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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 422 - (show annotations) (download)
Fri Aug 14 16:42:55 2009 UTC (4 years, 11 months ago) by ph10
File size: 167156 byte(s)
Fix --regex(p)=pattern not working in pcregrep.

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8 was out-of-date, and there was no check on the pcre_compile() error code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

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