After change 22 below (new libtool files), the configure script
1920 seems to know about icc (and also ecc). Therefore, I have commented out
1921 this hack in configure.in.
1922
1923 7. Added support for pkg-config (2 patches were sent in).
1924
1925 8. Negated POSIX character classes that used a combination of internal tables
1926 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1927 [[:^ascii]]. Typically, they would match almost any characters. The other
1928 POSIX classes were not broken in this way.
1929
1930 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1931 to find the match, as PCRE was deluded into thinking that the match had to
1932 start at the start point or following a newline. The same bug applied to
1933 patterns with negative forward assertions or any backward assertions
1934 preceding ".*" at the start, unless the pattern required a fixed first
1935 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1936
1937 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1938 starting at the last subject character, bytes beyond the end of the subject
1939 string were read.
1940
1941 11. Renamed the variable "class" as "classbits" to make life easier for C++
1942 users. (Previously there was a macro definition, but it apparently wasn't
1943 enough.)
1944
1945 12. Added the new field "tables" to the extra data so that tables can be passed
1946 in at exec time, or the internal tables can be re-selected. This allows
1947 a compiled regex to be saved and re-used at a later time by a different
1948 program that might have everything at different addresses.
1949
1950 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1951 -R library as well as a -L library.
1952
1953 14. The debugging options of pcretest (-d on the command line or D on a
1954 pattern) showed incorrect output for anything following an extended class
1955 that contained multibyte characters and which was followed by a quantifier.
1956
1957 15. Added optional support for general category Unicode character properties
1958 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1959 support. It adds about 90K to the size of the library. The meanings of the
1960 inbuilt class escapes such as \d and \s have NOT been changed.
1961
1962 16. Updated pcredemo.c to include calls to free() to release the memory for the
1963 compiled pattern.
1964
1965 17. The generated file chartables.c was being created in the source directory
1966 instead of in the building directory. This caused the build to fail if the
1967 source directory was different from the building directory, and was
1968 read-only.
1969
1970 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1971 file. No doubt somebody will tell me if they don't make sense... Also added
1972 Dan Mooney's comments about building on OpenVMS.
1973
1974 19. Added support for partial matching via the PCRE_PARTIAL option for
1975 pcre_exec() and the \P data escape in pcretest.
1976
1977 20. Extended pcretest with 3 new pattern features:
1978
1979 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1980 write the compiled pattern to the file whose name is "rest-of-line".
1981 This is a straight binary dump of the data, with the saved pointer to
1982 the character tables forced to be NULL. The study data, if any, is
1983 written too. After writing, pcretest reads a new pattern.
1984
1985 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1986 compiled pattern from the given file. There must not be any
1987 occurrences of "<" in the file name (pretty unlikely); if there are,
1988 pcretest will instead treat the initial "<" as a pattern delimiter.
1989 After reading in the pattern, pcretest goes on to read data lines as
1990 usual.
1991
1992 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1993 and 16-bit fields in a compiled pattern, to simulate a pattern that
1994 was compiled on a host of opposite endianness.
1995
1996 21. The pcre-exec() function can now cope with patterns that were compiled on
1997 hosts of opposite endianness, with this restriction:
1998
1999 As for any compiled expression that is saved and used later, the tables
2000 pointer field cannot be preserved; the extra_data field in the arguments
2001 to pcre_exec() should be used to pass in a tables address if a value
2002 other than the default internal tables were used at compile time.
2003
2004 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2005 now diagnosed as an error. Previously, most of the time, a negative number
2006 would have been treated as zero, but if in addition "ovector" was passed as
2007 NULL, a crash could occur.
2008
2009 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2010 new versions from the libtool 1.5 distribution (the last one is a copy of
2011 a file called libtool.m4). This seems to have fixed the need to patch
2012 "configure" to support Darwin 1.3 (which I used to do). However, I still
2013 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2014 workstation).
2015
2016 24. Changed the PCRE licence to be the more standard "BSD" licence.
2017
2018
2019 Version 4.5 01-Dec-03
2020 ---------------------
2021
2022 1. There has been some re-arrangement of the code for the match() function so
2023 that it can be compiled in a version that does not call itself recursively.
2024 Instead, it keeps those local variables that need separate instances for
2025 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2026 needs to "recurse". Keeping track of where control must go is done by means
2027 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2028 hide most of the details from the main code, and operates only if
2029 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2030 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2031 operating.
2032
2033 To make it easier for callers to provide specially tailored get/free
2034 functions for this usage, two new functions, pcre_stack_malloc, and
2035 pcre_stack_free, are used. They are always called in strict stacking order,
2036 and the size of block requested is always the same.
2037
2038 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2039 PCRE has been compiled to use the stack or the heap for recursion. The
2040 -C option of pcretest uses this to show which version is compiled.
2041
2042 A new data escape \S, is added to pcretest; it causes the amounts of store
2043 obtained and freed by both kinds of malloc/free at match time to be added
2044 to the output.
2045
2046 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2047 what's available on my current Linux desktop machine.
2048
2049 3. When matching a UTF-8 string, the test for a valid string at the start has
2050 been extended. If start_offset is not zero, PCRE now checks that it points
2051 to a byte that is the start of a UTF-8 character. If not, it returns
2052 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2053 this is necessary because there may be backward assertions in the pattern.
2054 When matching the same subject several times, it may save resources to use
2055 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2056
2057 4. The code for checking the validity of UTF-8 strings has been tightened so
2058 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2059 containing "overlong sequences".
2060
2061 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2062 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2063 should have been "&", but it just so happened that all the cases this let
2064 through by mistake were picked up later in the function.
2065
2066 6. I had used a variable called "isblank" - this is a C99 function, causing
2067 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2068
2069 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2070 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2071 so the version is shown; (c) stop "make test" echoing "./RunTest".
2072
2073 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2074
2075 9. The prototype for memmove() for systems that don't have it was using
2076 size_t, but the inclusion of the header that defines size_t was later. I've
2077 moved the #includes for the C headers earlier to avoid this.
2078
2079 10. Added some adjustments to the code to make it easier to compiler on certain
2080 special systems:
2081
2082 (a) Some "const" qualifiers were missing.
2083 (b) Added the macro EXPORT before all exported functions; by default this
2084 is defined to be empty.
2085 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2086 that it reads its output file name as an argument instead of writing
2087 to the standard output and assuming this can be redirected.
2088
2089 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2090 class containing characters with values greater than 255, PCRE compilation
2091 went into a loop.
2092
2093 12. A recursive reference to a subpattern that was within another subpattern
2094 that had a minimum quantifier of zero caused PCRE to crash. For example,
2095 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2096 recursion. If the recursively-called subpattern itself had a zero repeat,
2097 that was OK.
2098
2099 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2100 buffer into which it was copied (for escape processing) was still set at
2101 1024, so long lines caused crashes.
2102
2103 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2104 "internal error: code overflow...". This applied to any character class
2105 that was followed by a possessive quantifier.
2106
2107 15. Modified the Makefile to add libpcre.la as a prerequisite for
2108 libpcreposix.la because I was told this is needed for a parallel build to
2109 work.
2110
2111 16. If a pattern that contained .* following optional items at the start was
2112 studied, the wrong optimizing data was generated, leading to matching
2113 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2114 matching string must start with a or b or c. The correct conclusion for
2115 this pattern is that a match can start with any character.
2116
2117
2118 Version 4.4 13-Aug-03
2119 ---------------------
2120
2121 1. In UTF-8 mode, a character class containing characters with values between
2122 127 and 255 was not handled correctly if the compiled pattern was studied.
2123 In fixing this, I have also improved the studying algorithm for such
2124 classes (slightly).
2125
2126 2. Three internal functions had redundant arguments passed to them. Removal
2127 might give a very teeny performance improvement.
2128
2129 3. Documentation bug: the value of the capture_top field in a callout is *one
2130 more than* the number of the hightest numbered captured substring.
2131
2132 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2133 in incorrectly linking with a previously installed version. They now link
2134 explicitly with libpcre.la.
2135
2136 5. configure.in no longer needs to recognize Cygwin specially.
2137
2138 6. A problem in pcre.in for Windows platforms is fixed.
2139
2140 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2141 pcretest, it used to include the size of the study block as part of its
2142 output. Unfortunately, the structure contains a field that has a different
2143 size on different hardware architectures. This meant that the tests that
2144 showed this size failed. As the block is currently always of a fixed size,
2145 this information isn't actually particularly useful in pcretest output, so
2146 I have just removed it.
2147
2148 8. Three pre-processor statements accidentally did not start in column 1.
2149 Sadly, there are *still* compilers around that complain, even though
2150 standard C has not required this for well over a decade. Sigh.
2151
2152 9. In pcretest, the code for checking callouts passed small integers in the
2153 callout_data field, which is a void * field. However, some picky compilers
2154 complained about the casts involved for this on 64-bit systems. Now
2155 pcretest passes the address of the small integer instead, which should get
2156 rid of the warnings.
2157
2158 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2159 both compile and run time, and gives an error if an invalid UTF-8 sequence
2160 is found. There is a option for disabling this check in cases where the
2161 string is known to be correct and/or the maximum performance is wanted.
2162
2163 11. In response to a bug report, I changed one line in Makefile.in from
2164
2165 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2166 to
2167 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2168
2169 to look similar to other lines, but I have no way of telling whether this
2170 is the right thing to do, as I do not use Windows. No doubt I'll get told
2171 if it's wrong...
2172
2173
2174 Version 4.3 21-May-03
2175 ---------------------
2176
2177 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2178 Makefile.
2179
2180 2. Some refactoring to improve the quality of the code:
2181
2182 (i) The utf8_table... variables are now declared "const".
2183
2184 (ii) The code for \cx, which used the "case flipping" table to upper case
2185 lower case letters, now just substracts 32. This is ASCII-specific,
2186 but the whole concept of \cx is ASCII-specific, so it seems
2187 reasonable.
2188
2189 (iii) PCRE was using its character types table to recognize decimal and
2190 hexadecimal digits in the pattern. This is silly, because it handles
2191 only 0-9, a-f, and A-F, but the character types table is locale-
2192 specific, which means strange things might happen. A private
2193 table is now used for this - though it costs 256 bytes, a table is
2194 much faster than multiple explicit tests. Of course, the standard
2195 character types table is still used for matching digits in subject
2196 strings against \d.
2197
2198 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2199 ending in _t are). So I've renamed it as ESC_tee.
2200
2201 3. The first argument for regexec() in the POSIX wrapper should have been
2202 defined as "const".
2203
2204 4. Changed pcretest to use malloc() for its buffers so that they can be
2205 Electric Fenced for debugging.
2206
2207 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2208 to read one or more bytes before the start of the subject string. Often this
2209 had no effect on PCRE's behaviour, but in some circumstances it could
2210 provoke a segmentation fault.
2211
2212 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2213 to try to read one or more bytes before the start of the subject string.
2214
2215 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2216 UTF-8 support could misbehave in various ways if the subject string
2217 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2218 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2219 back over UTF-8 characters.)
2220
2221
2222 Version 4.2 14-Apr-03
2223 ---------------------
2224
2225 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2226
2227 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2228 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2229 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2230 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2231 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2232 and BUILD_EXEEXT
2233 Note: automatic setting of the BUILD variables is not yet working
2234 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2235 compile-time but not at link-time
2236 [LINK]: use for linking executables only
2237 make different versions for Windows and non-Windows
2238 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2239 libraries
2240 [LINK_FOR_BUILD]: new variable
2241 [OBJEXT]: use throughout
2242 [EXEEXT]: use throughout
2243 <winshared>: new target
2244 <wininstall>: new target
2245 <dftables.o>: use native compiler
2246 <dftables>: use native linker
2247 <install>: handle Windows platform correctly
2248 <clean>: ditto
2249 <check>: ditto
2250 copy DLL to top builddir before testing
2251
2252 As part of these changes, -no-undefined was removed again. This was reported
2253 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2254 in any case.
2255
2256 3. Some tidies to get rid of compiler warnings:
2257
2258 . In the match_data structure, match_limit was an unsigned long int, whereas
2259 match_call_count was an int. I've made them both unsigned long ints.
2260
2261 . In pcretest the fact that a const uschar * doesn't automatically cast to
2262 a void * provoked a warning.
2263
2264 . Turning on some more compiler warnings threw up some "shadow" variables
2265 and a few more missing casts.
2266
2267 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2268 option, a class that contained a single character with a value between 128
2269 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2270
2271 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2272 option, a class that contained several characters, but with at least one
2273 whose value was between 128 and 255 caused PCRE to crash.
2274
2275
2276 Version 4.1 12-Mar-03
2277 ---------------------
2278
2279 1. Compiling with gcc -pedantic found a couple of places where casts were
2280 needed, and a string in dftables.c that was longer than standard compilers are
2281 required to support.
2282
2283 2. Compiling with Sun's compiler found a few more places where the code could
2284 be tidied up in order to avoid warnings.
2285
2286 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2287 first of these names is deprecated in the latest Autoconf in favour of the name
2288 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2289 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2290 analogy I have changed it to CFLAGS_FOR_BUILD.
2291
2292 4. Added -no-undefined to the linking command in the Makefile, because this is
2293 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2294 linking step for the pcreposix library.
2295
2296 5. PCRE was failing to diagnose the case of two named groups with the same
2297 name.
2298
2299 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2300 literal character that is needed in the subject for a match, and scans along to
2301 ensure that it is present before embarking on the full matching process. This
2302 saves time in cases of nested unlimited repeats that are never going to match.
2303 Problem: the scan can take a lot of time if the subject is very long (e.g.
2304 megabytes), thus penalizing straightforward matches. It is now done only if the
2305 amount of subject to be scanned is less than 1000 bytes.
2306
2307 7. A lesser problem with the same optimization is that it was recording the
2308 first character of an anchored pattern as "needed", thus provoking a search
2309 right along the subject, even when the first match of the pattern was going to
2310 fail. The "needed" character is now not set for anchored patterns, unless it
2311 follows something in the pattern that is of non-fixed length. Thus, it still
2312 fulfils its original purpose of finding quick non-matches in cases of nested
2313 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2314
2315
2316 Version 4.0 17-Feb-03
2317 ---------------------
2318
2319 1. If a comment in an extended regex that started immediately after a meta-item
2320 extended to the end of string, PCRE compiled incorrect data. This could lead to
2321 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2322
2323 2. Moved to autoconf 2.53 and libtool 1.4.2.
2324
2325 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2326 the special perltest8 script is no longer needed - all the tests can be run
2327 from a single perltest script.
2328
2329 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2330 by \s. It has now been removed in PCRE. This means it isn't recognized as
2331 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2332 class [:space:] *does* include VT, thereby creating a mess.
2333
2334 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2335 space and tab.
2336
2337 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2338 its new features into the main test script, reducing the number of scripts.
2339
2340 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2341 were backward compatible, and made the (?i) apply to the whole pattern, as if
2342 /i were given. Now it behaves more logically, and applies the option setting
2343 only to what follows. PCRE has been changed to follow suit. However, if it
2344 finds options settings right at the start of the pattern, it extracts them into
2345 the global options, as before. Thus, they show up in the info data.
2346
2347 8. Added support for the \Q...\E escape sequence. Characters in between are
2348 treated as literals. This is slightly different from Perl in that $ and @ are
2349 also handled as literals inside the quotes. In Perl, they will cause variable
2350 interpolation. Note the following examples:
2351
2352 Pattern PCRE matches Perl matches
2353
2354 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2355 \Qabc\$xyz\E abc\$xyz abc\$xyz
2356 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2357
2358 For compatibility with Perl, \Q...\E sequences are recognized inside character
2359 classes as well as outside them.
2360
2361 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2362 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2363 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2364 signed/unsigned warnings.
2365
2366 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2367 option for pcretest, so I've replaced it by a simple function that does just
2368 that job.
2369
2370 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2371 "pcregrep -".
2372
2373 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2374 Java package. This provides some syntactic sugar for simple cases of what my
2375 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2376 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2377 item, you can use this simplified notation. Note that only makes sense with
2378 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2379 greediness, whatever the setting of the PCRE_UNGREEDY option.
2380
2381 13. A change of greediness default within a pattern was not taking effect at
2382 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2383 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2384 was abstracted outside.
2385
2386 14. PCRE now supports the \G assertion. It is true when the current matching
2387 position is at the start point of the match. This differs from \A when the
2388 starting offset is non-zero. Used with the /g option of pcretest (or similar
2389 code), it works in the same way as it does for Perl's /g option. If all
2390 alternatives of a regex begin with \G, the expression is anchored to the start
2391 match position, and the "anchored" flag is set in the compiled expression.
2392
2393 15. Some bugs concerning the handling of certain option changes within patterns
2394 have been fixed. These applied to options other than (?ims). For example,
2395 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2396 been the other way round. Some of this was related to change 7 above.
2397
2398 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2399 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2400 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2401 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2402
2403 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2404 mode. Unlike ".", it always matches newline, whatever the setting of
2405 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2406 assertions. Perl allows it, but it doesn't (in general) work because it can't
2407 calculate the length of the lookbehind. At least, that's the case for Perl
2408 5.8.0 - I've been told they are going to document that it doesn't work in
2409 future.
2410
2411 18. Added an error diagnosis for escapes that PCRE does not support: these are
2412 \L, \l, \N, \P, \p, \U, \u, and \X.
2413
2414 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2415 reading past the end of the pattern in cases such as /[abcd/.
2416
2417 20. PCRE was getting more memory than necessary for patterns with classes that
2418 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2419
2420 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2421 compiling PCRE for use with Virtual Pascal.
2422
2423 22. Small fix to the Makefile to make it work properly if the build is done
2424 outside the source tree.
2425
2426 23. Added a new extension: a condition to go with recursion. If a conditional
2427 subpattern starts with (?(R) the "true" branch is used if recursion has
2428 happened, whereas the "false" branch is used only at the top level.
2429
2430 24. When there was a very long string of literal characters (over 255 bytes
2431 without UTF support, over 250 bytes with UTF support), the computation of how
2432 much memory was required could be incorrect, leading to segfaults or other
2433 strange effects.
2434
2435 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2436 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2437 there was a subsequent back reference to those brackets. This meant that, for
2438 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2439 possible to check for precisely this case. All we can do is abandon the
2440 optimization if .* occurs inside capturing brackets when there are any back
2441 references whatsoever. (See below for a better fix that came later.)
2442
2443 26. The handling of the optimization for finding the first character of a
2444 non-anchored pattern, and for finding a character that is required later in the
2445 match were failing in some cases. This didn't break the matching; it just
2446 failed to optimize when it could. The way this is done has been re-implemented.
2447
2448 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2449
2450 28. Added a new feature that provides some of the functionality that Perl
2451 provides with (?{...}). The facility is termed a "callout". The way it is done
2452 in PCRE is for the caller to provide an optional function, by setting
2453 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2454 global variable. By default it is unset, which disables all calling out. To get
2455 the function called, the regex must include (?C) at appropriate points. This
2456 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2457 This provides a means of identifying different callout points. When PCRE
2458 reaches such a point in the regex, if pcre_callout has been set, the external
2459 function is called. It is provided with data in a structure called
2460 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2461 matching continues; if it returns a non-zero value, the match at the current
2462 point fails. However, backtracking will occur if possible. [This was changed
2463 later and other features added - see item 49 below.]
2464
2465 29. pcretest is upgraded to test the callout functionality. It provides a
2466 callout function that displays information. By default, it shows the start of
2467 the match and the current position in the text. There are some new data escapes
2468 to vary what happens:
2469
2470 \C+ in addition, show current contents of captured substrings
2471 \C- do not supply a callout function
2472 \C!n return 1 when callout number n is reached
2473 \C!n!m return 1 when callout number n is reached for the mth time
2474
2475 30. If pcregrep was called with the -l option and just a single file name, it
2476 output "<stdin>" if a match was found, instead of the file name.
2477
2478 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2479 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2480 pcre_exec(). This saves a malloc/free per call. The default value of
2481 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2482 when configuring.
2483
2484 32. The default maximum size of a compiled pattern is 64K. There have been a
2485 few cases of people hitting this limit. The code now uses macros to handle the
2486 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2487 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2488 configuring. Tests 2 and 5 work only with 2-byte links because they output
2489 debugging information about compiled patterns.
2490
2491 33. Internal code re-arrangements:
2492
2493 (a) Moved the debugging function for printing out a compiled regex into
2494 its own source file (printint.c) and used #include to pull it into
2495 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2496 separate copies.
2497
2498 (b) Defined the list of op-code names for debugging as a macro in
2499 internal.h so that it is next to the definition of the opcodes.
2500
2501 (c) Defined a table of op-code lengths for simpler skipping along compiled
2502 code. This is again a macro in internal.h so that it is next to the
2503 definition of the opcodes.
2504
2505 34. Added support for recursive calls to individual subpatterns, along the
2506 lines of Robin Houston's patch (but implemented somewhat differently).
2507
2508 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2509 allow it to read and process whole directories in Win32. This code was
2510 contributed by Lionel Fourquaux; it has not been tested by me.
2511
2512 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2513 used to name a group. Names consist of alphanumerics and underscores, and must
2514 be unique. Back references use the syntax (?P=name) and recursive calls use
2515 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2516 numbers. The function pcre_fullinfo() can be used after compilation to extract
2517 a name/number map. There are three relevant calls:
2518
2519 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2520 PCRE_INFO_NAMECOUNT yields the number of entries
2521 PCRE_INFO_NAMETABLE yields a pointer to the map.
2522
2523 The map is a vector of fixed-size entries. The size of each entry depends on
2524 the length of the longest name used. The first two bytes of each entry are the
2525 group number, most significant byte first. There follows the corresponding
2526 name, zero terminated. The names are in alphabetical order.
2527
2528 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2529 case instead of 255. Making it the same both with and without UTF-8 support
2530 means that the same test output works with both.
2531
2532 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2533 calling malloc() with a zero argument.
2534
2535 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2536 optimization. I've improved things by keeping a bitmap of backreferences with
2537 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2538 fact referenced, the optimization can be applied. It is unlikely that a
2539 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2540 the match to follow \n) will appear inside brackets with a number greater than
2541 31, but if it does, any back reference > 31 suppresses the optimization.
2542
2543 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2544 of disabling numbered capturing parentheses. Any opening parenthesis that is
2545 not followed by ? behaves as if it were followed by ?: but named parentheses
2546 can still be used for capturing (and they will acquire numbers in the usual
2547 way).
2548
2549 41. Redesigned the return codes from the match() function into yes/no/error so
2550 that errors can be passed back from deep inside the nested calls. A malloc
2551 failure while inside a recursive subpattern call now causes the
2552 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2553
2554 42. It is now possible to set a limit on the number of times the match()
2555 function is called in a call to pcre_exec(). This facility makes it possible to
2556 limit the amount of recursion and backtracking, though not in a directly
2557 obvious way, because the match() function is used in a number of different
2558 circumstances. The count starts from zero for each position in the subject
2559 string (for non-anchored patterns). The default limit is, for compatibility, a
2560 large number, namely 10 000 000. You can change this in two ways:
2561
2562 (a) When configuring PCRE before making, you can use --with-match-limit=n
2563 to set a default value for the compiled library.
2564
2565 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2566 a different value is set. See 45 below.
2567
2568 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2569
2570 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2571 of things that can be changed at compile time. The first argument specifies
2572 what is wanted and the second points to where the information is to be placed.
2573 The current list of available information is:
2574
2575 PCRE_CONFIG_UTF8
2576
2577 The output is an integer that is set to one if UTF-8 support is available;
2578 otherwise it is set to zero.
2579
2580 PCRE_CONFIG_NEWLINE
2581
2582 The output is an integer that it set to the value of the code that is used for
2583 newline. It is either LF (10) or CR (13).
2584
2585 PCRE_CONFIG_LINK_SIZE
2586
2587 The output is an integer that contains the number of bytes used for internal
2588 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2589
2590 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2591
2592 The output is an integer that contains the threshold above which the POSIX
2593 interface uses malloc() for output vectors. See item 31 above.
2594
2595 PCRE_CONFIG_MATCH_LIMIT
2596
2597 The output is an unsigned integer that contains the default limit of the number
2598 of match() calls in a pcre_exec() execution. See 42 above.
2599
2600 44. pcretest has been upgraded by the addition of the -C option. This causes it
2601 to extract all the available output from the new pcre_config() function, and to
2602 output it. The program then exits immediately.
2603
2604 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2605 order to support additional features. One way would have been to define
2606 pcre_exec2() (for example) with extra arguments, but this would not have been
2607 extensible, and would also have required all calls to the original function to
2608 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2609 is used for passing in "extra" data from pcre_study().
2610
2611 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2612 contains the following fields:
2613
2614 flags a bitmap indicating which of the following fields are set
2615 study_data opaque data from pcre_study()
2616 match_limit a way of specifying a limit on match() calls for a specific
2617 call to pcre_exec()
2618 callout_data data for callouts (see 49 below)
2619
2620 The flag bits are also defined in pcre.h, and are
2621
2622 PCRE_EXTRA_STUDY_DATA
2623 PCRE_EXTRA_MATCH_LIMIT
2624 PCRE_EXTRA_CALLOUT_DATA
2625
2626 The pcre_study() function now returns one of these new pcre_extra blocks, with
2627 the actual study data pointed to by the study_data field, and the
2628 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2629 before. That is, this change is entirely upwards-compatible and requires no
2630 change to existing code.
2631
2632 If you want to pass in additional data to pcre_exec(), you can either place it
2633 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2634 block.
2635
2636 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2637 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2638 times with different match limits, until it finds the minimum value needed for
2639 pcre_exec() to complete. The value is then output. This can be instructive; for
2640 most simple matches the number is quite small, but for pathological cases it
2641 gets very large very quickly.
2642
2643 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2644 returns the size of the data block pointed to by the study_data field in a
2645 pcre_extra block, that is, the value that was passed as the argument to
2646 pcre_malloc() when PCRE was getting memory in which to place the information
2647 created by pcre_study(). The fourth argument should point to a size_t variable.
2648 pcretest has been extended so that this information is shown after a successful
2649 pcre_study() call when information about the compiled regex is being displayed.
2650
2651 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2652 because what follows is always an absolute path. (Later: it turns out that this
2653 is more than cosmetic for MinGW, because it doesn't like empty path
2654 components.)
2655
2656 49. Some changes have been made to the callout feature (see 28 above):
2657
2658 (i) A callout function now has three choices for what it returns:
2659
2660 0 => success, carry on matching
2661 > 0 => failure at this point, but backtrack if possible
2662 < 0 => serious error, return this value from pcre_exec()
2663
2664 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2665 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2666 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2667 use by callout functions. It will never be used by PCRE itself.
2668
2669 (ii) The pcre_extra structure (see 45 above) has a void * field called
2670 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2671 pcre_callout_block structure has a field of the same name. The contents of
2672 the field passed in the pcre_extra structure are passed to the callout
2673 function in the corresponding field in the callout block. This makes it
2674 easier to use the same callout-containing regex from multiple threads. For
2675 testing, the pcretest program has a new data escape
2676
2677 \C*n pass the number n (may be negative) as callout_data
2678
2679 If the callout function in pcretest receives a non-zero value as
2680 callout_data, it returns that value.
2681
2682 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2683 there were some redundant $(CFLAGS) in commands that are now specified as
2684 $(LINK), which already includes $(CFLAGS).
2685
2686 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2687 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2688 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2689 one-byte characters throughout. Note that case-insensitive matching applies
2690 only to characters whose values are less than 256. PCRE doesn't support the
2691 notion of cases for higher-valued characters.
2692
2693 (i) A character class whose characters are all within 0-255 is handled as
2694 a bit map, and the map is inverted for negative classes. Previously, a
2695 character > 255 always failed to match such a class; however it should
2696 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2697
2698 (ii) A negated character class with a single character < 255 is coded as
2699 "not this character" (OP_NOT). This wasn't working properly when the test
2700 character was multibyte, either singly or repeated.
2701
2702 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2703 mode, for example: \x{100}{2,3}.
2704
2705 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2706 singly or repeated) now correctly test multibyte characters. However,
2707 PCRE doesn't recognize any characters with values greater than 255 as
2708 digits, spaces, or word characters. Such characters always match \D, \S,
2709 and \W, and never match \d, \s, or \w.
2710
2711 (v) Classes may now contain characters and character ranges with values
2712 greater than 255. For example: [ab\x{100}-\x{400}].
2713
2714 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2715 PCRE in UTF-8 mode.
2716
2717 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2718 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2719 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2720 value.)
2721
2722 53. The single man page has become too large. I have therefore split it up into
2723 a number of separate man pages. These also give rise to individual HTML pages;
2724 these are now put in a separate directory, and there is an index.html page that
2725 lists them all. Some hyperlinking between the pages has been installed.
2726
2727 54. Added convenience functions for handling named capturing parentheses.
2728
2729 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2730 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2731 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2732 are faulted.
2733
2734 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2735 calling configure. These values are used when compiling the dftables.c program
2736 which is run to generate the source of the default character tables. They
2737 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2738 you will need to set these values.
2739
2740 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2741
2742
2743 Version 3.9 02-Jan-02
2744 ---------------------
2745
2746 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2747
2748 2. If --disable-static was given, the building process failed when trying to
2749 build pcretest and pcregrep. (For some reason it was using libtool to compile
2750 them, which is not right, as they aren't part of the library.)
2751
2752
2753 Version 3.8 18-Dec-01
2754 ---------------------
2755
2756 1. The experimental UTF-8 code was completely screwed up. It was packing the
2757 bytes in the wrong order. How dumb can you get?
2758
2759
2760 Version 3.7 29-Oct-01
2761 ---------------------
2762
2763 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2764 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2765 this didn't happen under Solaris 8, where I normally test things.
2766
2767 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2768 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2769 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2770 this fix an hour or so after the initial 3.7 release.)
2771
2772
2773 Version 3.6 23-Oct-01
2774 ---------------------
2775
2776 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2777 offsets passed as NULL with zero offset count.
2778
2779 2. The config.guess and config.sub files had not been updated when I moved to
2780 the latest autoconf.
2781
2782
2783 Version 3.5 15-Aug-01
2784 ---------------------
2785
2786 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2787 had been forgotten.
2788
2789 2. By using declared but undefined structures, we can avoid using "void"
2790 definitions in pcre.h while keeping the internal definitions of the structures
2791 private.
2792
2793 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2794 user point of view, this means that both static and shared libraries are built
2795 by default, but this can be individually controlled. More of the work of
2796 handling this static/shared cases is now inside libtool instead of PCRE's make
2797 file.
2798
2799 4. The pcretest utility is now installed along with pcregrep because it is
2800 useful for users (to test regexs) and by doing this, it automatically gets
2801 relinked by libtool. The documentation has been turned into a man page, so
2802 there are now .1, .txt, and .html versions in /doc.
2803
2804 5. Upgrades to pcregrep:
2805 (i) Added long-form option names like gnu grep.
2806 (ii) Added --help to list all options with an explanatory phrase.
2807 (iii) Added -r, --recursive to recurse into sub-directories.
2808 (iv) Added -f, --file to read patterns from a file.
2809
2810 6. pcre_exec() was referring to its "code" argument before testing that
2811 argument for NULL (and giving an error if it was NULL).
2812
2813 7. Upgraded Makefile.in to allow for compiling in a different directory from
2814 the source directory.
2815
2816 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2817 options bits, the pointer it was passed was to an int instead of to an unsigned
2818 long int. This mattered only on 64-bit systems.
2819
2820 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2821 generated) instead of pcre.in, which it its source. Also made the same change
2822 in several of the .c files.
2823
2824 10. A new release of gcc defines printf() as a macro, which broke pcretest
2825 because it had an ifdef in the middle of a string argument for printf(). Fixed
2826 by using separate calls to printf().
2827
2828 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2829 script, to force use of CR or LF instead of \n in the source. On non-Unix
2830 systems, the value can be set in config.h.
2831
2832 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2833 absolute limit. Changed the text of the error message to make this clear, and
2834 likewise updated the man page.
2835
2836 13. The limit of 99 on the number of capturing subpatterns has been removed.
2837 The new limit is 65535, which I hope will not be a "real" limit.
2838
2839
2840 Version 3.4 22-Aug-00
2841 ---------------------
2842
2843 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2844
2845 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2846
2847
2848 Version 3.3 01-Aug-00
2849 ---------------------
2850
2851 1. If an octal character was given, but the value was greater than \377, it
2852 was not getting masked to the least significant bits, as documented. This could
2853 lead to crashes in some systems.
2854
2855 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2856 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2857
2858 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2859 These just pass their arguments on to (pcre_free)(), but they are provided
2860 because some uses of PCRE bind it to non-C systems that can call its functions,
2861 but cannot call free() or pcre_free() directly.
2862
2863 4. Add "make test" as a synonym for "make check". Corrected some comments in
2864 the Makefile.
2865
2866 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2867 Makefile.
2868
2869 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2870 command called pgrep for grepping around the active processes.
2871
2872 7. Added the beginnings of support for UTF-8 character strings.
2873
2874 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2875 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2876 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2877 out for the ar command.)
2878
2879
2880 Version 3.2 12-May-00
2881 ---------------------
2882
2883 This is purely a bug fixing release.
2884
2885 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2886 of ZA. This was just one example of several cases that could provoke this bug,
2887 which was introduced by change 9 of version 2.00. The code for breaking
2888 infinite loops after an iteration that matches an empty string was't working
2889 correctly.
2890
2891 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2892 when matched against abbab (for example). After matching an empty string, it
2893 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2894 caused it to match further down the string than it should.
2895
2896 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2897 was there because it doesn't seem to be needed, and it causes trouble on some
2898 systems, as it is not a Standard C header. It has been removed.
2899
2900 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2901 were reported on the Macintosh. The changes were from
2902
2903 while ((c = *(++ptr)) != 0 && c != '\n');
2904 to
2905 while ((c = *(++ptr)) != 0 && c != '\n') ;
2906
2907 Totally extraordinary, but if that's what it takes...
2908
2909 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2910 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2911 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2912 assumes the way PCRE uses memmove() (always moving upwards).
2913
2914 6. PCRE is being used in one environment where strchr() is not available. There
2915 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2916 faster code anyway.
2917
2918
2919 Version 3.1 09-Feb-00
2920 ---------------------
2921
2922 The only change in this release is the fixing of some bugs in Makefile.in for
2923 the "install" target:
2924
2925 (1) It was failing to install pcreposix.h.
2926
2927 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2928
2929
2930 Version 3.0 01-Feb-00
2931 ---------------------
2932
2933 1. Add support for the /+ modifier to perltest (to output $` like it does in
2934 pcretest).
2935
2936 2. Add support for the /g modifier to perltest.
2937
2938 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2939 matches null strings.
2940
2941 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2942 pattern. Perl treats empty patterns specially - it reuses the most recent
2943 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2944 effect.
2945
2946 5. The POSIX interface was broken in that it was just handing over the POSIX
2947 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2948 required a bigger vector, with some working space on the end. This means that
2949 the POSIX wrapper now has to get and free some memory, and copy the results.
2950
2951 6. Added some simple autoconf support, placing the test data and the
2952 documentation in separate directories, re-organizing some of the
2953 information files, and making it build pcre-config (a GNU standard). Also added
2954 libtool support for building PCRE as a shared library, which is now the
2955 default.
2956
2957 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2958 09 are not valid octal constants. Single digits will be used for minor values
2959 less than 10.
2960
2961 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2962 existing programs that set these in the POSIX interface can use PCRE without
2963 modification.
2964
2965 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2966 return all that pcre_info() returns, plus additional data. The pcre_info()
2967 function is retained for compatibility, but is considered to be obsolete.
2968
2969 10. Added experimental recursion feature (?R) to handle one common case that
2970 Perl 5.6 will be able to do with (?p{...}).
2971
2972 11. Added support for POSIX character classes like [:alpha:], which Perl is
2973 adopting.
2974
2975
2976 Version 2.08 31-Aug-99
2977 ----------------------
2978
2979 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2980 trying to match at the startoffset position, but instead was moving forward to
2981 the next newline as if a previous match had failed.
2982
2983 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2984 and could get into a loop if a null string was matched other than at the start
2985 of the subject.
2986
2987 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2988 be distinguished at compile time, and for completeness also added PCRE_DATE.
2989
2990 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2991 in GnuWin32 environments.
2992
2993
2994 Version 2.07 29-Jul-99
2995 ----------------------
2996
2997 1. The documentation is now supplied in plain text form and HTML as well as in
2998 the form of man page sources.
2999
3000 2. C++ compilers don't like assigning (void *) values to other pointer types.
3001 In particular this affects malloc(). Although there is no problem in Standard
3002 C, I've put in casts to keep C++ compilers happy.
3003
3004 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3005 should be (const char *).
3006
3007 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3008 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3009 However, I haven't made this a standard facility. The documentation doesn't
3010 mention it, and the Makefile doesn't support it.
3011
3012 5. The Makefile now contains an "install" target, with editable destinations at
3013 the top of the file. The pcretest program is not installed.
3014
3015 6. pgrep -V now gives the PCRE version number and date.
3016
3017 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3018 causing the entire string to be ignored, instead of just the last character.
3019
3020 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3021 non-matching string, it can take a very, very long time, even for strings of
3022 quite modest length, because of the nested recursion. PCRE now does better in
3023 some of these cases. It does this by remembering the last required literal
3024 character in the pattern, and pre-searching the subject to ensure it is present
3025 before running the real match. In other words, it applies a heuristic to detect
3026 some types of certain failure quickly, and in the above example, if presented
3027 with a string that has no trailing " it gives "no match" very quickly.
3028
3029 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3030 other alternatives are tried instead.
3031
3032
3033 Version 2.06 09-Jun-99
3034 ----------------------
3035
3036 1. Change pcretest's output for amount of store used to show just the code
3037 space, because the remainder (the data block) varies in size between 32-bit and
3038 64-bit systems.
3039
3040 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3041 start matching at. This allows lookbehinds to work when searching for multiple
3042 occurrences in a string.
3043
3044 3. Added additional options to pcretest for testing multiple occurrences:
3045
3046 /+ outputs the rest of the string that follows a match
3047 /g loops for multiple occurrences, using the new startoffset argument
3048 /G loops for multiple occurrences by passing an incremented pointer
3049
3050 4. PCRE wasn't doing the "first character" optimization for patterns starting
3051 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3052 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3053 the letter 'x'. On long subject strings, this gives a significant speed-up.
3054
3055
3056 Version 2.05 21-Apr-99
3057 ----------------------
3058
3059 1. Changed the type of magic_number from int to long int so that it works
3060 properly on 16-bit systems.
3061
3062 2. Fixed a bug which caused patterns starting with .* not to work correctly
3063 when the subject string contained newline characters. PCRE was assuming
3064 anchoring for such patterns in all cases, which is not correct because .* will
3065 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3066 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3067 must be retried after every newline in the subject.
3068
3069
3070 Version 2.04 18-Feb-99
3071 ----------------------
3072
3073 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3074 computation of the store needed to hold the pattern was incorrect (too large).
3075 If such patterns were nested a few deep, this could multiply and become a real
3076 problem.
3077
3078 2. Added /M option to pcretest to show the memory requirement of a specific
3079 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3080
3081 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3082 compiled in such a way that the backtracking after subsequent failure was
3083 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3084 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3085
3086
3087 Version 2.03 02-Feb-99
3088 ----------------------
3089
3090 1. Fixed typo and small mistake in man page.
3091
3092 2. Added 4th condition (GPL supersedes if conflict) and created separate
3093 LICENCE file containing the conditions.
3094
3095 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3096 Perl, that is the internal \ allows the delimiter to be included in the
3097 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3098 the final delimiter, add \ to the end of the pattern (to test the error).
3099
3100 4. Added the convenience functions for extracting substrings after a successful
3101 match. Updated pcretest to make it able to test these functions.
3102
3103
3104 Version 2.02 14-Jan-99
3105 ----------------------
3106
3107 1. Initialized the working variables associated with each extraction so that
3108 their saving and restoring doesn't refer to uninitialized store.
3109
3110 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3111 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3112 fix the problem.
3113
3114 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3115 calls, and wasn't printing the correct value for compiling calls. Increased the
3116 default value of LOOPREPEAT, and the number of significant figures in the
3117 times.
3118
3119 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3120
3121 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3122 a building problem on Windows NT with a FAT file system.
3123
3124
3125 Version 2.01 21-Oct-98
3126 ----------------------
3127
3128 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3129 to character tables built by pcre_maketables() in the current locale. If NULL
3130 is passed, the default tables are used.
3131
3132
3133 Version 2.00 24-Sep-98
3134 ----------------------
3135
3136 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3137 it any more.
3138
3139 2. Allow quantification of (?>) groups, and make it work correctly.
3140
3141 3. The first character computation wasn't working for (?>) groups.
3142
3143 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3144 end of the subject) and add 5.005's \z, which really does match only at the
3145 very end of the subject.
3146
3147 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3148
3149 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3150 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3151 localized options. All options to pcre_study() were also removed.
3152
3153 7. Add other new features from 5.005:
3154
3155 $(?<= positive lookbehind
3156 $(?<! negative lookbehind
3157 (?imsx-imsx) added the unsetting capability
3158 such a setting is global if at outer level; local otherwise
3159 (?imsx-imsx:) non-capturing groups with option setting
3160 (?(cond)re|re) conditional pattern matching
3161
3162 A backreference to itself in a repeated group matches the previous
3163 captured string.
3164
3165 8. General tidying up of studying (both automatic and via "study")
3166 consequential on the addition of new assertions.
3167
3168 9. As in 5.005, unlimited repeated groups that could match an empty substring
3169 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3170 runtime if any iteration does actually match an empty substring.
3171
3172 10. Include the RunTest script in the distribution.
3173
3174 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3175 discrepancies, some of which were old and were also with respect to 5.004. They
3176 have now been fixed.
3177
3178
3179 Version 1.09 28-Apr-98
3180 ----------------------
3181
3182 1. A negated single character class followed by a quantifier with a minimum
3183 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3184 program crashes, or just wrong answers. This did not apply to negated classes
3185 containing more than one character, or to minima other than one.
3186
3187
3188 Version 1.08 27-Mar-98
3189 ----------------------
3190
3191 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3192
3193 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3194 latter must appear before anything that relies on it in the pattern.
3195
3196
3197 Version 1.07 16-Feb-98
3198 ----------------------
3199
3200 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3201 repeat of a potentially empty string).
3202
3203
3204 Version 1.06 23-Jan-98
3205 ----------------------
3206
3207 1. Added Markus Oberhumer's little patches for C++.
3208
3209 2. Literal strings longer than 255 characters were broken.
3210
3211
3212 Version 1.05 23-Dec-97
3213 ----------------------
3214
3215 1. Negated character classes containing more than one character were failing if
3216 PCRE_CASELESS was set at run time.
3217
3218
3219 Version 1.04 19-Dec-97
3220 ----------------------
3221
3222 1. Corrected the man page, where some "const" qualifiers had been omitted.
3223
3224 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3225 input syntax.
3226
3227 3. Fixed memory leak which occurred when a regex with back references was
3228 matched with an offsets vector that wasn't big enough. The temporary memory
3229 that is used in this case wasn't being freed if the match failed.
3230
3231 4. Tidied pcretest to ensure it frees memory that it gets.
3232
3233 5. Temporary memory was being obtained in the case where the passed offsets
3234 vector was exactly big enough.
3235
3236 6. Corrected definition of offsetof() from change 5 below.
3237
3238 7. I had screwed up change 6 below and broken the rules for the use of
3239 setjmp(). Now fixed.
3240
3241
3242 Version 1.03 18-Dec-97
3243 ----------------------
3244
3245 1. A erroneous regex with a missing opening parenthesis was correctly
3246 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3247 on some systems.
3248
3249 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3250 it was reported that one broken compiler failed on the former because "code" is
3251 also an independent variable.
3252
3253 3. The erroneous regex a[]b caused an array overrun reference.
3254
3255 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3256 fail on data ending with that character. (It was going on too far, and checking
3257 the next character, typically a binary zero.) This was specific to the
3258 optimized code for single-character negative classes.
3259
3260 5. Added a contributed patch from the TIN world which does the following:
3261
3262 + Add an undef for memmove, in case the the system defines a macro for it.
3263
3264 + Add a definition of offsetof(), in case there isn't one. (I don't know
3265 the reason behind this - offsetof() is part of the ANSI standard - but
3266 it does no harm).
3267
3268 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3269 most of the places where whitespace preceded '#'. I have given up and
3270 allowed the remaining 2 cases to be at the margin.
3271
3272 + Rename some variables in pcre to eliminate shadowing. This seems very
3273 pedantic, but does no harm, of course.
3274
3275 6. Moved the call to setjmp() into its own function, to get rid of warnings
3276 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3277
3278 7. Constructs such as \d{8,} were compiling into the equivalent of
3279 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3280 outcome, but in this particular case used more store than had been allocated,
3281 which caused the bug to be discovered because it threw up an internal error.
3282
3283 8. The debugging code in both pcre and pcretest for outputting the compiled
3284 form of a regex was going wrong in the case of back references followed by
3285 curly-bracketed repeats.
3286
3287
3288 Version 1.02 12-Dec-97
3289 ----------------------
3290
3291 1. Typos in pcre.3 and comments in the source fixed.
3292
3293 2. Applied a contributed patch to get rid of places where it used to remove
3294 'const' from variables, and fixed some signed/unsigned and uninitialized
3295 variable warnings.
3296
3297 3. Added the "runtest" target to Makefile.
3298
3299 4. Set default compiler flag to -O2 rather than just -O.
3300
3301
3302 Version 1.01 19-Nov-97
3303 ----------------------
3304
3305 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3306 like /([ab]*)*/, that is, for classes with more than one character in them.
3307
3308 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3309 as /((?>a*))*/ (a PCRE_EXTRA facility).
3310
3311
3312 Version 1.00 18-Nov-97
3313 ----------------------
3314
3315 1. Added compile-time macros to support systems such as SunOS4 which don't have
3316 memmove() or strerror() but have other things that can be used instead.
3317
3318 2. Arranged that "make clean" removes the executables.
3319
3320
3321 Version 0.99 27-Oct-97
3322 ----------------------
3323
3324 1. Fixed bug in code for optimizing classes with only one character. It was
3325 initializing a 32-byte map regardless, which could cause it to run off the end
3326 of the memory it had got.
3327
3328 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3329
3330
3331 Version 0.98 22-Oct-97
3332 ----------------------
3333
3334 1. Fixed bug in code for handling temporary memory usage when there are more
3335 back references than supplied space in the ovector. This could cause segfaults.
3336
3337
3338 Version 0.97 21-Oct-97
3339 ----------------------
3340
3341 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3342
3343 2. Optimized negated single characters not to use a bit map.
3344
3345 3. Brought error texts together as macro definitions; clarified some of them;
3346 fixed one that was wrong - it said "range out of order" when it meant "invalid
3347 escape sequence".
3348
3349 4. Changed some char * arguments to const char *.
3350
3351 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3352
3353 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3354 pcretest.
3355
3356
3357 Version 0.96 16-Oct-97
3358 ----------------------
3359
3360 1. Added a simple "pgrep" utility to the distribution.
3361
3362 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3363 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3364 where "ddd" means "one or more decimal digits".
3365
3366 3. Fixed serious bug. If a pattern had a back reference, but the call to
3367 pcre_exec() didn't supply a large enough ovector to record the related
3368 identifying subpattern, the match always failed. PCRE now remembers the number
3369 of the largest back reference, and gets some temporary memory in which to save
3370 the offsets during matching if necessary, in order to ensure that
3371 backreferences always work.
3372
3373 4. Increased the compatibility with Perl in a number of ways:
3374
3375 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3376 to request this handling. The option can be set at compile or exec time.
3377
3378 (b) $ matches before a terminating newline by default; an option
3379 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3380 mode). The option can be set at compile or exec time.
3381
3382 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3383 the same as Perl's. If the decimal number it represents is less than 10
3384 or there aren't that many previous left capturing parentheses, an octal
3385 escape is read. Inside a character class, it's always an octal escape,
3386 even if it is a single digit.
3387
3388 (d) An escaped but undefined alphabetic character is taken as a literal,
3389 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3390 escapes.
3391
3392 (e) {0} is now permitted. (The previous item is removed from the compiled
3393 pattern).
3394
3395 5. Changed all the names of code files so that the basic parts are no longer
3396 than 10 characters, and abolished the teeny "globals.c" file.
3397
3398 6. Changed the handling of character classes; they are now done with a 32-byte
3399 bit map always.
3400
3401 7. Added the -d and /D options to pcretest to make it possible to look at the
3402 internals of compilation without having to recompile pcre.
3403
3404
3405 Version 0.95 23-Sep-97
3406 ----------------------
3407
3408 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3409 \x20 at the start of a run of normal characters. These were being treated as
3410 real characters, instead of the source characters being re-checked.
3411
3412
3413 Version 0.94 18-Sep-97
3414 ----------------------
3415
3416 1. The functions are now thread-safe, with the caveat that the global variables
3417 containing pointers to malloc() and free() or alternative functions are the
3418 same for all threads.
3419
3420 2. Get pcre_study() to generate a bitmap of initial characters for non-
3421 anchored patterns when this is possible, and use it if passed to pcre_exec().
3422
3423
3424 Version 0.93 15-Sep-97
3425 ----------------------
3426
3427 1. /(b)|(:+)/ was computing an incorrect first character.
3428
3429 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3430 but not actually doing anything yet.
3431
3432 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3433 as Perl does (e.g. [-az] or [az-]).
3434
3435 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3436 all possible positions.
3437
3438 5. Split up into different modules to avoid including unneeded functions in a
3439 compiled binary. However, compile and exec are still in one module. The "study"
3440 function is split off.
3441
3442 6. The character tables are now in a separate module whose source is generated
3443 by an auxiliary program - but can then be edited by hand if required. There are
3444 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3445 toupper() in the code.
3446
3447 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3448 make them global. Abolish the function for setting them, as the caller can now
3449 set them directly.
3450
3451
3452 Version 0.92 11-Sep-97
3453 ----------------------
3454
3455 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3456 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3457
3458 2. Caseless matching was not working in character classes if the characters in
3459 the pattern were in upper case.
3460
3461 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3462
3463 4. Make PCRE_ANCHORED public and accept as a compile option.
3464
3465 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3466 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3467 pass them.
3468
3469 6. Give an error if bad option bits passed at compile or run time.
3470
3471 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3472 pcretest to cause it to pass that flag.
3473
3474 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3475 options, and the first character, if set.
3476
3477 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3478
3479
3480 Version 0.91 10-Sep-97
3481 ----------------------
3482
3483 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3484 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3485
3486 2. PCRE was looping on encountering an indefinitely repeated back reference to
3487 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3488 Perl does - treats the match as successful.
3489
3490 ****

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