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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 423 - (show annotations) (download)
Sat Aug 15 18:17:31 2009 UTC (5 years, 4 months ago) by ph10
File size: 167300 byte(s)
Add libpcreposix.pc for pkg-config, which was missing.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.00 ??-???-??
5 ----------------------
6
7 1. The table for translating pcre_compile() error codes into POSIX error codes
8 was out-of-date, and there was no check on the pcre_compile() error code
9 being within the table. This could lead to an OK return being given in
10 error.
11
12 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
13 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
14 in a Windows environment.
15
16 3. The pcregrep --count option prints the count for each file even when it is
17 zero, as does GNU grep. However, pcregrep was also printing all files when
18 --files-with-matches was added. Now, when both options are given, it prints
19 counts only for those files that have at least one match. (GNU grep just
20 prints the file name in this circumstance, but including the count seems
21 more useful - otherwise, why use --count?) Also ensured that the
22 combination -clh just lists non-zero counts, with no names.
23
24 4. The long form of the pcregrep -F option was incorrectly implemented as
25 --fixed_strings instead of --fixed-strings. This is an incompatible change,
26 but it seems right to fix it, and I didn't think it was worth preserving
27 the old behaviour.
28
29 5. The command line items --regex=pattern and --regexp=pattern were not
30 recognized by pcregrep, which required --regex pattern or --regexp pattern
31 (with a space rather than an '='). The man page documented the '=' forms,
32 which are compatible with GNU grep; these now work.
33
34 6. No libpcreposix.pc file was created for pkg-config; there was just
35 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
36
37
38 Version 7.9 11-Apr-09
39 ---------------------
40
41 1. When building with support for bzlib/zlib (pcregrep) and/or readline
42 (pcretest), all targets were linked against these libraries. This included
43 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
44 libraries. This caused unwanted dependencies to be created. This problem
45 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
46 pcretest is linked with readline.
47
48 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
49 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
50 moved outside it again, because FALSE and TRUE are already defined in AIX,
51 but BOOL is not.
52
53 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
54 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
55
56 4. The pcregrep documentation said spaces were inserted as well as colons (or
57 hyphens) following file names and line numbers when outputting matching
58 lines. This is not true; no spaces are inserted. I have also clarified the
59 wording for the --colour (or --color) option.
60
61 5. In pcregrep, when --colour was used with -o, the list of matching strings
62 was not coloured; this is different to GNU grep, so I have changed it to be
63 the same.
64
65 6. When --colo(u)r was used in pcregrep, only the first matching substring in
66 each matching line was coloured. Now it goes on to look for further matches
67 of any of the test patterns, which is the same behaviour as GNU grep.
68
69 7. A pattern that could match an empty string could cause pcregrep to loop; it
70 doesn't make sense to accept an empty string match in pcregrep, so I have
71 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
72 seems to be how GNU grep behaves.
73
74 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
75 start or after a newline", because the conditional assertion was not being
76 correctly handled. The rule now is that both the assertion and what follows
77 in the first alternative must satisfy the test.
78
79 9. If auto-callout was enabled in a pattern with a conditional group whose
80 condition was an assertion, PCRE could crash during matching, both with
81 pcre_exec() and pcre_dfa_exec().
82
83 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
84 used for matching.
85
86 11. Unicode property support in character classes was not working for
87 characters (bytes) greater than 127 when not in UTF-8 mode.
88
89 12. Added the -M command line option to pcretest.
90
91 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
92
93 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
94
95 16. Added comments and documentation about mis-use of no_arg in the C++
96 wrapper.
97
98 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
99 from Martin Jerabek that uses macro names for all relevant character and
100 string constants.
101
102 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
103 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
104 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
105 these, but not everybody uses configure.
106
107 19. A conditional group that had only one branch was not being correctly
108 recognized as an item that could match an empty string. This meant that an
109 enclosing group might also not be so recognized, causing infinite looping
110 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
111 with the subject "ab", where knowledge that the repeated group can match
112 nothing is needed in order to break the loop.
113
114 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
115 exec(), but without supplying a callout function, matching went wrong.
116
117 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
118 leak if the size of the offset vector was greater than 30. When the vector
119 is smaller, the saved offsets during recursion go onto a local stack
120 vector, but for larger vectors malloc() is used. It was failing to free
121 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
122 error, in fact).
123
124 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
125 heapframe that is used only when UTF-8 support is enabled. This caused no
126 problem, but was untidy.
127
128 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
129 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
130 included within another project.
131
132 24. Steven Van Ingelgem's patches to add more options to the CMake support,
133 slightly modified by me:
134
135 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
136 not building pcregrep.
137
138 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
139 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
140
141 25. Forward references, both numeric and by name, in patterns that made use of
142 duplicate group numbers, could behave incorrectly or give incorrect errors,
143 because when scanning forward to find the reference group, PCRE was not
144 taking into account the duplicate group numbers. A pattern such as
145 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
146
147 26. Changed a few more instances of "const unsigned char *" to USPTR, making
148 the feature of a custom pointer more persuasive (as requested by a user).
149
150 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
151 pcretest.c, inside #ifndefs, because it seems they are sometimes already
152 pre-defined.
153
154 28. Added support for (*UTF8) at the start of a pattern.
155
156 29. Arrange for flags added by the "release type" setting in CMake to be shown
157 in the configuration summary.
158
159
160 Version 7.8 05-Sep-08
161 ---------------------
162
163 1. Replaced UCP searching code with optimized version as implemented for Ad
164 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
165 stage table and inline lookup instead of a function, giving speed ups of 2
166 to 5 times on some simple patterns that I tested. Permission was given to
167 distribute the MultiStage2.py script that generates the tables (it's not in
168 the tarball, but is in the Subversion repository).
169
170 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
171 scripts.
172
173 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
174 a group with a zero qualifier. The result of the study could be incorrect,
175 or the function might crash, depending on the pattern.
176
177 4. Caseless matching was not working for non-ASCII characters in back
178 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
179 It now works when Unicode Property Support is available.
180
181 5. In pcretest, an escape such as \x{de} in the data was always generating
182 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
183 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
184 truncation.
185
186 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
187
188 7. Added two (int) casts to pcregrep when printing the difference of two
189 pointers, in case they are 64-bit values.
190
191 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
192 test 2 if it fails.
193
194 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
195 and a #define of that name to empty if it is not externally set. This is to
196 allow users of MSVC to set it if necessary.
197
198 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
199 the convenience functions in the pcre_get.c source file.
200
201 11. An option change at the start of a pattern that had top-level alternatives
202 could cause overwriting and/or a crash. This command provoked a crash in
203 some environments:
204
205 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
206
207 This potential security problem was recorded as CVE-2008-2371.
208
209 12. For a pattern where the match had to start at the beginning or immediately
210 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
211 pcre_dfa_exec() could read past the end of the passed subject if there was
212 no match. To help with detecting such bugs (e.g. with valgrind), I modified
213 pcretest so that it places the subject at the end of its malloc-ed buffer.
214
215 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
216 exec() might read past the end of the data buffer in UTF-8 mode.
217
218 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
219 the data contained the byte 0x85 as part of a UTF-8 character within its
220 first line. This applied both to normal and DFA matching.
221
222 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
223 /^[^d]*?$/8 failed to match "abc".
224
225 16. Added a missing copyright notice to pcrecpp_internal.h.
226
227 17. Make it more clear in the documentation that values returned from
228 pcre_exec() in ovector are byte offsets, not character counts.
229
230 18. Tidied a few places to stop certain compilers from issuing warnings.
231
232 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
233 supplied by Stefan Weber. I made a further small update for 7.8 because
234 there is a change of source arrangements: the pcre_searchfuncs.c module is
235 replaced by pcre_ucd.c.
236
237
238 Version 7.7 07-May-08
239 ---------------------
240
241 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
242 a string to a long long, pretend we don't even have a long long." This is
243 done by checking for the strtoq, strtoll, and _strtoi64 functions.
244
245 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
246 pre-7.6 versions, which defined a global no_arg variable instead of putting
247 it in the RE class. (See also #8 below.)
248
249 3. Remove a line of dead code, identified by coverity and reported by Nuno
250 Lopes.
251
252 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
253
254 (1) The include/exclude patterns were being applied to the whole pathnames
255 of files, instead of just to the final components.
256
257 (2) If there was more than one level of directory, the subdirectories were
258 skipped unless they satisfied the include/exclude conditions. This is
259 inconsistent with GNU grep (and could even be seen as contrary to the
260 pcregrep specification - which I improved to make it absolutely clear).
261 The action now is always to scan all levels of directory, and just
262 apply the include/exclude patterns to regular files.
263
264 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
265 --exclude_dir in the tests to avoid scanning .svn directories.
266
267 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
268 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
269 doesn't support NULs in patterns.
270
271 7. Added some missing "const"s to declarations of static tables in
272 pcre_compile.c and pcre_dfa_exec.c.
273
274 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
275 caused by fix #2 above. (Subsequently also a second patch to fix the
276 first patch. And a third patch - this was a messy problem.)
277
278 9. Applied Craig's patch to remove the use of push_back().
279
280 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
281 matching function regexec().
282
283 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
284 which, however, unlike Perl's \g{...}, are subroutine calls, not back
285 references. PCRE supports relative numbers with this syntax (I don't think
286 Oniguruma does).
287
288 12. Previously, a group with a zero repeat such as (...){0} was completely
289 omitted from the compiled regex. However, this means that if the group
290 was called as a subroutine from elsewhere in the pattern, things went wrong
291 (an internal error was given). Such groups are now left in the compiled
292 pattern, with a new opcode that causes them to be skipped at execution
293 time.
294
295 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
296 to the way PCRE behaves:
297
298 (a) A lone ] character is dis-allowed (Perl treats it as data).
299
300 (b) A back reference to an unmatched subpattern matches an empty string
301 (Perl fails the current match path).
302
303 (c) A data ] in a character class must be notated as \] because if the
304 first data character in a class is ], it defines an empty class. (In
305 Perl it is not possible to have an empty class.) The empty class []
306 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
307 The negative empty class [^] matches any one character, independently
308 of the DOTALL setting.
309
310 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
311 non-existent subpattern following a character class starting with ']' and
312 containing () gave an internal compiling error instead of "reference to
313 non-existent subpattern". Fortunately, when the pattern did exist, the
314 compiled code was correct. (When scanning forwards to check for the
315 existencd of the subpattern, it was treating the data ']' as terminating
316 the class, so got the count wrong. When actually compiling, the reference
317 was subsequently set up correctly.)
318
319 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
320 it was being rejected as not supported by pcre_dfa_exec(), even though
321 other assertions are supported. I have made pcre_dfa_exec() support
322 (*FAIL).
323
324 16. The implementation of 13c above involved the invention of a new opcode,
325 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
326 cannot be changed at match time, I realized I could make a small
327 improvement to matching performance by compiling OP_ALLANY instead of
328 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
329 on the OP_ANY path.
330
331 17. Compiling pcretest on Windows with readline support failed without the
332 following two fixes: (1) Make the unistd.h include conditional on
333 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
334
335 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
336 ncurses library to be included for pcretest when ReadLine support is
337 requested, but also to allow for it to be overridden. This patch came from
338 Daniel Bergström.
339
340 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
341 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
342 any errors with the current Unicode tables. Thanks to Peter Kankowski for
343 spotting this.
344
345
346 Version 7.6 28-Jan-08
347 ---------------------
348
349 1. A character class containing a very large number of characters with
350 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
351 overflow.
352
353 2. Patch to cut out the "long long" test in pcrecpp_unittest when
354 HAVE_LONG_LONG is not defined.
355
356 3. Applied Christian Ehrlicher's patch to update the CMake build files to
357 bring them up to date and include new features. This patch includes:
358
359 - Fixed PH's badly added libz and libbz2 support.
360 - Fixed a problem with static linking.
361 - Added pcredemo. [But later removed - see 7 below.]
362 - Fixed dftables problem and added an option.
363 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
364 HAVE_LONG_LONG.
365 - Added readline support for pcretest.
366 - Added an listing of the option settings after cmake has run.
367
368 4. A user submitted a patch to Makefile that makes it easy to create
369 "pcre.dll" under mingw when using Configure/Make. I added stuff to
370 Makefile.am that cause it to include this special target, without
371 affecting anything else. Note that the same mingw target plus all
372 the other distribution libraries and programs are now supported
373 when configuring with CMake (see 6 below) instead of with
374 Configure/Make.
375
376 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
377 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
378 exported in the Windows port". It has not yet been confirmed that the patch
379 solves the problem, but it does no harm.
380
381 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
382 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
383 with CMake, and also correct the comment about stack recursion.
384
385 7. Remove the automatic building of pcredemo from the ./configure system and
386 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
387 of a program that users should build themselves after PCRE is installed, so
388 building it automatically is not really right. What is more, it gave
389 trouble in some build environments.
390
391 8. Further tidies to CMakeLists.txt from Sheri and Christian.
392
393
394 Version 7.5 10-Jan-08
395 ---------------------
396
397 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
398 values in parens when parsing an RE using the C++ wrapper."
399
400 2. Negative specials like \S did not work in character classes in UTF-8 mode.
401 Characters greater than 255 were excluded from the class instead of being
402 included.
403
404 3. The same bug as (2) above applied to negated POSIX classes such as
405 [:^space:].
406
407 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
408 defined or documented. It seems to have been a typo for PCRE_STATIC, so
409 I have changed it.
410
411 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
412 first named subpattern) and a construct such as (?&a) would reference the
413 first named subpattern whose name started with "a" (in other words, the
414 length check was missing). Both these problems are fixed. "Subpattern name
415 expected" is now given for (?&) (a zero-length name), and this patch also
416 makes it give the same error for \k'' (previously it complained that that
417 was a reference to a non-existent subpattern).
418
419 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
420 this is right because (?- can be followed by option settings as well as by
421 digits. I have, however, made the messages clearer.
422
423 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
424 than the number used in the conditional) now cause a compile-time error.
425 This is actually not compatible with Perl, which accepts such patterns, but
426 treats the conditional as always being FALSE (as PCRE used to), but it
427 seems to me that giving a diagnostic is better.
428
429 8. Change "alphameric" to the more common word "alphanumeric" in comments
430 and messages.
431
432 9. Fix two occurrences of "backslash" in comments that should have been
433 "backspace".
434
435 10. Remove two redundant lines of code that can never be obeyed (their function
436 was moved elsewhere).
437
438 11. The program that makes PCRE's Unicode character property table had a bug
439 which caused it to generate incorrect table entries for sequences of
440 characters that have the same character type, but are in different scripts.
441 It amalgamated them into a single range, with the script of the first of
442 them. In other words, some characters were in the wrong script. There were
443 thirteen such cases, affecting characters in the following ranges:
444
445 U+002b0 - U+002c1
446 U+0060c - U+0060d
447 U+0061e - U+00612
448 U+0064b - U+0065e
449 U+0074d - U+0076d
450 U+01800 - U+01805
451 U+01d00 - U+01d77
452 U+01d9b - U+01dbf
453 U+0200b - U+0200f
454 U+030fc - U+030fe
455 U+03260 - U+0327f
456 U+0fb46 - U+0fbb1
457 U+10450 - U+1049d
458
459 12. The -o option (show only the matching part of a line) for pcregrep was not
460 compatible with GNU grep in that, if there was more than one match in a
461 line, it showed only the first of them. It now behaves in the same way as
462 GNU grep.
463
464 13. If the -o and -v options were combined for pcregrep, it printed a blank
465 line for every non-matching line. GNU grep prints nothing, and pcregrep now
466 does the same. The return code can be used to tell if there were any
467 non-matching lines.
468
469 14. Added --file-offsets and --line-offsets to pcregrep.
470
471 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
472 infinitely looping recursion. The bug was that positive lookaheads were not
473 being skipped when checking for a possible empty match (negative lookaheads
474 and both kinds of lookbehind were skipped).
475
476 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
477 inclusion of <windows.h> to before rather than after the definition of
478 INVALID_FILE_ATTRIBUTES (patch from David Byron).
479
480 17. Specifying a possessive quantifier with a specific limit for a Unicode
481 character property caused pcre_compile() to compile bad code, which led at
482 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
483 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
484 caused the error; without that there was no problem.
485
486 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
487
488 19. Added --enable-pcretest-libreadline.
489
490 20. In pcrecpp.cc, the variable 'count' was incremented twice in
491 RE::GlobalReplace(). As a result, the number of replacements returned was
492 double what it should be. I removed one of the increments, but Craig sent a
493 later patch that removed the other one (the right fix) and added unit tests
494 that check the return values (which was not done before).
495
496 21. Several CMake things:
497
498 (1) Arranged that, when cmake is used on Unix, the libraries end up with
499 the names libpcre and libpcreposix, not just pcre and pcreposix.
500
501 (2) The above change means that pcretest and pcregrep are now correctly
502 linked with the newly-built libraries, not previously installed ones.
503
504 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
505
506 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
507 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
508 UTF-8 newline character). The key issue is that the pattern starts .*;
509 this means that the match must be either at the beginning, or after a
510 newline. The bug was in the code for advancing after a failed match and
511 checking that the new position followed a newline. It was not taking
512 account of UTF-8 characters correctly.
513
514 23. PCRE was behaving differently from Perl in the way it recognized POSIX
515 character classes. PCRE was not treating the sequence [:...:] as a
516 character class unless the ... were all letters. Perl, however, seems to
517 allow any characters between [: and :], though of course it rejects as
518 unknown any "names" that contain non-letters, because all the known class
519 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
520 for example, whereas PCRE did not - it did not recognize a POSIX character
521 class. This seemed a bit dangerous, so the code has been changed to be
522 closer to Perl. The behaviour is not identical to Perl, because PCRE will
523 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
524 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
525 Perl does, and where it didn't before.
526
527 24. Rewrite so as to remove the single use of %n from pcregrep because in some
528 Windows environments %n is disabled by default.
529
530
531 Version 7.4 21-Sep-07
532 ---------------------
533
534 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
535 means that a class such as [\s] counted as "explicit reference to CR or
536 LF". That isn't really right - the whole point of the change was to try to
537 help when there was an actual mention of one of the two characters. So now
538 the change happens only if \r or \n (or a literal CR or LF) character is
539 encountered.
540
541 2. The 32-bit options word was also used for 6 internal flags, but the numbers
542 of both had grown to the point where there were only 3 bits left.
543 Fortunately, there was spare space in the data structure, and so I have
544 moved the internal flags into a new 16-bit field to free up more option
545 bits.
546
547 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
548 but did not set the internal JCHANGED flag - either of these is enough to
549 control the way the "get" function works - but the PCRE_INFO_JCHANGED
550 facility is supposed to tell if (?J) was ever used, so now (?J) at the
551 start sets both bits.
552
553 4. Added options (at build time, compile time, exec time) to change \R from
554 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
555
556 5. doc/pcresyntax.html was missing from the distribution.
557
558 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
559 compatibility, even though it is no longer used.
560
561 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
562 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
563 windows.h file is present (where different names are used). [This was
564 reversed later after testing - see 16 below.]
565
566 8. Changed all #include <config.h> to #include "config.h". There were also
567 some further <pcre.h> cases that I changed to "pcre.h".
568
569 9. When pcregrep was used with the --colour option, it missed the line ending
570 sequence off the lines that it output.
571
572 10. It was pointed out to me that arrays of string pointers cause lots of
573 relocations when a shared library is dynamically loaded. A technique of
574 using a single long string with a table of offsets can drastically reduce
575 these. I have refactored PCRE in four places to do this. The result is
576 dramatic:
577
578 Originally: 290
579 After changing UCP table: 187
580 After changing error message table: 43
581 After changing table of "verbs" 36
582 After changing table of Posix names 22
583
584 Thanks to the folks working on Gregex for glib for this insight.
585
586 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
587 unicode-properties was also set.
588
589 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
590
591 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
592 checked only for CRLF.
593
594 14. Added casts to pcretest.c to avoid compiler warnings.
595
596 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
597
598 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
599 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
600 entirely. This removes changes made in 7 above.
601
602 17. The CMake files have been updated, and there is now more information about
603 building with CMake in the NON-UNIX-USE document.
604
605
606 Version 7.3 28-Aug-07
607 ---------------------
608
609 1. In the rejigging of the build system that eventually resulted in 7.1, the
610 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
611 brackets there is not right, since it causes compilers to look for an
612 installed pcre.h, not the version that is in the source that is being
613 compiled (which of course may be different). I have changed it back to:
614
615 #include "pcre.h"
616
617 I have a vague recollection that the change was concerned with compiling in
618 different directories, but in the new build system, that is taken care of
619 by the VPATH setting the Makefile.
620
621 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
622 when the subject happened to end in the byte 0x85 (e.g. if the last
623 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
624 characters but of course it shouldn't be taken as a newline when it is part
625 of another character. The bug was that, for an unlimited repeat of . in
626 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
627 characters when looking for a newline.
628
629 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
630
631 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
632 in debug output.
633
634 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
635 long printing in the pcrecpp unittest when running under MinGW.
636
637 6. ESC_K was left out of the EBCDIC table.
638
639 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
640 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
641 limit also applies to "virtual nesting" when a pattern is recursive, and in
642 this case 1000 isn't so big. I have been able to remove this limit at the
643 expense of backing off one optimization in certain circumstances. Normally,
644 when pcre_exec() would call its internal match() function recursively and
645 immediately return the result unconditionally, it uses a "tail recursion"
646 feature to save stack. However, when a subpattern that can match an empty
647 string has an unlimited repetition quantifier, it no longer makes this
648 optimization. That gives it a stack frame in which to save the data for
649 checking that an empty string has been matched. Previously this was taken
650 from the 1000-entry workspace that had been reserved. So now there is no
651 explicit limit, but more stack is used.
652
653 8. Applied Daniel's patches to solve problems with the import/export magic
654 syntax that is required for Windows, and which was going wrong for the
655 pcreposix and pcrecpp parts of the library. These were overlooked when this
656 problem was solved for the main library.
657
658 9. There were some crude static tests to avoid integer overflow when computing
659 the size of patterns that contain repeated groups with explicit upper
660 limits. As the maximum quantifier is 65535, the maximum group length was
661 set at 30,000 so that the product of these two numbers did not overflow a
662 32-bit integer. However, it turns out that people want to use groups that
663 are longer than 30,000 bytes (though not repeat them that many times).
664 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
665 made it possible to implement the integer overflow checks in a much more
666 dynamic way, which I have now done. The artificial limitation on group
667 length has been removed - we now have only the limit on the total length of
668 the compiled pattern, which depends on the LINK_SIZE setting.
669
670 10. Fixed a bug in the documentation for get/copy named substring when
671 duplicate names are permitted. If none of the named substrings are set, the
672 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
673 empty string.
674
675 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
676 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
677 because the ] is interpreted as the first data character and the
678 terminating ] is not found. PCRE has been made compatible with Perl in this
679 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
680 cause memory overwriting.
681
682 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
683 string has been matched (to stop an infinite loop). It was not recognizing
684 a conditional subpattern that could match an empty string if that
685 subpattern was within another subpattern. For example, it looped when
686 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
687 condition was not nested. This bug has been fixed.
688
689 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
690 past the start of the subject in the presence of bytes with the top bit
691 set, for example "\x8aBCD".
692
693 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
694 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
695
696 14. Optimized (?!) to (*FAIL).
697
698 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
699 This restricts code points to be within the range 0 to 0x10FFFF, excluding
700 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
701 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
702 does: it's just the validity check that is more restrictive.
703
704 16. Inserted checks for integer overflows during escape sequence (backslash)
705 processing, and also fixed erroneous offset values for syntax errors during
706 backslash processing.
707
708 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
709 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
710
711 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
712 caused an overrun.
713
714 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
715 something other than just ASCII characters) inside a group that had an
716 unlimited repeat caused a loop at compile time (while checking to see
717 whether the group could match an empty string).
718
719 20. Debugging a pattern containing \p or \P could cause a crash. For example,
720 [\P{Any}] did so. (Error in the code for printing property names.)
721
722 21. An orphan \E inside a character class could cause a crash.
723
724 22. A repeated capturing bracket such as (A)? could cause a wild memory
725 reference during compilation.
726
727 23. There are several functions in pcre_compile() that scan along a compiled
728 expression for various reasons (e.g. to see if it's fixed length for look
729 behind). There were bugs in these functions when a repeated \p or \P was
730 present in the pattern. These operators have additional parameters compared
731 with \d, etc, and these were not being taken into account when moving along
732 the compiled data. Specifically:
733
734 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
735 length.
736
737 (b) An item such as \pL+ within a repeated group could cause crashes or
738 loops.
739
740 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
741 "reference to non-existent subpattern" error.
742
743 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
744
745 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
746 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
747
748 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
749
750 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
751 character were causing crashes (broken optimization).
752
753 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
754 \p or \P) caused a compile-time loop.
755
756 28. More problems have arisen in unanchored patterns when CRLF is a valid line
757 break. For example, the unstudied pattern [\r\n]A does not match the string
758 "\r\nA" because change 7.0/46 below moves the current point on by two
759 characters after failing to match at the start. However, the pattern \nA
760 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
761 the same is true. There doesn't seem any very clean way out of this, but
762 what I have chosen to do makes the common cases work: PCRE now takes note
763 of whether there can be an explicit match for \r or \n anywhere in the
764 pattern, and if so, 7.0/46 no longer applies. As part of this change,
765 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
766 pattern has explicit CR or LF references.
767
768 29. Added (*CR) etc for changing newline setting at start of pattern.
769
770
771 Version 7.2 19-Jun-07
772 ---------------------
773
774 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
775 which is apparently normally available under Windows.
776
777 2. Re-jig the pcregrep tests with different newline settings in an attempt
778 to make them independent of the local environment's newline setting.
779
780 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
781
782 4. Some of the "internals" tests were previously cut out when the link size
783 was not 2, because the output contained actual offsets. The recent new
784 "Z" feature of pcretest means that these can be cut out, making the tests
785 usable with all link sizes.
786
787 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
788 stack recursion. This gives a massive performance boost under BSD, but just
789 a small improvement under Linux. However, it saves one field in the frame
790 in all cases.
791
792 6. Added more features from the forthcoming Perl 5.10:
793
794 (a) (?-n) (where n is a string of digits) is a relative subroutine or
795 recursion call. It refers to the nth most recently opened parentheses.
796
797 (b) (?+n) is also a relative subroutine call; it refers to the nth next
798 to be opened parentheses.
799
800 (c) Conditions that refer to capturing parentheses can be specified
801 relatively, for example, (?(-2)... or (?(+3)...
802
803 (d) \K resets the start of the current match so that everything before
804 is not part of it.
805
806 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
807
808 (f) \g{name} is another synonym - part of Perl 5.10's unification of
809 reference syntax.
810
811 (g) (?| introduces a group in which the numbering of parentheses in each
812 alternative starts with the same number.
813
814 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
815
816 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
817 PCRE_INFO_JCHANGED.
818
819 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
820 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
821 for detecting groups that can match an empty string.
822
823 9. A pattern with a very large number of alternatives (more than several
824 hundred) was running out of internal workspace during the pre-compile
825 phase, where pcre_compile() figures out how much memory will be needed. A
826 bit of new cunning has reduced the workspace needed for groups with
827 alternatives. The 1000-alternative test pattern now uses 12 bytes of
828 workspace instead of running out of the 4096 that are available.
829
830 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
831
832 11. Applied patch from Google to remove an optimization that didn't quite work.
833 The report of the bug said:
834
835 pcrecpp::RE("a*").FullMatch("aaa") matches, while
836 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
837 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
838
839 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
840 it matched the wrong number of bytes.
841
842
843 Version 7.1 24-Apr-07
844 ---------------------
845
846 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
847 that is more "standard", making use of automake and other Autotools. There
848 is some re-arrangement of the files and adjustment of comments consequent
849 on this.
850
851 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
852 for recursive directory scanning broke on some systems because the files
853 are not scanned in any specific order and on different systems the order
854 was different. A call to "sort" has been inserted into RunGrepTest for the
855 approprate test as a short-term fix. In the longer term there may be an
856 alternative.
857
858 3. I had an email from Eric Raymond about problems translating some of PCRE's
859 man pages to HTML (despite the fact that I distribute HTML pages, some
860 people do their own conversions for various reasons). The problems
861 concerned the use of low-level troff macros .br and .in. I have therefore
862 removed all such uses from the man pages (some were redundant, some could
863 be replaced by .nf/.fi pairs). The 132html script that I use to generate
864 HTML has been updated to handle .nf/.fi and to complain if it encounters
865 .br or .in.
866
867 4. Updated comments in configure.ac that get placed in config.h.in and also
868 arranged for config.h to be included in the distribution, with the name
869 config.h.generic, for the benefit of those who have to compile without
870 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
871
872 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
873 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
874 makevp.bat for the current PCRE, using the additional files
875 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
876
877 6. A Windows user reported a minor discrepancy with test 2, which turned out
878 to be caused by a trailing space on an input line that had got lost in his
879 copy. The trailing space was an accident, so I've just removed it.
880
881 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
882 that is needed.
883
884 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
885 as "const" (a) because they are and (b) because it helps the PHP
886 maintainers who have recently made a script to detect big data structures
887 in the php code that should be moved to the .rodata section. I remembered
888 to update Builducptable as well, so it won't revert if ucptable.h is ever
889 re-created.
890
891 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
892 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
893 order to be able to cut out the UTF-8 tables in the latter when UTF-8
894 support is not required. This saves 1.5-2K of code, which is important in
895 some applications.
896
897 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
898 so as not to refer to the tables, even though these functions will never be
899 called when UTF-8 support is disabled. Otherwise there are problems with a
900 shared library.
901
902 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
903
904 (a) It was defining its arguments as char * instead of void *.
905
906 (b) It was assuming that all moves were upwards in memory; this was true
907 a long time ago when I wrote it, but is no longer the case.
908
909 The emulated memove() is provided for those environments that have neither
910 memmove() nor bcopy(). I didn't think anyone used it these days, but that
911 is clearly not the case, as these two bugs were recently reported.
912
913 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
914 and Detrail to create the HTML documentation, the .txt form of the man
915 pages, and it removes trailing spaces from listed files. It also creates
916 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
917 case, it wraps all the #defines with #ifndefs. This script should be run
918 before "make dist".
919
920 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
921 with Unicode property support.
922
923 (a) For a maximizing quantifier, if the two different cases of the
924 character were of different lengths in their UTF-8 codings (there are
925 some cases like this - I found 11), and the matching function had to
926 back up over a mixture of the two cases, it incorrectly assumed they
927 were both the same length.
928
929 (b) When PCRE was configured to use the heap rather than the stack for
930 recursion during matching, it was not correctly preserving the data for
931 the other case of a UTF-8 character when checking ahead for a match
932 while processing a minimizing repeat. If the check also involved
933 matching a wide character, but failed, corruption could cause an
934 erroneous result when trying to check for a repeat of the original
935 character.
936
937 13. Some tidying changes to the testing mechanism:
938
939 (a) The RunTest script now detects the internal link size and whether there
940 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
941 values substituted by "configure". (The RunGrepTest script already did
942 this for UTF-8.) The configure.ac script no longer substitutes the
943 relevant variables.
944
945 (b) The debugging options /B and /D in pcretest show the compiled bytecode
946 with length and offset values. This means that the output is different
947 for different internal link sizes. Test 2 is skipped for link sizes
948 other than 2 because of this, bypassing the problem. Unfortunately,
949 there was also a test in test 3 (the locale tests) that used /B and
950 failed for link sizes other than 2. Rather than cut the whole test out,
951 I have added a new /Z option to pcretest that replaces the length and
952 offset values with spaces. This is now used to make test 3 independent
953 of link size. (Test 2 will be tidied up later.)
954
955 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
956 segmentation fault instead of returning the appropriate error message.
957
958 15. In multiline mode when the newline sequence was set to "any", the pattern
959 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
960 This doesn't seem right; it now treats the CRLF combination as the line
961 ending, and so does not match in that case. It's only a pattern such as ^$
962 that would hit this one: something like ^ABC$ would have failed after \r
963 and then tried again after \r\n.
964
965 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
966 in an attempt to make files that differ only in their line terminators
967 compare equal. This works on Linux.
968
969 17. Under certain error circumstances pcregrep might try to free random memory
970 as it exited. This is now fixed, thanks to valgrind.
971
972 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
973 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
974 was because its rules for how to advance for /g after matching an empty
975 string at the end of a line did not allow for this case. They now check for
976 it specially.
977
978 20. pcretest is supposed to handle patterns and data of any length, by
979 extending its buffers when necessary. It was getting this wrong when the
980 buffer for a data line had to be extended.
981
982 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
983 CRLF as a newline sequence.
984
985 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
986 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
987 I have nevertheless tidied it up.
988
989 23. Added some casts to kill warnings from HP-UX ia64 compiler.
990
991 24. Added a man page for pcre-config.
992
993
994 Version 7.0 19-Dec-06
995 ---------------------
996
997 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
998 moving to gcc 4.1.1.
999
1000 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1001 sys/time.h, which is documented as needed for this function. It doesn't
1002 seem to matter on Linux, but it showed up on some releases of OS X.
1003
1004 3. It seems that there are systems where bytes whose values are greater than
1005 127 match isprint() in the "C" locale. The "C" locale should be the
1006 default when a C program starts up. In most systems, only ASCII printing
1007 characters match isprint(). This difference caused the output from pcretest
1008 to vary, making some of the tests fail. I have changed pcretest so that:
1009
1010 (a) When it is outputting text in the compiled version of a pattern, bytes
1011 other than 32-126 are always shown as hex escapes.
1012
1013 (b) When it is outputting text that is a matched part of a subject string,
1014 it does the same, unless a different locale has been set for the match
1015 (using the /L modifier). In this case, it uses isprint() to decide.
1016
1017 4. Fixed a major bug that caused incorrect computation of the amount of memory
1018 required for a compiled pattern when options that changed within the
1019 pattern affected the logic of the preliminary scan that determines the
1020 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1021 that the computed length was too small. The symptoms of this bug were
1022 either the PCRE error "internal error: code overflow" from pcre_compile(),
1023 or a glibc crash with a message such as "pcretest: free(): invalid next
1024 size (fast)". Examples of patterns that provoked this bug (shown in
1025 pcretest format) are:
1026
1027 /(?-x: )/x
1028 /(?x)(?-x: \s*#\s*)/
1029 /((?i)[\x{c0}])/8
1030 /(?i:[\x{c0}])/8
1031
1032 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1033 is now done differently.
1034
1035 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1036 wrapper classes; (b) implement a new function in the C++ scanner that is
1037 more efficient than the old way of doing things because it avoids levels of
1038 recursion in the regex matching; (c) add a paragraph to the documentation
1039 for the FullMatch() function.
1040
1041 6. The escape sequence \n was being treated as whatever was defined as
1042 "newline". Not only was this contrary to the documentation, which states
1043 that \n is character 10 (hex 0A), but it also went horribly wrong when
1044 "newline" was defined as CRLF. This has been fixed.
1045
1046 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1047 was being set to -1 for the "end of line" case (supposedly a value that no
1048 character can have). Though this value is never used (the check for end of
1049 line is "zero bytes in current character"), it caused compiler complaints.
1050 I've changed it to 0xffffffff.
1051
1052 8. In pcre_version.c, the version string was being built by a sequence of
1053 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1054 string (as it is for production releases) called a macro with an empty
1055 argument. The C standard says the result of this is undefined. The gcc
1056 compiler treats it as an empty string (which was what was wanted) but it is
1057 reported that Visual C gives an error. The source has been hacked around to
1058 avoid this problem.
1059
1060 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1061 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1062 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1063 of them did).
1064
1065 10. Originally, pcretest opened its input and output without "b"; then I was
1066 told that "b" was needed in some environments, so it was added for release
1067 5.0 to both the input and output. (It makes no difference on Unix-like
1068 systems.) Later I was told that it is wrong for the input on Windows. I've
1069 now abstracted the modes into two macros, to make it easier to fiddle with
1070 them, and removed "b" from the input mode under Windows.
1071
1072 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1073
1074 12. Added -help and --help to pcretest as an official way of being reminded
1075 of the options.
1076
1077 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1078 and pcrecpp.cc because they annoy compilers at high warning levels.
1079
1080 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1081
1082 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1083 scripts are not C programs :-) and which was not noticed because it works
1084 on Linux.
1085
1086 16. pcretest is supposed to handle any length of pattern and data line (as one
1087 line or as a continued sequence of lines) by extending its input buffer if
1088 necessary. This feature was broken for very long pattern lines, leading to
1089 a string of junk being passed to pcre_compile() if the pattern was longer
1090 than about 50K.
1091
1092 17. I have done a major re-factoring of the way pcre_compile() computes the
1093 amount of memory needed for a compiled pattern. Previously, there was code
1094 that made a preliminary scan of the pattern in order to do this. That was
1095 OK when PCRE was new, but as the facilities have expanded, it has become
1096 harder and harder to keep it in step with the real compile phase, and there
1097 have been a number of bugs (see for example, 4 above). I have now found a
1098 cunning way of running the real compile function in a "fake" mode that
1099 enables it to compute how much memory it would need, while actually only
1100 ever using a few hundred bytes of working memory and without too many
1101 tests of the mode. This should make future maintenance and development
1102 easier. A side effect of this work is that the limit of 200 on the nesting
1103 depth of parentheses has been removed (though this was never a serious
1104 limitation, I suspect). However, there is a downside: pcre_compile() now
1105 runs more slowly than before (30% or more, depending on the pattern). I
1106 hope this isn't a big issue. There is no effect on runtime performance.
1107
1108 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1109 newline (only possible for the last line of a file) and it was a
1110 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1111
1112 19. Added additional timing features to pcretest. (1) The -tm option now times
1113 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1114 separate command line item, by a number that specifies the number of
1115 repeats to use when timing. The default is 50000; this gives better
1116 precision, but takes uncomfortably long for very large patterns.
1117
1118 20. Extended pcre_study() to be more clever in cases where a branch of a
1119 subpattern has no definite first character. For example, (a*|b*)[cd] would
1120 previously give no result from pcre_study(). Now it recognizes that the
1121 first character must be a, b, c, or d.
1122
1123 21. There was an incorrect error "recursive call could loop indefinitely" if
1124 a subpattern (or the entire pattern) that was being tested for matching an
1125 empty string contained only one non-empty item after a nested subpattern.
1126 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1127 incorrectly, because the \d was being skipped in the check.
1128
1129 22. The pcretest program now has a new pattern option /B and a command line
1130 option -b, which is equivalent to adding /B to every pattern. This causes
1131 it to show the compiled bytecode, without the additional information that
1132 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1133 is the same as /B/I).
1134
1135 23. A new optimization is now able automatically to treat some sequences such
1136 as a*b as a*+b. More specifically, if something simple (such as a character
1137 or a simple class like \d) has an unlimited quantifier, and is followed by
1138 something that cannot possibly match the quantified thing, the quantifier
1139 is automatically "possessified".
1140
1141 24. A recursive reference to a subpattern whose number was greater than 39
1142 went wrong under certain circumstances in UTF-8 mode. This bug could also
1143 have affected the operation of pcre_study().
1144
1145 25. Realized that a little bit of performance could be had by replacing
1146 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1147
1148 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1149
1150 27. Possessive quantifiers such as a++ were previously implemented by turning
1151 them into atomic groups such as ($>a+). Now they have their own opcodes,
1152 which improves performance. This includes the automatically created ones
1153 from 23 above.
1154
1155 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1156 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1157 the first matched character to be a colon. This applied both to named and
1158 numbered groups.
1159
1160 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1161
1162 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1163 building PCRE on Windows easier, so I have included it in the distribution.
1164
1165 31. There is now a check in pcretest against a ridiculously large number being
1166 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1167 loop, the loop is abandoned.
1168
1169 32. Forward references to subpatterns in conditions such as (?(2)...) where
1170 subpattern 2 is defined later cause pcre_compile() to search forwards in
1171 the pattern for the relevant set of parentheses. This search went wrong
1172 when there were unescaped parentheses in a character class, parentheses
1173 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1174
1175 33. "Subroutine" calls and backreferences were previously restricted to
1176 referencing subpatterns earlier in the regex. This restriction has now
1177 been removed.
1178
1179 34. Added a number of extra features that are going to be in Perl 5.10. On the
1180 whole, these are just syntactic alternatives for features that PCRE had
1181 previously implemented using the Python syntax or my own invention. The
1182 other formats are all retained for compatibility.
1183
1184 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1185 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1186 also .NET compatible.
1187
1188 (b) A recursion or subroutine call to a named group can now be defined as
1189 (?&name) as well as (?P>name).
1190
1191 (c) A backreference to a named group can now be defined as \k<name> or
1192 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1193 5.10, are also .NET compatible.
1194
1195 (d) A conditional reference to a named group can now use the syntax
1196 (?(<name>) or (?('name') as well as (?(name).
1197
1198 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1199 groups (named and numbered) that are never evaluated inline, but can be
1200 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1201 is always false. There may be only one alternative in such a group.
1202
1203 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1204 as the simple (?(R). The condition is true only if the most recent
1205 recursion is that of the given number or name. It does not search out
1206 through the entire recursion stack.
1207
1208 (g) The escape \gN or \g{N} has been added, where N is a positive or
1209 negative number, specifying an absolute or relative reference.
1210
1211 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1212 some "unreachable code" warnings.
1213
1214 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1215 things, this adds five new scripts.
1216
1217 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1218 There were also incompatibilities regarding the handling of \Q..\E inside
1219 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1220 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1221
1222 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1223 matches an empty string, and forcibly breaks the loop. There were bugs in
1224 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1225 against aaaa the result was just "a" rather than "aaaa", for example. Two
1226 separate and independent bugs (that affected different cases) have been
1227 fixed.
1228
1229 39. Refactored the code to abolish the use of different opcodes for small
1230 capturing bracket numbers. This is a tidy that I avoided doing when I
1231 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1232 The new approach is not only tidier, it makes it possible to reduce the
1233 memory needed to fix the previous bug (38).
1234
1235 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1236 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1237 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1238 mode.
1239
1240 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1241 report.
1242
1243 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1244 copy construction and assignment in the C++ wrapper.
1245
1246 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1247 couple of bugs that could have given wrong results in the "--newline=crlf"
1248 case.
1249
1250 44. Added a number of casts and did some reorganization of signed/unsigned int
1251 variables following suggestions from Dair Grant. Also renamed the variable
1252 "this" as "item" because it is a C++ keyword.
1253
1254 45. Arranged for dftables to add
1255
1256 #include "pcre_internal.h"
1257
1258 to pcre_chartables.c because without it, gcc 4.x may remove the array
1259 definition from the final binary if PCRE is built into a static library and
1260 dead code stripping is activated.
1261
1262 46. For an unanchored pattern, if a match attempt fails at the start of a
1263 newline sequence, and the newline setting is CRLF or ANY, and the next two
1264 characters are CRLF, advance by two characters instead of one.
1265
1266
1267 Version 6.7 04-Jul-06
1268 ---------------------
1269
1270 1. In order to handle tests when input lines are enormously long, pcretest has
1271 been re-factored so that it automatically extends its buffers when
1272 necessary. The code is crude, but this _is_ just a test program. The
1273 default size has been increased from 32K to 50K.
1274
1275 2. The code in pcre_study() was using the value of the re argument before
1276 testing it for NULL. (Of course, in any sensible call of the function, it
1277 won't be NULL.)
1278
1279 3. The memmove() emulation function in pcre_internal.h, which is used on
1280 systems that lack both memmove() and bcopy() - that is, hardly ever -
1281 was missing a "static" storage class specifier.
1282
1283 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1284 containing an extended class (one that cannot be represented by a bitmap
1285 because it contains high-valued characters or Unicode property items, e.g.
1286 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1287 pattern, but PCRE should not loop if you do not (it no longer does).
1288 [Detail: two cases were found: (a) a repeated subpattern containing an
1289 extended class; (b) a recursive reference to a subpattern that followed a
1290 previous extended class. It wasn't skipping over the extended class
1291 correctly when UTF-8 mode was not set.]
1292
1293 5. A negated single-character class was not being recognized as fixed-length
1294 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1295 compile error "lookbehind assertion is not fixed length".
1296
1297 6. The RunPerlTest auxiliary script was showing an unexpected difference
1298 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1299 write a Perl script that can interpret lines of an input file either as
1300 byte characters or as UTF-8, which is what "perltest" was being required to
1301 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1302 can't do is switch easily at run time between having the "use utf8;" pragma
1303 or not. In the end, I fudged it by using the RunPerlTest script to insert
1304 "use utf8;" explicitly for the UTF-8 tests.
1305
1306 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1307 the end of the subject string, contrary to the documentation and to what
1308 Perl does. This was true of both matching functions. Now it matches only at
1309 the start of the subject and immediately after *internal* newlines.
1310
1311 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1312 a pointer to an int instead of a pointer to an unsigned long int. This
1313 caused problems on 64-bit systems.
1314
1315 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1316 instance of the 'standard' template library not being so standard".
1317
1318 10. There was no check on the number of named subpatterns nor the maximum
1319 length of a subpattern name. The product of these values is used to compute
1320 the size of the memory block for a compiled pattern. By supplying a very
1321 long subpattern name and a large number of named subpatterns, the size
1322 computation could be caused to overflow. This is now prevented by limiting
1323 the length of names to 32 characters, and the number of named subpatterns
1324 to 10,000.
1325
1326 11. Subpatterns that are repeated with specific counts have to be replicated in
1327 the compiled pattern. The size of memory for this was computed from the
1328 length of the subpattern and the repeat count. The latter is limited to
1329 65535, but there was no limit on the former, meaning that integer overflow
1330 could in principle occur. The compiled length of a repeated subpattern is
1331 now limited to 30,000 bytes in order to prevent this.
1332
1333 12. Added the optional facility to have named substrings with the same name.
1334
1335 13. Added the ability to use a named substring as a condition, using the
1336 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1337 are numbers (not recommended). Forward references are permitted.
1338
1339 14. Added forward references in named backreferences (if you see what I mean).
1340
1341 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1342 pattern could run off the end of the subject. For example, the pattern
1343 "(?s)(.{1,5})"8 did this with the subject "ab".
1344
1345 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1346 PCRE_CASELESS was set when matching characters that were quantified with ?
1347 or *.
1348
1349 17. A character class other than a single negated character that had a minimum
1350 but no maximum quantifier - for example [ab]{6,} - was not handled
1351 correctly by pce_dfa_exec(). It would match only one character.
1352
1353 18. A valid (though odd) pattern that looked like a POSIX character
1354 class but used an invalid character after [ (for example [[,abc,]]) caused
1355 pcre_compile() to give the error "Failed: internal error: code overflow" or
1356 in some cases to crash with a glibc free() error. This could even happen if
1357 the pattern terminated after [[ but there just happened to be a sequence of
1358 letters, a binary zero, and a closing ] in the memory that followed.
1359
1360 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1361 over the years. Originally (before any Unicode support), just the bottom 8
1362 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1363 output from "man perlunicode" includes this:
1364
1365 The regular expression compiler produces polymorphic opcodes. That
1366 is, the pattern adapts to the data and automatically switches to
1367 the Unicode character scheme when presented with Unicode data--or
1368 instead uses a traditional byte scheme when presented with byte
1369 data.
1370
1371 Sadly, a wide octal escape does not cause a switch, and in a string with
1372 no other multibyte characters, these octal escapes are treated as before.
1373 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1374 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1375 Unicode string.
1376
1377 I have not perpetrated such confusion in PCRE. Up till now, it took just
1378 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1379 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1380 translate to the appropriate multibyte character.
1381
1382 29. Applied some refactoring to reduce the number of warnings from Microsoft
1383 and Borland compilers. This has included removing the fudge introduced
1384 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1385 a warning about an unused variable.
1386
1387 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1388 characters since release 4.0, because Perl (from release 5.004) does not.
1389 [Or at least, is documented not to: some releases seem to be in conflict
1390 with the documentation.] However, when a pattern was studied with
1391 pcre_study() and all its branches started with \s, PCRE still included VT
1392 as a possible starting character. Of course, this did no harm; it just
1393 caused an unnecessary match attempt.
1394
1395 22. Removed a now-redundant internal flag bit that recorded the fact that case
1396 dependency changed within the pattern. This was once needed for "required
1397 byte" processing, but is no longer used. This recovers a now-scarce options
1398 bit. Also moved the least significant internal flag bit to the most-
1399 significant bit of the word, which was not previously used (hangover from
1400 the days when it was an int rather than a uint) to free up another bit for
1401 the future.
1402
1403 23. Added support for CRLF line endings as well as CR and LF. As well as the
1404 default being selectable at build time, it can now be changed at runtime
1405 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1406 specify that it is scanning data with non-default line endings.
1407
1408 24. Changed the definition of CXXLINK to make it agree with the definition of
1409 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1410
1411 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1412 recursions. This makes a big different to stack usage for some patterns.
1413
1414 26. If a subpattern containing a named recursion or subroutine reference such
1415 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1416 the space required for the compiled pattern went wrong and gave too small a
1417 value. Depending on the environment, this could lead to "Failed: internal
1418 error: code overflow at offset 49" or "glibc detected double free or
1419 corruption" errors.
1420
1421 27. Applied patches from Google (a) to support the new newline modes and (b) to
1422 advance over multibyte UTF-8 characters in GlobalReplace.
1423
1424 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1425 difference for some implementation of PCRE in some Windows version.
1426
1427 29. Added some extra testing facilities to pcretest:
1428
1429 \q<number> in a data line sets the "match limit" value
1430 \Q<number> in a data line sets the "match recursion limt" value
1431 -S <number> sets the stack size, where <number> is in megabytes
1432
1433 The -S option isn't available for Windows.
1434
1435
1436 Version 6.6 06-Feb-06
1437 ---------------------
1438
1439 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1440 in pcreposix.h. I have copied the definition from pcre.h.
1441
1442 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1443 because pcre.h is no longer a built file.
1444
1445 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1446 not normally included in the compiled code.
1447
1448
1449 Version 6.5 01-Feb-06
1450 ---------------------
1451
1452 1. When using the partial match feature with pcre_dfa_exec(), it was not
1453 anchoring the second and subsequent partial matches at the new starting
1454 point. This could lead to incorrect results. For example, with the pattern
1455 /1234/, partially matching against "123" and then "a4" gave a match.
1456
1457 2. Changes to pcregrep:
1458
1459 (a) All non-match returns from pcre_exec() were being treated as failures
1460 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1461 error message is output. Some extra information is given for the
1462 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1463 probably the only errors that are likely to be caused by users (by
1464 specifying a regex that has nested indefinite repeats, for instance).
1465 If there are more than 20 of these errors, pcregrep is abandoned.
1466
1467 (b) A binary zero was treated as data while matching, but terminated the
1468 output line if it was written out. This has been fixed: binary zeroes
1469 are now no different to any other data bytes.
1470
1471 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1472 used to set a locale for matching. The --locale=xxxx long option has
1473 been added (no short equivalent) to specify a locale explicitly on the
1474 pcregrep command, overriding the environment variables.
1475
1476 (d) When -B was used with -n, some line numbers in the output were one less
1477 than they should have been.
1478
1479 (e) Added the -o (--only-matching) option.
1480
1481 (f) If -A or -C was used with -c (count only), some lines of context were
1482 accidentally printed for the final match.
1483
1484 (g) Added the -H (--with-filename) option.
1485
1486 (h) The combination of options -rh failed to suppress file names for files
1487 that were found from directory arguments.
1488
1489 (i) Added the -D (--devices) and -d (--directories) options.
1490
1491 (j) Added the -F (--fixed-strings) option.
1492
1493 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1494
1495 (l) Added the --colo(u)r option.
1496
1497 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1498 is not present by default.
1499
1500 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1501 items such as (?R) or (?1), when the recursion could match a number of
1502 alternatives. If it matched one of the alternatives, but subsequently,
1503 outside the recursion, there was a failure, the code tried to back up into
1504 the recursion. However, because of the way PCRE is implemented, this is not
1505 possible, and the result was an incorrect result from the match.
1506
1507 In order to prevent this happening, the specification of recursion has
1508 been changed so that all such subpatterns are automatically treated as
1509 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1510
1511 4. I had overlooked the fact that, in some locales, there are characters for
1512 which isalpha() is true but neither isupper() nor islower() are true. In
1513 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1514 and ordfeminine) are like this. This affected the treatment of \w and \W
1515 when they appeared in character classes, but not when they appeared outside
1516 a character class. The bit map for "word" characters is now created
1517 separately from the results of isalnum() instead of just taking it from the
1518 upper, lower, and digit maps. (Plus the underscore character, of course.)
1519
1520 5. The above bug also affected the handling of POSIX character classes such as
1521 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1522 permanent tables. Instead, the bit maps for such a class were previously
1523 created as the appropriate unions of the upper, lower, and digit bitmaps.
1524 Now they are created by subtraction from the [[:word:]] class, which has
1525 its own bitmap.
1526
1527 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1528 It is created by subtracting the vertical space characters (\x09, \x0a,
1529 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1530 subtraction was done in the overall bitmap for a character class, meaning
1531 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1532 be recognized. This bug has been fixed.
1533
1534 7. Patches from the folks at Google:
1535
1536 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1537 real life, but is still worth protecting against".
1538
1539 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1540 regular expressions".
1541
1542 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1543 have it.
1544
1545 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1546 "configure" and the latter not, in order to fix a problem somebody had
1547 with compiling the Arg class on HP-UX.
1548
1549 (e) Improve the error-handling of the C++ wrapper a little bit.
1550
1551 (f) New tests for checking recursion limiting.
1552
1553 8. The pcre_memmove() function, which is used only if the environment does not
1554 have a standard memmove() function (and is therefore rarely compiled),
1555 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1556 returning a result (though PCRE never actually uses the result).
1557
1558 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1559 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1560 returned instead of calling malloc() with an overflowing number that would
1561 most likely cause subsequent chaos.
1562
1563 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1564
1565 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1566 with this option is matched, the nmatch and pmatch options of regexec() are
1567 ignored.
1568
1569 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1570 provided in case anyone wants to the the POSIX interface with UTF-8
1571 strings.
1572
1573 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1574 C++ linking (needed for some HP-UX environments).
1575
1576 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1577 (unused parameter) and in the pcre_printint() function (omitted "default"
1578 switch label when the default is to do nothing).
1579
1580 15. Added some code to make it possible, when PCRE is compiled as a C++
1581 library, to replace subject pointers for pcre_exec() with a smart pointer
1582 class, thus making it possible to process discontinuous strings.
1583
1584 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1585 much the same function. They were added by different people who were trying
1586 to make PCRE easy to compile on non-Unix systems. It has been suggested
1587 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1588 for compiling on Windows systems. I have therefore replaced it with
1589 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1590 defaults to "extern" for C or "extern C" for C++, which works fine on
1591 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1592 SCOPE with something explicit in config.h. In addition:
1593
1594 (a) pcreposix.h still had just "extern" instead of either of these macros;
1595 I have replaced it with PCRE_DATA_SCOPE.
1596
1597 (b) Functions such as _pcre_xclass(), which are internal to the library,
1598 but external in the C sense, all had PCRE_EXPORT in their definitions.
1599 This is apparently wrong for the Windows case, so I have removed it.
1600 (It makes no difference on Unix-like systems.)
1601
1602 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1603 of recursive calls to match(). This is different to MATCH_LIMIT because
1604 that limits the total number of calls to match(), not all of which increase
1605 the depth of recursion. Limiting the recursion depth limits the amount of
1606 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1607 when PCRE is compiled, and changed at run time. A patch from Google adds
1608 this functionality to the C++ interface.
1609
1610 18. Changes to the handling of Unicode character properties:
1611
1612 (a) Updated the table to Unicode 4.1.0.
1613
1614 (b) Recognize characters that are not in the table as "Cn" (undefined).
1615
1616 (c) I revised the way the table is implemented to a much improved format
1617 which includes recognition of ranges. It now supports the ranges that
1618 are defined in UnicodeData.txt, and it also amalgamates other
1619 characters into ranges. This has reduced the number of entries in the
1620 table from around 16,000 to around 3,000, thus reducing its size
1621 considerably. I realized I did not need to use a tree structure after
1622 all - a binary chop search is just as efficient. Having reduced the
1623 number of entries, I extended their size from 6 bytes to 8 bytes to
1624 allow for more data.
1625
1626 (d) Added support for Unicode script names via properties such as \p{Han}.
1627
1628 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1629 matching that character.
1630
1631 20. When matching a repeated Unicode property with a minimum greater than zero,
1632 (for example \pL{2,}), PCRE could look past the end of the subject if it
1633 reached it while seeking the minimum number of characters. This could
1634 happen only if some of the characters were more than one byte long, because
1635 there is a check for at least the minimum number of bytes.
1636
1637 21. Refactored the implementation of \p and \P so as to be more general, to
1638 allow for more different types of property in future. This has changed the
1639 compiled form incompatibly. Anybody with saved compiled patterns that use
1640 \p or \P will have to recompile them.
1641
1642 22. Added "Any" and "L&" to the supported property types.
1643
1644 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1645 but give a compile time error if the value is greater than 0xff.
1646
1647 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1648 accidentally not being installed or uninstalled.
1649
1650 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1651 made were to insert the current release number. This seemed silly, because
1652 it made things harder for people building PCRE on systems that don't run
1653 "configure". I have turned pcre.h into a distributed file, no longer built
1654 by "configure", with the version identification directly included. There is
1655 no longer a pcre.h.in file.
1656
1657 However, this change necessitated a change to the pcre-config script as
1658 well. It is built from pcre-config.in, and one of the substitutions was the
1659 release number. I have updated configure.ac so that ./configure now finds
1660 the release number by grepping pcre.h.
1661
1662 26. Added the ability to run the tests under valgrind.
1663
1664
1665 Version 6.4 05-Sep-05
1666 ---------------------
1667
1668 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1669 "--" to be printed when multiple files were scanned, even when none of the
1670 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1671 consider it to be a bug, and have restored the previous behaviour.
1672
1673 2. A couple of code tidies to get rid of compiler warnings.
1674
1675 3. The pcretest program used to cheat by referring to symbols in the library
1676 whose names begin with _pcre_. These are internal symbols that are not
1677 really supposed to be visible externally, and in some environments it is
1678 possible to suppress them. The cheating is now confined to including
1679 certain files from the library's source, which is a bit cleaner.
1680
1681 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1682 file's purpose clearer.
1683
1684 5. Reorganized pcre_ucp_findchar().
1685
1686
1687 Version 6.3 15-Aug-05
1688 ---------------------
1689
1690 1. The file libpcre.pc.in did not have general read permission in the tarball.
1691
1692 2. There were some problems when building without C++ support:
1693
1694 (a) If C++ support was not built, "make install" and "make test" still
1695 tried to test it.
1696
1697 (b) There were problems when the value of CXX was explicitly set. Some
1698 changes have been made to try to fix these, and ...
1699
1700 (c) --disable-cpp can now be used to explicitly disable C++ support.
1701
1702 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1703 backslash in a target when C++ was disabled. This confuses some
1704 versions of "make", apparently. Using an intermediate variable solves
1705 this. (Same for CPP_LOBJ.)
1706
1707 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1708 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1709 necessary on certain architectures.
1710
1711 4. Added a setting of -export-symbols-regex to the link command to remove
1712 those symbols that are exported in the C sense, but actually are local
1713 within the library, and not documented. Their names all begin with
1714 "_pcre_". This is not a perfect job, because (a) we have to except some
1715 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1716 available (and never for static libraries). I have made a note to try to
1717 find a way round (a) in the future.
1718
1719
1720 Version 6.2 01-Aug-05
1721 ---------------------
1722
1723 1. There was no test for integer overflow of quantifier values. A construction
1724 such as {1111111111111111} would give undefined results. What is worse, if
1725 a minimum quantifier for a parenthesized subpattern overflowed and became
1726 negative, the calculation of the memory size went wrong. This could have
1727 led to memory overwriting.
1728
1729 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1730
1731 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1732 operating environments where this matters.
1733
1734 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1735 PCRE options from within the C++ wrapper.
1736
1737 5. Named capturing subpatterns were not being correctly counted when a pattern
1738 was compiled. This caused two problems: (a) If there were more than 100
1739 such subpatterns, the calculation of the memory needed for the whole
1740 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1741 back references of the form \12, where the number was greater than 9, were
1742 not recognized as back references, even though there were sufficient
1743 previous subpatterns.
1744
1745 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1746 versions of gcc, e.g. 2.95.4.
1747
1748
1749 Version 6.1 21-Jun-05
1750 ---------------------
1751
1752 1. There was one reference to the variable "posix" in pcretest.c that was not
1753 surrounded by "#if !defined NOPOSIX".
1754
1755 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1756 the cross-check on the old pcre_info() function, for the benefit of the
1757 cut-down version of PCRE that is currently imported into Exim.
1758
1759 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1760 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1761 patterns that start (?i) but I don't think that matters. The use of (?i) is
1762 just an example; this all applies to the other options as well.
1763
1764 4. Since libtool seems to echo the compile commands it is issuing, the output
1765 from "make" can be reduced a bit by putting "@" in front of each libtool
1766 compile command.
1767
1768 5. Patch from the folks at Google for configure.in to be a bit more thorough
1769 in checking for a suitable C++ installation before trying to compile the
1770 C++ stuff. This should fix a reported problem when a compiler was present,
1771 but no suitable headers.
1772
1773 6. The man pages all had just "PCRE" as their title. I have changed them to
1774 be the relevant file name. I have also arranged that these names are
1775 retained in the file doc/pcre.txt, which is a concatenation in text format
1776 of all the man pages except the little individual ones for each function.
1777
1778 7. The NON-UNIX-USE file had not been updated for the different set of source
1779 files that come with release 6. I also added a few comments about the C++
1780 wrapper.
1781
1782
1783 Version 6.0 07-Jun-05
1784 ---------------------
1785
1786 1. Some minor internal re-organization to help with my DFA experiments.
1787
1788 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1789 didn't matter for the library itself when fully configured, but did matter
1790 when compiling without UCP support, or within Exim, where the ucp files are
1791 not imported.
1792
1793 3. Refactoring of the library code to split up the various functions into
1794 different source modules. The addition of the new DFA matching code (see
1795 below) to a single monolithic source would have made it really too
1796 unwieldy, quite apart from causing all the code to be include in a
1797 statically linked application, when only some functions are used. This is
1798 relevant even without the DFA addition now that patterns can be compiled in
1799 one application and matched in another.
1800
1801 The downside of splitting up is that there have to be some external
1802 functions and data tables that are used internally in different modules of
1803 the library but which are not part of the API. These have all had their
1804 names changed to start with "_pcre_" so that they are unlikely to clash
1805 with other external names.
1806
1807 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1808 a different (DFA) algorithm. Although it is slower than the original
1809 function, it does have some advantages for certain types of matching
1810 problem.
1811
1812 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1813 including restarting after a partial match.
1814
1815 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1816 defined when compiling for Windows was sent to me. I have put it into the
1817 code, though I have no means of testing or verifying it.
1818
1819 7. Added the pcre_refcount() auxiliary function.
1820
1821 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1822 match before or at the first newline in the subject string. In pcretest,
1823 the /f option on a pattern can be used to set this.
1824
1825 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1826 would behave wrongly. This has been present in PCRE since release 4.0.
1827
1828 10. A number of changes to the pcregrep command:
1829
1830 (a) Refactored how -x works; insert ^(...)$ instead of setting
1831 PCRE_ANCHORED and checking the length, in preparation for adding
1832 something similar for -w.
1833
1834 (b) Added the -w (match as a word) option.
1835
1836 (c) Refactored the way lines are read and buffered so as to have more
1837 than one at a time available.
1838
1839 (d) Implemented a pcregrep test script.
1840
1841 (e) Added the -M (multiline match) option. This allows patterns to match
1842 over several lines of the subject. The buffering ensures that at least
1843 8K, or the rest of the document (whichever is the shorter) is available
1844 for matching (and similarly the previous 8K for lookbehind assertions).
1845
1846 (f) Changed the --help output so that it now says
1847
1848 -w, --word-regex(p)
1849
1850 instead of two lines, one with "regex" and the other with "regexp"
1851 because that confused at least one person since the short forms are the
1852 same. (This required a bit of code, as the output is generated
1853 automatically from a table. It wasn't just a text change.)
1854
1855 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1856 option but starts with a hyphen. Could be a pattern or a path name
1857 starting with a hyphen, for instance.
1858
1859 (h) "-" can be given as a file name to represent stdin.
1860
1861 (i) When file names are being printed, "(standard input)" is used for
1862 the standard input, for compatibility with GNU grep. Previously
1863 "<stdin>" was used.
1864
1865 (j) The option --label=xxx can be used to supply a name to be used for
1866 stdin when file names are being printed. There is no short form.
1867
1868 (k) Re-factored the options decoding logic because we are going to add
1869 two more options that take data. Such options can now be given in four
1870 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1871
1872 (l) Added the -A, -B, and -C options for requesting that lines of context
1873 around matches be printed.
1874
1875 (m) Added the -L option to print the names of files that do not contain
1876 any matching lines, that is, the complement of -l.
1877
1878 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1879 continue to scan other files.
1880
1881 (o) The -s option was incorrectly implemented. For compatibility with other
1882 greps, it now suppresses the error message for a non-existent or non-
1883 accessible file (but not the return code). There is a new option called
1884 -q that suppresses the output of matching lines, which was what -s was
1885 previously doing.
1886
1887 (p) Added --include and --exclude options to specify files for inclusion
1888 and exclusion when recursing.
1889
1890 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1891 Hopefully, it now does.
1892
1893 12. Missing cast in pcre_study().
1894
1895 13. Added an "uninstall" target to the makefile.
1896
1897 14. Replaced "extern" in the function prototypes in Makefile.in with
1898 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1899 world, but is set differently for Windows.
1900
1901 15. Added a second compiling function called pcre_compile2(). The only
1902 difference is that it has an extra argument, which is a pointer to an
1903 integer error code. When there is a compile-time failure, this is set
1904 non-zero, in addition to the error test pointer being set to point to an
1905 error message. The new argument may be NULL if no error number is required
1906 (but then you may as well call pcre_compile(), which is now just a
1907 wrapper). This facility is provided because some applications need a
1908 numeric error indication, but it has also enabled me to tidy up the way
1909 compile-time errors are handled in the POSIX wrapper.
1910
1911 16. Added VPATH=.libs to the makefile; this should help when building with one
1912 prefix path and installing with another. (Or so I'm told by someone who
1913 knows more about this stuff than I do.)
1914
1915 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1916 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1917 match everything, including newlines. This is not POSIX-compatible, but
1918 somebody wanted the feature. From pcretest it can be activated by using
1919 both the P and the s flags.
1920
1921 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1922
1923 19. libpcre.pc was being incorrectly installed as executable.
1924
1925 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1926 it now also looks for '\r' so that it will work unmodified on Windows.
1927
1928 21. Added Google's contributed C++ wrapper to the distribution.
1929
1930 22. Added some untidy missing memory free() calls in pcretest, to keep
1931 Electric Fence happy when testing.
1932
1933
1934
1935 Version 5.0 13-Sep-04
1936 ---------------------
1937
1938 1. Internal change: literal characters are no longer packed up into items
1939 containing multiple characters in a single byte-string. Each character
1940 is now matched using a separate opcode. However, there may be more than one
1941 byte in the character in UTF-8 mode.
1942
1943 2. The pcre_callout_block structure has two new fields: pattern_position and
1944 next_item_length. These contain the offset in the pattern to the next match
1945 item, and its length, respectively.
1946
1947 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1948 insertion of callouts before each pattern item. Added the /C option to
1949 pcretest to make use of this.
1950
1951 4. On the advice of a Windows user, the lines
1952
1953 #if defined(_WIN32) || defined(WIN32)
1954 _setmode( _fileno( stdout ), 0x8000 );
1955 #endif /* defined(_WIN32) || defined(WIN32) */
1956
1957 have been added to the source of pcretest. This apparently does useful
1958 magic in relation to line terminators.
1959
1960 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1961 for the benefit of those environments where the "b" makes a difference.
1962
1963 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1964 to know about it. I have put a hack into configure.in that adds in code
1965 to set GCC=yes if CC=icc. This seems to end up at a point in the
1966 generated configure script that is early enough to affect the setting of
1967 compiler options, which is what is needed, but I have no means of testing
1968 whether it really works. (The user who reported this had patched the
1969 generated configure script, which of course I cannot do.)
1970
1971 LATER: After change 22 below (new libtool files), the configure script
1972 seems to know about icc (and also ecc). Therefore, I have commented out
1973 this hack in configure.in.
1974
1975 7. Added support for pkg-config (2 patches were sent in).
1976
1977 8. Negated POSIX character classes that used a combination of internal tables
1978 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1979 [[:^ascii]]. Typically, they would match almost any characters. The other
1980 POSIX classes were not broken in this way.
1981
1982 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1983 to find the match, as PCRE was deluded into thinking that the match had to
1984 start at the start point or following a newline. The same bug applied to
1985 patterns with negative forward assertions or any backward assertions
1986 preceding ".*" at the start, unless the pattern required a fixed first
1987 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1988
1989 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1990 starting at the last subject character, bytes beyond the end of the subject
1991 string were read.
1992
1993 11. Renamed the variable "class" as "classbits" to make life easier for C++
1994 users. (Previously there was a macro definition, but it apparently wasn't
1995 enough.)
1996
1997 12. Added the new field "tables" to the extra data so that tables can be passed
1998 in at exec time, or the internal tables can be re-selected. This allows
1999 a compiled regex to be saved and re-used at a later time by a different
2000 program that might have everything at different addresses.
2001
2002 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2003 -R library as well as a -L library.
2004
2005 14. The debugging options of pcretest (-d on the command line or D on a
2006 pattern) showed incorrect output for anything following an extended class
2007 that contained multibyte characters and which was followed by a quantifier.
2008
2009 15. Added optional support for general category Unicode character properties
2010 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2011 support. It adds about 90K to the size of the library. The meanings of the
2012 inbuilt class escapes such as \d and \s have NOT been changed.
2013
2014 16. Updated pcredemo.c to include calls to free() to release the memory for the
2015 compiled pattern.
2016
2017 17. The generated file chartables.c was being created in the source directory
2018 instead of in the building directory. This caused the build to fail if the
2019 source directory was different from the building directory, and was
2020 read-only.
2021
2022 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2023 file. No doubt somebody will tell me if they don't make sense... Also added
2024 Dan Mooney's comments about building on OpenVMS.
2025
2026 19. Added support for partial matching via the PCRE_PARTIAL option for
2027 pcre_exec() and the \P data escape in pcretest.
2028
2029 20. Extended pcretest with 3 new pattern features:
2030
2031 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2032 write the compiled pattern to the file whose name is "rest-of-line".
2033 This is a straight binary dump of the data, with the saved pointer to
2034 the character tables forced to be NULL. The study data, if any, is
2035 written too. After writing, pcretest reads a new pattern.
2036
2037 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2038 compiled pattern from the given file. There must not be any
2039 occurrences of "<" in the file name (pretty unlikely); if there are,
2040 pcretest will instead treat the initial "<" as a pattern delimiter.
2041 After reading in the pattern, pcretest goes on to read data lines as
2042 usual.
2043
2044 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2045 and 16-bit fields in a compiled pattern, to simulate a pattern that
2046 was compiled on a host of opposite endianness.
2047
2048 21. The pcre-exec() function can now cope with patterns that were compiled on
2049 hosts of opposite endianness, with this restriction:
2050
2051 As for any compiled expression that is saved and used later, the tables
2052 pointer field cannot be preserved; the extra_data field in the arguments
2053 to pcre_exec() should be used to pass in a tables address if a value
2054 other than the default internal tables were used at compile time.
2055
2056 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2057 now diagnosed as an error. Previously, most of the time, a negative number
2058 would have been treated as zero, but if in addition "ovector" was passed as
2059 NULL, a crash could occur.
2060
2061 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2062 new versions from the libtool 1.5 distribution (the last one is a copy of
2063 a file called libtool.m4). This seems to have fixed the need to patch
2064 "configure" to support Darwin 1.3 (which I used to do). However, I still
2065 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2066 workstation).
2067
2068 24. Changed the PCRE licence to be the more standard "BSD" licence.
2069
2070
2071 Version 4.5 01-Dec-03
2072 ---------------------
2073
2074 1. There has been some re-arrangement of the code for the match() function so
2075 that it can be compiled in a version that does not call itself recursively.
2076 Instead, it keeps those local variables that need separate instances for
2077 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2078 needs to "recurse". Keeping track of where control must go is done by means
2079 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2080 hide most of the details from the main code, and operates only if
2081 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2082 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2083 operating.
2084
2085 To make it easier for callers to provide specially tailored get/free
2086 functions for this usage, two new functions, pcre_stack_malloc, and
2087 pcre_stack_free, are used. They are always called in strict stacking order,
2088 and the size of block requested is always the same.
2089
2090 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2091 PCRE has been compiled to use the stack or the heap for recursion. The
2092 -C option of pcretest uses this to show which version is compiled.
2093
2094 A new data escape \S, is added to pcretest; it causes the amounts of store
2095 obtained and freed by both kinds of malloc/free at match time to be added
2096 to the output.
2097
2098 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2099 what's available on my current Linux desktop machine.
2100
2101 3. When matching a UTF-8 string, the test for a valid string at the start has
2102 been extended. If start_offset is not zero, PCRE now checks that it points
2103 to a byte that is the start of a UTF-8 character. If not, it returns
2104 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2105 this is necessary because there may be backward assertions in the pattern.
2106 When matching the same subject several times, it may save resources to use
2107 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2108
2109 4. The code for checking the validity of UTF-8 strings has been tightened so
2110 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2111 containing "overlong sequences".
2112
2113 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2114 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2115 should have been "&", but it just so happened that all the cases this let
2116 through by mistake were picked up later in the function.
2117
2118 6. I had used a variable called "isblank" - this is a C99 function, causing
2119 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2120
2121 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2122 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2123 so the version is shown; (c) stop "make test" echoing "./RunTest".
2124
2125 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2126
2127 9. The prototype for memmove() for systems that don't have it was using
2128 size_t, but the inclusion of the header that defines size_t was later. I've
2129 moved the #includes for the C headers earlier to avoid this.
2130
2131 10. Added some adjustments to the code to make it easier to compiler on certain
2132 special systems:
2133
2134 (a) Some "const" qualifiers were missing.
2135 (b) Added the macro EXPORT before all exported functions; by default this
2136 is defined to be empty.
2137 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2138 that it reads its output file name as an argument instead of writing
2139 to the standard output and assuming this can be redirected.
2140
2141 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2142 class containing characters with values greater than 255, PCRE compilation
2143 went into a loop.
2144
2145 12. A recursive reference to a subpattern that was within another subpattern
2146 that had a minimum quantifier of zero caused PCRE to crash. For example,
2147 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2148 recursion. If the recursively-called subpattern itself had a zero repeat,
2149 that was OK.
2150
2151 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2152 buffer into which it was copied (for escape processing) was still set at
2153 1024, so long lines caused crashes.
2154
2155 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2156 "internal error: code overflow...". This applied to any character class
2157 that was followed by a possessive quantifier.
2158
2159 15. Modified the Makefile to add libpcre.la as a prerequisite for
2160 libpcreposix.la because I was told this is needed for a parallel build to
2161 work.
2162
2163 16. If a pattern that contained .* following optional items at the start was
2164 studied, the wrong optimizing data was generated, leading to matching
2165 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2166 matching string must start with a or b or c. The correct conclusion for
2167 this pattern is that a match can start with any character.
2168
2169
2170 Version 4.4 13-Aug-03
2171 ---------------------
2172
2173 1. In UTF-8 mode, a character class containing characters with values between
2174 127 and 255 was not handled correctly if the compiled pattern was studied.
2175 In fixing this, I have also improved the studying algorithm for such
2176 classes (slightly).
2177
2178 2. Three internal functions had redundant arguments passed to them. Removal
2179 might give a very teeny performance improvement.
2180
2181 3. Documentation bug: the value of the capture_top field in a callout is *one
2182 more than* the number of the hightest numbered captured substring.
2183
2184 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2185 in incorrectly linking with a previously installed version. They now link
2186 explicitly with libpcre.la.
2187
2188 5. configure.in no longer needs to recognize Cygwin specially.
2189
2190 6. A problem in pcre.in for Windows platforms is fixed.
2191
2192 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2193 pcretest, it used to include the size of the study block as part of its
2194 output. Unfortunately, the structure contains a field that has a different
2195 size on different hardware architectures. This meant that the tests that
2196 showed this size failed. As the block is currently always of a fixed size,
2197 this information isn't actually particularly useful in pcretest output, so
2198 I have just removed it.
2199
2200 8. Three pre-processor statements accidentally did not start in column 1.
2201 Sadly, there are *still* compilers around that complain, even though
2202 standard C has not required this for well over a decade. Sigh.
2203
2204 9. In pcretest, the code for checking callouts passed small integers in the
2205 callout_data field, which is a void * field. However, some picky compilers
2206 complained about the casts involved for this on 64-bit systems. Now
2207 pcretest passes the address of the small integer instead, which should get
2208 rid of the warnings.
2209
2210 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2211 both compile and run time, and gives an error if an invalid UTF-8 sequence
2212 is found. There is a option for disabling this check in cases where the
2213 string is known to be correct and/or the maximum performance is wanted.
2214
2215 11. In response to a bug report, I changed one line in Makefile.in from
2216
2217 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2218 to
2219 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2220
2221 to look similar to other lines, but I have no way of telling whether this
2222 is the right thing to do, as I do not use Windows. No doubt I'll get told
2223 if it's wrong...
2224
2225
2226 Version 4.3 21-May-03
2227 ---------------------
2228
2229 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2230 Makefile.
2231
2232 2. Some refactoring to improve the quality of the code:
2233
2234 (i) The utf8_table... variables are now declared "const".
2235
2236 (ii) The code for \cx, which used the "case flipping" table to upper case
2237 lower case letters, now just substracts 32. This is ASCII-specific,
2238 but the whole concept of \cx is ASCII-specific, so it seems
2239 reasonable.
2240
2241 (iii) PCRE was using its character types table to recognize decimal and
2242 hexadecimal digits in the pattern. This is silly, because it handles
2243 only 0-9, a-f, and A-F, but the character types table is locale-
2244 specific, which means strange things might happen. A private
2245 table is now used for this - though it costs 256 bytes, a table is
2246 much faster than multiple explicit tests. Of course, the standard
2247 character types table is still used for matching digits in subject
2248 strings against \d.
2249
2250 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2251 ending in _t are). So I've renamed it as ESC_tee.
2252
2253 3. The first argument for regexec() in the POSIX wrapper should have been
2254 defined as "const".
2255
2256 4. Changed pcretest to use malloc() for its buffers so that they can be
2257 Electric Fenced for debugging.
2258
2259 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2260 to read one or more bytes before the start of the subject string. Often this
2261 had no effect on PCRE's behaviour, but in some circumstances it could
2262 provoke a segmentation fault.
2263
2264 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2265 to try to read one or more bytes before the start of the subject string.
2266
2267 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2268 UTF-8 support could misbehave in various ways if the subject string
2269 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2270 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2271 back over UTF-8 characters.)
2272
2273
2274 Version 4.2 14-Apr-03
2275 ---------------------
2276
2277 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2278
2279 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2280 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2281 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2282 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2283 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2284 and BUILD_EXEEXT
2285 Note: automatic setting of the BUILD variables is not yet working
2286 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2287 compile-time but not at link-time
2288 [LINK]: use for linking executables only
2289 make different versions for Windows and non-Windows
2290 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2291 libraries
2292 [LINK_FOR_BUILD]: new variable
2293 [OBJEXT]: use throughout
2294 [EXEEXT]: use throughout
2295 <winshared>: new target
2296 <wininstall>: new target
2297 <dftables.o>: use native compiler
2298 <dftables>: use native linker
2299 <install>: handle Windows platform correctly
2300 <clean>: ditto
2301 <check>: ditto
2302 copy DLL to top builddir before testing
2303
2304 As part of these changes, -no-undefined was removed again. This was reported
2305 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2306 in any case.
2307
2308 3. Some tidies to get rid of compiler warnings:
2309
2310 . In the match_data structure, match_limit was an unsigned long int, whereas
2311 match_call_count was an int. I've made them both unsigned long ints.
2312
2313 . In pcretest the fact that a const uschar * doesn't automatically cast to
2314 a void * provoked a warning.
2315
2316 . Turning on some more compiler warnings threw up some "shadow" variables
2317 and a few more missing casts.
2318
2319 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2320 option, a class that contained a single character with a value between 128
2321 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2322
2323 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2324 option, a class that contained several characters, but with at least one
2325 whose value was between 128 and 255 caused PCRE to crash.
2326
2327
2328 Version 4.1 12-Mar-03
2329 ---------------------
2330
2331 1. Compiling with gcc -pedantic found a couple of places where casts were
2332 needed, and a string in dftables.c that was longer than standard compilers are
2333 required to support.
2334
2335 2. Compiling with Sun's compiler found a few more places where the code could
2336 be tidied up in order to avoid warnings.
2337
2338 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2339 first of these names is deprecated in the latest Autoconf in favour of the name
2340 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2341 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2342 analogy I have changed it to CFLAGS_FOR_BUILD.
2343
2344 4. Added -no-undefined to the linking command in the Makefile, because this is
2345 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2346 linking step for the pcreposix library.
2347
2348 5. PCRE was failing to diagnose the case of two named groups with the same
2349 name.
2350
2351 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2352 literal character that is needed in the subject for a match, and scans along to
2353 ensure that it is present before embarking on the full matching process. This
2354 saves time in cases of nested unlimited repeats that are never going to match.
2355 Problem: the scan can take a lot of time if the subject is very long (e.g.
2356 megabytes), thus penalizing straightforward matches. It is now done only if the
2357 amount of subject to be scanned is less than 1000 bytes.
2358
2359 7. A lesser problem with the same optimization is that it was recording the
2360 first character of an anchored pattern as "needed", thus provoking a search
2361 right along the subject, even when the first match of the pattern was going to
2362 fail. The "needed" character is now not set for anchored patterns, unless it
2363 follows something in the pattern that is of non-fixed length. Thus, it still
2364 fulfils its original purpose of finding quick non-matches in cases of nested
2365 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2366
2367
2368 Version 4.0 17-Feb-03
2369 ---------------------
2370
2371 1. If a comment in an extended regex that started immediately after a meta-item
2372 extended to the end of string, PCRE compiled incorrect data. This could lead to
2373 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2374
2375 2. Moved to autoconf 2.53 and libtool 1.4.2.
2376
2377 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2378 the special perltest8 script is no longer needed - all the tests can be run
2379 from a single perltest script.
2380
2381 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2382 by \s. It has now been removed in PCRE. This means it isn't recognized as
2383 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2384 class [:space:] *does* include VT, thereby creating a mess.
2385
2386 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2387 space and tab.
2388
2389 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2390 its new features into the main test script, reducing the number of scripts.
2391
2392 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2393 were backward compatible, and made the (?i) apply to the whole pattern, as if
2394 /i were given. Now it behaves more logically, and applies the option setting
2395 only to what follows. PCRE has been changed to follow suit. However, if it
2396 finds options settings right at the start of the pattern, it extracts them into
2397 the global options, as before. Thus, they show up in the info data.
2398
2399 8. Added support for the \Q...\E escape sequence. Characters in between are
2400 treated as literals. This is slightly different from Perl in that $ and @ are
2401 also handled as literals inside the quotes. In Perl, they will cause variable
2402 interpolation. Note the following examples:
2403
2404 Pattern PCRE matches Perl matches
2405
2406 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2407 \Qabc\$xyz\E abc\$xyz abc\$xyz
2408 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2409
2410 For compatibility with Perl, \Q...\E sequences are recognized inside character
2411 classes as well as outside them.
2412
2413 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2414 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2415 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2416 signed/unsigned warnings.
2417
2418 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2419 option for pcretest, so I've replaced it by a simple function that does just
2420 that job.
2421
2422 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2423 "pcregrep -".
2424
2425 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2426 Java package. This provides some syntactic sugar for simple cases of what my
2427 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2428 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2429 item, you can use this simplified notation. Note that only makes sense with
2430 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2431 greediness, whatever the setting of the PCRE_UNGREEDY option.
2432
2433 13. A change of greediness default within a pattern was not taking effect at
2434 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2435 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2436 was abstracted outside.
2437
2438 14. PCRE now supports the \G assertion. It is true when the current matching
2439 position is at the start point of the match. This differs from \A when the
2440 starting offset is non-zero. Used with the /g option of pcretest (or similar
2441 code), it works in the same way as it does for Perl's /g option. If all
2442 alternatives of a regex begin with \G, the expression is anchored to the start
2443 match position, and the "anchored" flag is set in the compiled expression.
2444
2445 15. Some bugs concerning the handling of certain option changes within patterns
2446 have been fixed. These applied to options other than (?ims). For example,
2447 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2448 been the other way round. Some of this was related to change 7 above.
2449
2450 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2451 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2452 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2453 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2454
2455 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2456 mode. Unlike ".", it always matches newline, whatever the setting of
2457 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2458 assertions. Perl allows it, but it doesn't (in general) work because it can't
2459 calculate the length of the lookbehind. At least, that's the case for Perl
2460 5.8.0 - I've been told they are going to document that it doesn't work in
2461 future.
2462
2463 18. Added an error diagnosis for escapes that PCRE does not support: these are
2464 \L, \l, \N, \P, \p, \U, \u, and \X.
2465
2466 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2467 reading past the end of the pattern in cases such as /[abcd/.
2468
2469 20. PCRE was getting more memory than necessary for patterns with classes that
2470 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2471
2472 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2473 compiling PCRE for use with Virtual Pascal.
2474
2475 22. Small fix to the Makefile to make it work properly if the build is done
2476 outside the source tree.
2477
2478 23. Added a new extension: a condition to go with recursion. If a conditional
2479 subpattern starts with (?(R) the "true" branch is used if recursion has
2480 happened, whereas the "false" branch is used only at the top level.
2481
2482 24. When there was a very long string of literal characters (over 255 bytes
2483 without UTF support, over 250 bytes with UTF support), the computation of how
2484 much memory was required could be incorrect, leading to segfaults or other
2485 strange effects.
2486
2487 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2488 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2489 there was a subsequent back reference to those brackets. This meant that, for
2490 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2491 possible to check for precisely this case. All we can do is abandon the
2492 optimization if .* occurs inside capturing brackets when there are any back
2493 references whatsoever. (See below for a better fix that came later.)
2494
2495 26. The handling of the optimization for finding the first character of a
2496 non-anchored pattern, and for finding a character that is required later in the
2497 match were failing in some cases. This didn't break the matching; it just
2498 failed to optimize when it could. The way this is done has been re-implemented.
2499
2500 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2501
2502 28. Added a new feature that provides some of the functionality that Perl
2503 provides with (?{...}). The facility is termed a "callout". The way it is done
2504 in PCRE is for the caller to provide an optional function, by setting
2505 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2506 global variable. By default it is unset, which disables all calling out. To get
2507 the function called, the regex must include (?C) at appropriate points. This
2508 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2509 This provides a means of identifying different callout points. When PCRE
2510 reaches such a point in the regex, if pcre_callout has been set, the external
2511 function is called. It is provided with data in a structure called
2512 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2513 matching continues; if it returns a non-zero value, the match at the current
2514 point fails. However, backtracking will occur if possible. [This was changed
2515 later and other features added - see item 49 below.]
2516
2517 29. pcretest is upgraded to test the callout functionality. It provides a
2518 callout function that displays information. By default, it shows the start of
2519 the match and the current position in the text. There are some new data escapes
2520 to vary what happens:
2521
2522 \C+ in addition, show current contents of captured substrings
2523 \C- do not supply a callout function
2524 \C!n return 1 when callout number n is reached
2525 \C!n!m return 1 when callout number n is reached for the mth time
2526
2527 30. If pcregrep was called with the -l option and just a single file name, it
2528 output "<stdin>" if a match was found, instead of the file name.
2529
2530 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2531 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2532 pcre_exec(). This saves a malloc/free per call. The default value of
2533 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2534 when configuring.
2535
2536 32. The default maximum size of a compiled pattern is 64K. There have been a
2537 few cases of people hitting this limit. The code now uses macros to handle the
2538 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2539 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2540 configuring. Tests 2 and 5 work only with 2-byte links because they output
2541 debugging information about compiled patterns.
2542
2543 33. Internal code re-arrangements:
2544
2545 (a) Moved the debugging function for printing out a compiled regex into
2546 its own source file (printint.c) and used #include to pull it into
2547 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2548 separate copies.
2549
2550 (b) Defined the list of op-code names for debugging as a macro in
2551 internal.h so that it is next to the definition of the opcodes.
2552
2553 (c) Defined a table of op-code lengths for simpler skipping along compiled
2554 code. This is again a macro in internal.h so that it is next to the
2555 definition of the opcodes.
2556
2557 34. Added support for recursive calls to individual subpatterns, along the
2558 lines of Robin Houston's patch (but implemented somewhat differently).
2559
2560 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2561 allow it to read and process whole directories in Win32. This code was
2562 contributed by Lionel Fourquaux; it has not been tested by me.
2563
2564 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2565 used to name a group. Names consist of alphanumerics and underscores, and must
2566 be unique. Back references use the syntax (?P=name) and recursive calls use
2567 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2568 numbers. The function pcre_fullinfo() can be used after compilation to extract
2569 a name/number map. There are three relevant calls:
2570
2571 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2572 PCRE_INFO_NAMECOUNT yields the number of entries
2573 PCRE_INFO_NAMETABLE yields a pointer to the map.
2574
2575 The map is a vector of fixed-size entries. The size of each entry depends on
2576 the length of the longest name used. The first two bytes of each entry are the
2577 group number, most significant byte first. There follows the corresponding
2578 name, zero terminated. The names are in alphabetical order.
2579
2580 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2581 case instead of 255. Making it the same both with and without UTF-8 support
2582 means that the same test output works with both.
2583
2584 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2585 calling malloc() with a zero argument.
2586
2587 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2588 optimization. I've improved things by keeping a bitmap of backreferences with
2589 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2590 fact referenced, the optimization can be applied. It is unlikely that a
2591 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2592 the match to follow \n) will appear inside brackets with a number greater than
2593 31, but if it does, any back reference > 31 suppresses the optimization.
2594
2595 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2596 of disabling numbered capturing parentheses. Any opening parenthesis that is
2597 not followed by ? behaves as if it were followed by ?: but named parentheses
2598 can still be used for capturing (and they will acquire numbers in the usual
2599 way).
2600
2601 41. Redesigned the return codes from the match() function into yes/no/error so
2602 that errors can be passed back from deep inside the nested calls. A malloc
2603 failure while inside a recursive subpattern call now causes the
2604 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2605
2606 42. It is now possible to set a limit on the number of times the match()
2607 function is called in a call to pcre_exec(). This facility makes it possible to
2608 limit the amount of recursion and backtracking, though not in a directly
2609 obvious way, because the match() function is used in a number of different
2610 circumstances. The count starts from zero for each position in the subject
2611 string (for non-anchored patterns). The default limit is, for compatibility, a
2612 large number, namely 10 000 000. You can change this in two ways:
2613
2614 (a) When configuring PCRE before making, you can use --with-match-limit=n
2615 to set a default value for the compiled library.
2616
2617 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2618 a different value is set. See 45 below.
2619
2620 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2621
2622 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2623 of things that can be changed at compile time. The first argument specifies
2624 what is wanted and the second points to where the information is to be placed.
2625 The current list of available information is:
2626
2627 PCRE_CONFIG_UTF8
2628
2629 The output is an integer that is set to one if UTF-8 support is available;
2630 otherwise it is set to zero.
2631
2632 PCRE_CONFIG_NEWLINE
2633
2634 The output is an integer that it set to the value of the code that is used for
2635 newline. It is either LF (10) or CR (13).
2636
2637 PCRE_CONFIG_LINK_SIZE
2638
2639 The output is an integer that contains the number of bytes used for internal
2640 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2641
2642 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2643
2644 The output is an integer that contains the threshold above which the POSIX
2645 interface uses malloc() for output vectors. See item 31 above.
2646
2647 PCRE_CONFIG_MATCH_LIMIT
2648
2649 The output is an unsigned integer that contains the default limit of the number
2650 of match() calls in a pcre_exec() execution. See 42 above.
2651
2652 44. pcretest has been upgraded by the addition of the -C option. This causes it
2653 to extract all the available output from the new pcre_config() function, and to
2654 output it. The program then exits immediately.
2655
2656 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2657 order to support additional features. One way would have been to define
2658 pcre_exec2() (for example) with extra arguments, but this would not have been
2659 extensible, and would also have required all calls to the original function to
2660 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2661 is used for passing in "extra" data from pcre_study().
2662
2663 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2664 contains the following fields:
2665
2666 flags a bitmap indicating which of the following fields are set
2667 study_data opaque data from pcre_study()
2668 match_limit a way of specifying a limit on match() calls for a specific
2669 call to pcre_exec()
2670 callout_data data for callouts (see 49 below)
2671
2672 The flag bits are also defined in pcre.h, and are
2673
2674 PCRE_EXTRA_STUDY_DATA
2675 PCRE_EXTRA_MATCH_LIMIT
2676 PCRE_EXTRA_CALLOUT_DATA
2677
2678 The pcre_study() function now returns one of these new pcre_extra blocks, with
2679 the actual study data pointed to by the study_data field, and the
2680 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2681 before. That is, this change is entirely upwards-compatible and requires no
2682 change to existing code.
2683
2684 If you want to pass in additional data to pcre_exec(), you can either place it
2685 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2686 block.
2687
2688 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2689 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2690 times with different match limits, until it finds the minimum value needed for
2691 pcre_exec() to complete. The value is then output. This can be instructive; for
2692 most simple matches the number is quite small, but for pathological cases it
2693 gets very large very quickly.
2694
2695 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2696 returns the size of the data block pointed to by the study_data field in a
2697 pcre_extra block, that is, the value that was passed as the argument to
2698 pcre_malloc() when PCRE was getting memory in which to place the information
2699 created by pcre_study(). The fourth argument should point to a size_t variable.
2700 pcretest has been extended so that this information is shown after a successful
2701 pcre_study() call when information about the compiled regex is being displayed.
2702
2703 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2704 because what follows is always an absolute path. (Later: it turns out that this
2705 is more than cosmetic for MinGW, because it doesn't like empty path
2706 components.)
2707
2708 49. Some changes have been made to the callout feature (see 28 above):
2709
2710 (i) A callout function now has three choices for what it returns:
2711
2712 0 => success, carry on matching
2713 > 0 => failure at this point, but backtrack if possible
2714 < 0 => serious error, return this value from pcre_exec()
2715
2716 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2717 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2718 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2719 use by callout functions. It will never be used by PCRE itself.
2720
2721 (ii) The pcre_extra structure (see 45 above) has a void * field called
2722 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2723 pcre_callout_block structure has a field of the same name. The contents of
2724 the field passed in the pcre_extra structure are passed to the callout
2725 function in the corresponding field in the callout block. This makes it
2726 easier to use the same callout-containing regex from multiple threads. For
2727 testing, the pcretest program has a new data escape
2728
2729 \C*n pass the number n (may be negative) as callout_data
2730
2731 If the callout function in pcretest receives a non-zero value as
2732 callout_data, it returns that value.
2733
2734 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2735 there were some redundant $(CFLAGS) in commands that are now specified as
2736 $(LINK), which already includes $(CFLAGS).
2737
2738 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2739 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2740 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2741 one-byte characters throughout. Note that case-insensitive matching applies
2742 only to characters whose values are less than 256. PCRE doesn't support the
2743 notion of cases for higher-valued characters.
2744
2745 (i) A character class whose characters are all within 0-255 is handled as
2746 a bit map, and the map is inverted for negative classes. Previously, a
2747 character > 255 always failed to match such a class; however it should
2748 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2749
2750 (ii) A negated character class with a single character < 255 is coded as
2751 "not this character" (OP_NOT). This wasn't working properly when the test
2752 character was multibyte, either singly or repeated.
2753
2754 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2755 mode, for example: \x{100}{2,3}.
2756
2757 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2758 singly or repeated) now correctly test multibyte characters. However,
2759 PCRE doesn't recognize any characters with values greater than 255 as
2760 digits, spaces, or word characters. Such characters always match \D, \S,
2761 and \W, and never match \d, \s, or \w.
2762
2763 (v) Classes may now contain characters and character ranges with values
2764 greater than 255. For example: [ab\x{100}-\x{400}].
2765
2766 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2767 PCRE in UTF-8 mode.
2768
2769 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2770 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2771 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2772 value.)
2773
2774 53. The single man page has become too large. I have therefore split it up into
2775 a number of separate man pages. These also give rise to individual HTML pages;
2776 these are now put in a separate directory, and there is an index.html page that
2777 lists them all. Some hyperlinking between the pages has been installed.
2778
2779 54. Added convenience functions for handling named capturing parentheses.
2780
2781 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2782 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2783 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2784 are faulted.
2785
2786 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2787 calling configure. These values are used when compiling the dftables.c program
2788 which is run to generate the source of the default character tables. They
2789 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2790 you will need to set these values.
2791
2792 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2793
2794
2795 Version 3.9 02-Jan-02
2796 ---------------------
2797
2798 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2799
2800 2. If --disable-static was given, the building process failed when trying to
2801 build pcretest and pcregrep. (For some reason it was using libtool to compile
2802 them, which is not right, as they aren't part of the library.)
2803
2804
2805 Version 3.8 18-Dec-01
2806 ---------------------
2807
2808 1. The experimental UTF-8 code was completely screwed up. It was packing the
2809 bytes in the wrong order. How dumb can you get?
2810
2811
2812 Version 3.7 29-Oct-01
2813 ---------------------
2814
2815 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2816 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2817 this didn't happen under Solaris 8, where I normally test things.
2818
2819 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2820 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2821 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2822 this fix an hour or so after the initial 3.7 release.)
2823
2824
2825 Version 3.6 23-Oct-01
2826 ---------------------
2827
2828 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2829 offsets passed as NULL with zero offset count.
2830
2831 2. The config.guess and config.sub files had not been updated when I moved to
2832 the latest autoconf.
2833
2834
2835 Version 3.5 15-Aug-01
2836 ---------------------
2837
2838 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2839 had been forgotten.
2840
2841 2. By using declared but undefined structures, we can avoid using "void"
2842 definitions in pcre.h while keeping the internal definitions of the structures
2843 private.
2844
2845 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2846 user point of view, this means that both static and shared libraries are built
2847 by default, but this can be individually controlled. More of the work of
2848 handling this static/shared cases is now inside libtool instead of PCRE's make
2849 file.
2850
2851 4. The pcretest utility is now installed along with pcregrep because it is
2852 useful for users (to test regexs) and by doing this, it automatically gets
2853 relinked by libtool. The documentation has been turned into a man page, so
2854 there are now .1, .txt, and .html versions in /doc.
2855
2856 5. Upgrades to pcregrep:
2857 (i) Added long-form option names like gnu grep.
2858 (ii) Added --help to list all options with an explanatory phrase.
2859 (iii) Added -r, --recursive to recurse into sub-directories.
2860 (iv) Added -f, --file to read patterns from a file.
2861
2862 6. pcre_exec() was referring to its "code" argument before testing that
2863 argument for NULL (and giving an error if it was NULL).
2864
2865 7. Upgraded Makefile.in to allow for compiling in a different directory from
2866 the source directory.
2867
2868 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2869 options bits, the pointer it was passed was to an int instead of to an unsigned
2870 long int. This mattered only on 64-bit systems.
2871
2872 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2873 generated) instead of pcre.in, which it its source. Also made the same change
2874 in several of the .c files.
2875
2876 10. A new release of gcc defines printf() as a macro, which broke pcretest
2877 because it had an ifdef in the middle of a string argument for printf(). Fixed
2878 by using separate calls to printf().
2879
2880 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2881 script, to force use of CR or LF instead of \n in the source. On non-Unix
2882 systems, the value can be set in config.h.
2883
2884 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2885 absolute limit. Changed the text of the error message to make this clear, and
2886 likewise updated the man page.
2887
2888 13. The limit of 99 on the number of capturing subpatterns has been removed.
2889 The new limit is 65535, which I hope will not be a "real" limit.
2890
2891
2892 Version 3.4 22-Aug-00
2893 ---------------------
2894
2895 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2896
2897 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2898
2899
2900 Version 3.3 01-Aug-00
2901 ---------------------
2902
2903 1. If an octal character was given, but the value was greater than \377, it
2904 was not getting masked to the least significant bits, as documented. This could
2905 lead to crashes in some systems.
2906
2907 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2908 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2909
2910 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2911 These just pass their arguments on to (pcre_free)(), but they are provided
2912 because some uses of PCRE bind it to non-C systems that can call its functions,
2913 but cannot call free() or pcre_free() directly.
2914
2915 4. Add "make test" as a synonym for "make check". Corrected some comments in
2916 the Makefile.
2917
2918 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2919 Makefile.
2920
2921 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2922 command called pgrep for grepping around the active processes.
2923
2924 7. Added the beginnings of support for UTF-8 character strings.
2925
2926 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2927 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2928 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2929 out for the ar command.)
2930
2931
2932 Version 3.2 12-May-00
2933 ---------------------
2934
2935 This is purely a bug fixing release.
2936
2937 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2938 of ZA. This was just one example of several cases that could provoke this bug,
2939 which was introduced by change 9 of version 2.00. The code for breaking
2940 infinite loops after an iteration that matches an empty string was't working
2941 correctly.
2942
2943 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2944 when matched against abbab (for example). After matching an empty string, it
2945 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2946 caused it to match further down the string than it should.
2947
2948 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2949 was there because it doesn't seem to be needed, and it causes trouble on some
2950 systems, as it is not a Standard C header. It has been removed.
2951
2952 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2953 were reported on the Macintosh. The changes were from
2954
2955 while ((c = *(++ptr)) != 0 && c != '\n');
2956 to
2957 while ((c = *(++ptr)) != 0 && c != '\n') ;
2958
2959 Totally extraordinary, but if that's what it takes...
2960
2961 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2962 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2963 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2964 assumes the way PCRE uses memmove() (always moving upwards).
2965
2966 6. PCRE is being used in one environment where strchr() is not available. There
2967 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2968 faster code anyway.
2969
2970
2971 Version 3.1 09-Feb-00
2972 ---------------------
2973
2974 The only change in this release is the fixing of some bugs in Makefile.in for
2975 the "install" target:
2976
2977 (1) It was failing to install pcreposix.h.
2978
2979 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2980
2981
2982 Version 3.0 01-Feb-00
2983 ---------------------
2984
2985 1. Add support for the /+ modifier to perltest (to output $` like it does in
2986 pcretest).
2987
2988 2. Add support for the /g modifier to perltest.
2989
2990 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2991 matches null strings.
2992
2993 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2994 pattern. Perl treats empty patterns specially - it reuses the most recent
2995 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2996 effect.
2997
2998 5. The POSIX interface was broken in that it was just handing over the POSIX
2999 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3000 required a bigger vector, with some working space on the end. This means that
3001 the POSIX wrapper now has to get and free some memory, and copy the results.
3002
3003 6. Added some simple autoconf support, placing the test data and the
3004 documentation in separate directories, re-organizing some of the
3005 information files, and making it build pcre-config (a GNU standard). Also added
3006 libtool support for building PCRE as a shared library, which is now the
3007 default.
3008
3009 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3010 09 are not valid octal constants. Single digits will be used for minor values
3011 less than 10.
3012
3013 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3014 existing programs that set these in the POSIX interface can use PCRE without
3015 modification.
3016
3017 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3018 return all that pcre_info() returns, plus additional data. The pcre_info()
3019 function is retained for compatibility, but is considered to be obsolete.
3020
3021 10. Added experimental recursion feature (?R) to handle one common case that
3022 Perl 5.6 will be able to do with (?p{...}).
3023
3024 11. Added support for POSIX character classes like [:alpha:], which Perl is
3025 adopting.
3026
3027
3028 Version 2.08 31-Aug-99
3029 ----------------------
3030
3031 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3032 trying to match at the startoffset position, but instead was moving forward to
3033 the next newline as if a previous match had failed.
3034
3035 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3036 and could get into a loop if a null string was matched other than at the start
3037 of the subject.
3038
3039 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3040 be distinguished at compile time, and for completeness also added PCRE_DATE.
3041
3042 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3043 in GnuWin32 environments.
3044
3045
3046 Version 2.07 29-Jul-99
3047 ----------------------
3048
3049 1. The documentation is now supplied in plain text form and HTML as well as in
3050 the form of man page sources.
3051
3052 2. C++ compilers don't like assigning (void *) values to other pointer types.
3053 In particular this affects malloc(). Although there is no problem in Standard
3054 C, I've put in casts to keep C++ compilers happy.
3055
3056 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3057 should be (const char *).
3058
3059 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3060 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3061 However, I haven't made this a standard facility. The documentation doesn't
3062 mention it, and the Makefile doesn't support it.
3063
3064 5. The Makefile now contains an "install" target, with editable destinations at
3065 the top of the file. The pcretest program is not installed.
3066
3067 6. pgrep -V now gives the PCRE version number and date.
3068
3069 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3070 causing the entire string to be ignored, instead of just the last character.
3071
3072 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3073 non-matching string, it can take a very, very long time, even for strings of
3074 quite modest length, because of the nested recursion. PCRE now does better in
3075 some of these cases. It does this by remembering the last required literal
3076 character in the pattern, and pre-searching the subject to ensure it is present
3077 before running the real match. In other words, it applies a heuristic to detect
3078 some types of certain failure quickly, and in the above example, if presented
3079 with a string that has no trailing " it gives "no match" very quickly.
3080
3081 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3082 other alternatives are tried instead.
3083
3084
3085 Version 2.06 09-Jun-99
3086 ----------------------
3087
3088 1. Change pcretest's output for amount of store used to show just the code
3089 space, because the remainder (the data block) varies in size between 32-bit and
3090 64-bit systems.
3091
3092 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3093 start matching at. This allows lookbehinds to work when searching for multiple
3094 occurrences in a string.
3095
3096 3. Added additional options to pcretest for testing multiple occurrences:
3097
3098 /+ outputs the rest of the string that follows a match
3099 /g loops for multiple occurrences, using the new startoffset argument
3100 /G loops for multiple occurrences by passing an incremented pointer
3101
3102 4. PCRE wasn't doing the "first character" optimization for patterns starting
3103 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3104 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3105 the letter 'x'. On long subject strings, this gives a significant speed-up.
3106
3107
3108 Version 2.05 21-Apr-99
3109 ----------------------
3110
3111 1. Changed the type of magic_number from int to long int so that it works
3112 properly on 16-bit systems.
3113
3114 2. Fixed a bug which caused patterns starting with .* not to work correctly
3115 when the subject string contained newline characters. PCRE was assuming
3116 anchoring for such patterns in all cases, which is not correct because .* will
3117 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3118 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3119 must be retried after every newline in the subject.
3120
3121
3122 Version 2.04 18-Feb-99
3123 ----------------------
3124
3125 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3126 computation of the store needed to hold the pattern was incorrect (too large).
3127 If such patterns were nested a few deep, this could multiply and become a real
3128 problem.
3129
3130 2. Added /M option to pcretest to show the memory requirement of a specific
3131 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3132
3133 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3134 compiled in such a way that the backtracking after subsequent failure was
3135 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3136 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3137
3138
3139 Version 2.03 02-Feb-99
3140 ----------------------
3141
3142 1. Fixed typo and small mistake in man page.
3143
3144 2. Added 4th condition (GPL supersedes if conflict) and created separate
3145 LICENCE file containing the conditions.
3146
3147 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3148 Perl, that is the internal \ allows the delimiter to be included in the
3149 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3150 the final delimiter, add \ to the end of the pattern (to test the error).
3151
3152 4. Added the convenience functions for extracting substrings after a successful
3153 match. Updated pcretest to make it able to test these functions.
3154
3155
3156 Version 2.02 14-Jan-99
3157 ----------------------
3158
3159 1. Initialized the working variables associated with each extraction so that
3160 their saving and restoring doesn't refer to uninitialized store.
3161
3162 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3163 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3164 fix the problem.
3165
3166 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3167 calls, and wasn't printing the correct value for compiling calls. Increased the
3168 default value of LOOPREPEAT, and the number of significant figures in the
3169 times.
3170
3171 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3172
3173 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3174 a building problem on Windows NT with a FAT file system.
3175
3176
3177 Version 2.01 21-Oct-98
3178 ----------------------
3179
3180 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3181 to character tables built by pcre_maketables() in the current locale. If NULL
3182 is passed, the default tables are used.
3183
3184
3185 Version 2.00 24-Sep-98
3186 ----------------------
3187
3188 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3189 it any more.
3190
3191 2. Allow quantification of (?>) groups, and make it work correctly.
3192
3193 3. The first character computation wasn't working for (?>) groups.
3194
3195 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3196 end of the subject) and add 5.005's \z, which really does match only at the
3197 very end of the subject.
3198
3199 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3200
3201 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3202 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3203 localized options. All options to pcre_study() were also removed.
3204
3205 7. Add other new features from 5.005:
3206
3207 $(?<= positive lookbehind
3208 $(?<! negative lookbehind
3209 (?imsx-imsx) added the unsetting capability
3210 such a setting is global if at outer level; local otherwise
3211 (?imsx-imsx:) non-capturing groups with option setting
3212 (?(cond)re|re) conditional pattern matching
3213
3214 A backreference to itself in a repeated group matches the previous
3215 captured string.
3216
3217 8. General tidying up of studying (both automatic and via "study")
3218 consequential on the addition of new assertions.
3219
3220 9. As in 5.005, unlimited repeated groups that could match an empty substring
3221 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3222 runtime if any iteration does actually match an empty substring.
3223
3224 10. Include the RunTest script in the distribution.
3225
3226 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3227 discrepancies, some of which were old and were also with respect to 5.004. They
3228 have now been fixed.
3229
3230
3231 Version 1.09 28-Apr-98
3232 ----------------------
3233
3234 1. A negated single character class followed by a quantifier with a minimum
3235 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3236 program crashes, or just wrong answers. This did not apply to negated classes
3237 containing more than one character, or to minima other than one.
3238
3239
3240 Version 1.08 27-Mar-98
3241 ----------------------
3242
3243 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3244
3245 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3246 latter must appear before anything that relies on it in the pattern.
3247
3248
3249 Version 1.07 16-Feb-98
3250 ----------------------
3251
3252 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3253 repeat of a potentially empty string).
3254
3255
3256 Version 1.06 23-Jan-98
3257 ----------------------
3258
3259 1. Added Markus Oberhumer's little patches for C++.
3260
3261 2. Literal strings longer than 255 characters were broken.
3262
3263
3264 Version 1.05 23-Dec-97
3265 ----------------------
3266
3267 1. Negated character classes containing more than one character were failing if
3268 PCRE_CASELESS was set at run time.
3269
3270
3271 Version 1.04 19-Dec-97
3272 ----------------------
3273
3274 1. Corrected the man page, where some "const" qualifiers had been omitted.
3275
3276 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3277 input syntax.
3278
3279 3. Fixed memory leak which occurred when a regex with back references was
3280 matched with an offsets vector that wasn't big enough. The temporary memory
3281 that is used in this case wasn't being freed if the match failed.
3282
3283 4. Tidied pcretest to ensure it frees memory that it gets.
3284
3285 5. Temporary memory was being obtained in the case where the passed offsets
3286 vector was exactly big enough.
3287
3288 6. Corrected definition of offsetof() from change 5 below.
3289
3290 7. I had screwed up change 6 below and broken the rules for the use of
3291 setjmp(). Now fixed.
3292
3293
3294 Version 1.03 18-Dec-97
3295 ----------------------
3296
3297 1. A erroneous regex with a missing opening parenthesis was correctly
3298 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3299 on some systems.
3300
3301 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3302 it was reported that one broken compiler failed on the former because "code" is
3303 also an independent variable.
3304
3305 3. The erroneous regex a[]b caused an array overrun reference.
3306
3307 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3308 fail on data ending with that character. (It was going on too far, and checking
3309 the next character, typically a binary zero.) This was specific to the
3310 optimized code for single-character negative classes.
3311
3312 5. Added a contributed patch from the TIN world which does the following:
3313
3314 + Add an undef for memmove, in case the the system defines a macro for it.
3315
3316 + Add a definition of offsetof(), in case there isn't one. (I don't know
3317 the reason behind this - offsetof() is part of the ANSI standard - but
3318 it does no harm).
3319
3320 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3321 most of the places where whitespace preceded '#'. I have given up and
3322 allowed the remaining 2 cases to be at the margin.
3323
3324 + Rename some variables in pcre to eliminate shadowing. This seems very
3325 pedantic, but does no harm, of course.
3326
3327 6. Moved the call to setjmp() into its own function, to get rid of warnings
3328 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3329
3330 7. Constructs such as \d{8,} were compiling into the equivalent of
3331 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3332 outcome, but in this particular case used more store than had been allocated,
3333 which caused the bug to be discovered because it threw up an internal error.
3334
3335 8. The debugging code in both pcre and pcretest for outputting the compiled
3336 form of a regex was going wrong in the case of back references followed by
3337 curly-bracketed repeats.
3338
3339
3340 Version 1.02 12-Dec-97
3341 ----------------------
3342
3343 1. Typos in pcre.3 and comments in the source fixed.
3344
3345 2. Applied a contributed patch to get rid of places where it used to remove
3346 'const' from variables, and fixed some signed/unsigned and uninitialized
3347 variable warnings.
3348
3349 3. Added the "runtest" target to Makefile.
3350
3351 4. Set default compiler flag to -O2 rather than just -O.
3352
3353
3354 Version 1.01 19-Nov-97
3355 ----------------------
3356
3357 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3358 like /([ab]*)*/, that is, for classes with more than one character in them.
3359
3360 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3361 as /((?>a*))*/ (a PCRE_EXTRA facility).
3362
3363
3364 Version 1.00 18-Nov-97
3365 ----------------------
3366
3367 1. Added compile-time macros to support systems such as SunOS4 which don't have
3368 memmove() or strerror() but have other things that can be used instead.
3369
3370 2. Arranged that "make clean" removes the executables.
3371
3372
3373 Version 0.99 27-Oct-97
3374 ----------------------
3375
3376 1. Fixed bug in code for optimizing classes with only one character. It was
3377 initializing a 32-byte map regardless, which could cause it to run off the end
3378 of the memory it had got.
3379
3380 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3381
3382
3383 Version 0.98 22-Oct-97
3384 ----------------------
3385
3386 1. Fixed bug in code for handling temporary memory usage when there are more
3387 back references than supplied space in the ovector. This could cause segfaults.
3388
3389
3390 Version 0.97 21-Oct-97
3391 ----------------------
3392
3393 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3394
3395 2. Optimized negated single characters not to use a bit map.
3396
3397 3. Brought error texts together as macro definitions; clarified some of them;
3398 fixed one that was wrong - it said "range out of order" when it meant "invalid
3399 escape sequence".
3400
3401 4. Changed some char * arguments to const char *.
3402
3403 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3404
3405 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3406 pcretest.
3407
3408
3409 Version 0.96 16-Oct-97
3410 ----------------------
3411
3412 1. Added a simple "pgrep" utility to the distribution.
3413
3414 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3415 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3416 where "ddd" means "one or more decimal digits".
3417
3418 3. Fixed serious bug. If a pattern had a back reference, but the call to
3419 pcre_exec() didn't supply a large enough ovector to record the related
3420 identifying subpattern, the match always failed. PCRE now remembers the number
3421 of the largest back reference, and gets some temporary memory in which to save
3422 the offsets during matching if necessary, in order to ensure that
3423 backreferences always work.
3424
3425 4. Increased the compatibility with Perl in a number of ways:
3426
3427 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3428 to request this handling. The option can be set at compile or exec time.
3429
3430 (b) $ matches before a terminating newline by default; an option
3431 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3432 mode). The option can be set at compile or exec time.
3433
3434 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3435 the same as Perl's. If the decimal number it represents is less than 10
3436 or there aren't that many previous left capturing parentheses, an octal
3437 escape is read. Inside a character class, it's always an octal escape,
3438 even if it is a single digit.
3439
3440 (d) An escaped but undefined alphabetic character is taken as a literal,
3441 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3442 escapes.
3443
3444 (e) {0} is now permitted. (The previous item is removed from the compiled
3445 pattern).
3446
3447 5. Changed all the names of code files so that the basic parts are no longer
3448 than 10 characters, and abolished the teeny "globals.c" file.
3449
3450 6. Changed the handling of character classes; they are now done with a 32-byte
3451 bit map always.
3452
3453 7. Added the -d and /D options to pcretest to make it possible to look at the
3454 internals of compilation without having to recompile pcre.
3455
3456
3457 Version 0.95 23-Sep-97
3458 ----------------------
3459
3460 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3461 \x20 at the start of a run of normal characters. These were being treated as
3462 real characters, instead of the source characters being re-checked.
3463
3464
3465 Version 0.94 18-Sep-97
3466 ----------------------
3467
3468 1. The functions are now thread-safe, with the caveat that the global variables
3469 containing pointers to malloc() and free() or alternative functions are the
3470 same for all threads.
3471
3472 2. Get pcre_study() to generate a bitmap of initial characters for non-
3473 anchored patterns when this is possible, and use it if passed to pcre_exec().
3474
3475
3476 Version 0.93 15-Sep-97
3477 ----------------------
3478
3479 1. /(b)|(:+)/ was computing an incorrect first character.
3480
3481 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3482 but not actually doing anything yet.
3483
3484 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3485 as Perl does (e.g. [-az] or [az-]).
3486
3487 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3488 all possible positions.
3489
3490 5. Split up into different modules to avoid including unneeded functions in a
3491 compiled binary. However, compile and exec are still in one module. The "study"
3492 function is split off.
3493
3494 6. The character tables are now in a separate module whose source is generated
3495 by an auxiliary program - but can then be edited by hand if required. There are
3496 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3497 toupper() in the code.
3498
3499 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3500 make them global. Abolish the function for setting them, as the caller can now
3501 set them directly.
3502
3503
3504 Version 0.92 11-Sep-97
3505 ----------------------
3506
3507 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3508 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3509
3510 2. Caseless matching was not working in character classes if the characters in
3511 the pattern were in upper case.
3512
3513 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3514
3515 4. Make PCRE_ANCHORED public and accept as a compile option.
3516
3517 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3518 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3519 pass them.
3520
3521 6. Give an error if bad option bits passed at compile or run time.
3522
3523 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3524 pcretest to cause it to pass that flag.
3525
3526 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3527 options, and the first character, if set.
3528
3529 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3530
3531
3532 Version 0.91 10-Sep-97
3533 ----------------------
3534
3535 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3536 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3537
3538 2. PCRE was looping on encountering an indefinitely repeated back reference to
3539 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3540 Perl does - treats the match as successful.
3541
3542 ****

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