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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 390 - (hide annotations) (download)
Tue Mar 17 10:58:16 2009 UTC (5 years, 8 months ago) by ph10
File size: 162528 byte(s)
Comments about no_arg in the C++ wrapper.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 376 Version 7.9 xx-xxx-09
5     ---------------------
6    
7     1. When building with support for bzlib/zlib (pcregrep) and/or readline
8     (pcretest), all targets were linked against these libraries. This included
9     libpcre, libpcreposix, and libpcrecpp, even though they do not use these
10     libraries. This caused unwanted dependencies to be created. This problem
11     has been fixed, and now only pcregrep is linked with bzlib/zlib and only
12     pcretest is linked with readline.
13    
14     2. The "typedef int BOOL" in pcre_internal.h that was included inside the
15     "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
16     moved outside it again, because FALSE and TRUE are already defined in AIX,
17     but BOOL is not.
18    
19     3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
20     PCRE_MATCH_LIMIT_RETURSION values as ints, when they should be long ints.
21    
22     4. The pcregrep documentation said spaces were inserted as well as colons (or
23     hyphens) following file names and line numbers when outputting matching
24     lines. This is not true; no spaces are inserted. I have also clarified the
25     wording for the --colour (or --color) option.
26 ph10 377
27 ph10 378 5. In pcregrep, when --colour was used with -o, the list of matching strings
28     was not coloured; this is different to GNU grep, so I have changed it to be
29     the same.
30    
31     6. When --colo(u)r was used in pcregrep, only the first matching substring in
32     each matching line was coloured. Now it goes on to look for further matches
33     of any of the test patterns, which is the same behaviour as GNU grep.
34 ph10 379
35     7. A pattern that could match an empty string could cause pcregrep to loop; it
36     doesn't make sense to accept an empty string match in pcregrep, so I have
37     locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
38     seems to be how GNU grep behaves.
39 ph10 380
40     8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
41     start or after a newline", because the conditional assertion was not being
42 ph10 381 correctly handled. The rule now is that both the assertion and what follows
43     in the first alternative must satisfy the test.
44    
45     9. If auto-callout was enabled in a pattern with a conditional group, PCRE
46     could crash during matching.
47 ph10 383
48     10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
49     used for matching.
50 ph10 384
51     11. Unicode property support in character classes was not working for
52 ph10 385 characters (bytes) greater than 127 when not in UTF-8 mode.
53 ph10 386
54     12. Added the -M command line option to pcretest.
55 ph10 388
56     14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
57 ph10 389
58     15. Added the PCRE_NO_START_OPTIMIZE match-time option.
59 ph10 390
60     16. Added comments and documentation about mis-use of no_arg in the C++
61     wrapper.
62 ph10 376
63    
64 ph10 374 Version 7.8 05-Sep-08
65 ph10 349 ---------------------
66    
67     1. Replaced UCP searching code with optimized version as implemented for Ad
68     Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
69 ph10 351 stage table and inline lookup instead of a function, giving speed ups of 2
70     to 5 times on some simple patterns that I tested. Permission was given to
71     distribute the MultiStage2.py script that generates the tables (it's not in
72     the tarball, but is in the Subversion repository).
73 ph10 349
74 ph10 351 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
75     scripts.
76 ph10 358
77 ph10 353 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
78     a group with a zero qualifier. The result of the study could be incorrect,
79 ph10 358 or the function might crash, depending on the pattern.
80    
81     4. Caseless matching was not working for non-ASCII characters in back
82 ph10 354 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
83 ph10 358 It now works when Unicode Property Support is available.
84    
85 ph10 355 5. In pcretest, an escape such as \x{de} in the data was always generating
86     a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
87     non-UTF-8 mode. If the value is greater than 255, it gives a warning about
88 ph10 358 truncation.
89 ph10 349
90 ph10 358 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
91    
92 ph10 357 7. Added two (int) casts to pcregrep when printing the difference of two
93     pointers, in case they are 64-bit values.
94 ph10 371
95     8. Added comments about Mac OS X stack usage to the pcrestack man page and to
96     test 2 if it fails.
97    
98 ph10 359 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
99 ph10 371 and a #define of that name to empty if it is not externally set. This is to
100     allow users of MSVC to set it if necessary.
101    
102     10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
103 ph10 359 the convenience functions in the pcre_get.c source file.
104 ph10 371
105 ph10 360 11. An option change at the start of a pattern that had top-level alternatives
106     could cause overwriting and/or a crash. This command provoked a crash in
107 ph10 371 some environments:
108    
109     printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
110    
111 ph10 360 This potential security problem was recorded as CVE-2008-2371.
112 ph10 371
113 ph10 361 12. For a pattern where the match had to start at the beginning or immediately
114     after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
115 ph10 371 pcre_dfa_exec() could read past the end of the passed subject if there was
116 ph10 361 no match. To help with detecting such bugs (e.g. with valgrind), I modified
117     pcretest so that it places the subject at the end of its malloc-ed buffer.
118 ph10 371
119 ph10 364 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
120 ph10 371 exec() might read past the end of the data buffer in UTF-8 mode.
121    
122 ph10 364 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
123 ph10 371 the data contained the byte 0x85 as part of a UTF-8 character within its
124     first line. This applied both to normal and DFA matching.
125    
126 ph10 366 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
127 ph10 371 /^[^d]*?$/8 failed to match "abc".
128 ph10 351
129 ph10 371 16. Added a missing copyright notice to pcrecpp_internal.h.
130    
131     17. Make it more clear in the documentation that values returned from
132 ph10 368 pcre_exec() in ovector are byte offsets, not character counts.
133 ph10 357
134 ph10 371 18. Tidied a few places to stop certain compilers from issuing warnings.
135 ph10 368
136 ph10 373 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
137     supplied by Stefan Weber. I made a further small update for 7.8 because
138 ph10 374 there is a change of source arrangements: the pcre_searchfuncs.c module is
139     replaced by pcre_ucd.c.
140 ph10 371
141 ph10 373
142 ph10 347 Version 7.7 07-May-08
143 ph10 321 ---------------------
144    
145     1. Applied Craig's patch to sort out a long long problem: "If we can't convert
146 ph10 345 a string to a long long, pretend we don't even have a long long." This is
147 ph10 321 done by checking for the strtoq, strtoll, and _strtoi64 functions.
148 ph10 345
149 ph10 322 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
150     pre-7.6 versions, which defined a global no_arg variable instead of putting
151 ph10 332 it in the RE class. (See also #8 below.)
152 ph10 345
153     3. Remove a line of dead code, identified by coverity and reported by Nuno
154     Lopes.
155    
156 ph10 324 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
157 ph10 321
158 ph10 324 (1) The include/exclude patterns were being applied to the whole pathnames
159 ph10 345 of files, instead of just to the final components.
160    
161 ph10 324 (2) If there was more than one level of directory, the subdirectories were
162     skipped unless they satisfied the include/exclude conditions. This is
163     inconsistent with GNU grep (and could even be seen as contrary to the
164     pcregrep specification - which I improved to make it absolutely clear).
165     The action now is always to scan all levels of directory, and just
166     apply the include/exclude patterns to regular files.
167 ph10 345
168 ph10 325 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
169 ph10 345 --exclude_dir in the tests to avoid scanning .svn directories.
170    
171 ph10 326 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
172 ph10 345 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
173 ph10 326 doesn't support NULs in patterns.
174 ph10 345
175     7. Added some missing "const"s to declarations of static tables in
176     pcre_compile.c and pcre_dfa_exec.c.
177    
178 ph10 328 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
179 ph10 329 caused by fix #2 above. (Subsequently also a second patch to fix the
180 ph10 345 first patch. And a third patch - this was a messy problem.)
181 ph10 321
182 ph10 345 9. Applied Craig's patch to remove the use of push_back().
183    
184 ph10 332 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
185     matching function regexec().
186 ph10 345
187     11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
188     which, however, unlike Perl's \g{...}, are subroutine calls, not back
189 ph10 333 references. PCRE supports relative numbers with this syntax (I don't think
190     Oniguruma does).
191 ph10 345
192     12. Previously, a group with a zero repeat such as (...){0} was completely
193 ph10 335 omitted from the compiled regex. However, this means that if the group
194     was called as a subroutine from elsewhere in the pattern, things went wrong
195 ph10 345 (an internal error was given). Such groups are now left in the compiled
196     pattern, with a new opcode that causes them to be skipped at execution
197 ph10 335 time.
198 ph10 345
199 ph10 341 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
200     to the way PCRE behaves:
201 ph10 345
202 ph10 336 (a) A lone ] character is dis-allowed (Perl treats it as data).
203 ph10 345
204     (b) A back reference to an unmatched subpattern matches an empty string
205 ph10 336 (Perl fails the current match path).
206 ph10 345
207 ph10 341 (c) A data ] in a character class must be notated as \] because if the
208 ph10 345 first data character in a class is ], it defines an empty class. (In
209 ph10 341 Perl it is not possible to have an empty class.) The empty class []
210     never matches; it forces failure and is equivalent to (*FAIL) or (?!).
211     The negative empty class [^] matches any one character, independently
212     of the DOTALL setting.
213 ph10 345
214     14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
215 ph10 340 non-existent subpattern following a character class starting with ']' and
216     containing () gave an internal compiling error instead of "reference to
217     non-existent subpattern". Fortunately, when the pattern did exist, the
218 ph10 345 compiled code was correct. (When scanning forwards to check for the
219 ph10 340 existencd of the subpattern, it was treating the data ']' as terminating
220 ph10 345 the class, so got the count wrong. When actually compiling, the reference
221 ph10 340 was subsequently set up correctly.)
222 ph10 345
223 ph10 341 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
224 ph10 345 it was being rejected as not supported by pcre_dfa_exec(), even though
225     other assertions are supported. I have made pcre_dfa_exec() support
226     (*FAIL).
227    
228 ph10 342 16. The implementation of 13c above involved the invention of a new opcode,
229 ph10 345 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
230     cannot be changed at match time, I realized I could make a small
231     improvement to matching performance by compiling OP_ALLANY instead of
232     OP_ANY for "." when DOTALL was set, and then removing the runtime tests
233     on the OP_ANY path.
234    
235     17. Compiling pcretest on Windows with readline support failed without the
236 ph10 343 following two fixes: (1) Make the unistd.h include conditional on
237 ph10 345 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
238    
239 ph10 344 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
240 ph10 345 ncurses library to be included for pcretest when ReadLine support is
241     requested, but also to allow for it to be overridden. This patch came from
242     Daniel Bergström.
243 ph10 347
244 ph10 346 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
245 ph10 347 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
246     any errors with the current Unicode tables. Thanks to Peter Kankowski for
247     spotting this.
248 ph10 324
249 ph10 332
250 ph10 319 Version 7.6 28-Jan-08
251 ph10 300 ---------------------
252    
253 ph10 302 1. A character class containing a very large number of characters with
254     codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
255     overflow.
256 ph10 309
257     2. Patch to cut out the "long long" test in pcrecpp_unittest when
258     HAVE_LONG_LONG is not defined.
259    
260 ph10 303 3. Applied Christian Ehrlicher's patch to update the CMake build files to
261 ph10 304 bring them up to date and include new features. This patch includes:
262 ph10 309
263 ph10 304 - Fixed PH's badly added libz and libbz2 support.
264     - Fixed a problem with static linking.
265 ph10 312 - Added pcredemo. [But later removed - see 7 below.]
266 ph10 304 - Fixed dftables problem and added an option.
267     - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
268     HAVE_LONG_LONG.
269     - Added readline support for pcretest.
270 ph10 309 - Added an listing of the option settings after cmake has run.
271    
272 ph10 314 4. A user submitted a patch to Makefile that makes it easy to create
273     "pcre.dll" under mingw when using Configure/Make. I added stuff to
274     Makefile.am that cause it to include this special target, without
275     affecting anything else. Note that the same mingw target plus all
276     the other distribution libraries and programs are now supported
277     when configuring with CMake (see 6 below) instead of with
278     Configure/Make.
279 ph10 309
280 ph10 308 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
281     This is an attempt to solve the reported problem "pcrecpp::no_arg is not
282 ph10 309 exported in the Windows port". It has not yet been confirmed that the patch
283     solves the problem, but it does no harm.
284 ph10 313
285 ph10 311 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
286 ph10 319 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
287 ph10 314 with CMake, and also correct the comment about stack recursion.
288 ph10 313
289 ph10 312 7. Remove the automatic building of pcredemo from the ./configure system and
290 ph10 313 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
291     of a program that users should build themselves after PCRE is installed, so
292     building it automatically is not really right. What is more, it gave
293 ph10 312 trouble in some build environments.
294 ph10 300
295 ph10 319 8. Further tidies to CMakeLists.txt from Sheri and Christian.
296 ph10 308
297 ph10 319
298 ph10 298 Version 7.5 10-Jan-08
299 ph10 263 ---------------------
300    
301     1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
302     values in parens when parsing an RE using the C++ wrapper."
303 ph10 286
304 ph10 264 2. Negative specials like \S did not work in character classes in UTF-8 mode.
305     Characters greater than 255 were excluded from the class instead of being
306     included.
307 ph10 286
308     3. The same bug as (2) above applied to negated POSIX classes such as
309 ph10 265 [:^space:].
310 ph10 286
311 ph10 267 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
312     defined or documented. It seems to have been a typo for PCRE_STATIC, so
313 ph10 286 I have changed it.
314    
315 ph10 268 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
316 ph10 286 first named subpattern) and a construct such as (?&a) would reference the
317     first named subpattern whose name started with "a" (in other words, the
318 ph10 272 length check was missing). Both these problems are fixed. "Subpattern name
319     expected" is now given for (?&) (a zero-length name), and this patch also
320     makes it give the same error for \k'' (previously it complained that that
321     was a reference to a non-existent subpattern).
322 ph10 286
323 ph10 269 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
324 ph10 286 this is right because (?- can be followed by option settings as well as by
325 ph10 269 digits. I have, however, made the messages clearer.
326 ph10 286
327 ph10 270 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
328     than the number used in the conditional) now cause a compile-time error.
329     This is actually not compatible with Perl, which accepts such patterns, but
330     treats the conditional as always being FALSE (as PCRE used to), but it
331     seems to me that giving a diagnostic is better.
332 ph10 286
333 ph10 274 8. Change "alphameric" to the more common word "alphanumeric" in comments
334 ph10 275 and messages.
335 ph10 286
336     9. Fix two occurrences of "backslash" in comments that should have been
337     "backspace".
338    
339     10. Remove two redundant lines of code that can never be obeyed (their function
340     was moved elsewhere).
341    
342     11. The program that makes PCRE's Unicode character property table had a bug
343     which caused it to generate incorrect table entries for sequences of
344 ph10 277 characters that have the same character type, but are in different scripts.
345 ph10 286 It amalgamated them into a single range, with the script of the first of
346 ph10 277 them. In other words, some characters were in the wrong script. There were
347     thirteen such cases, affecting characters in the following ranges:
348 ph10 286
349 ph10 277 U+002b0 - U+002c1
350     U+0060c - U+0060d
351 ph10 286 U+0061e - U+00612
352 ph10 277 U+0064b - U+0065e
353     U+0074d - U+0076d
354     U+01800 - U+01805
355     U+01d00 - U+01d77
356     U+01d9b - U+01dbf
357     U+0200b - U+0200f
358     U+030fc - U+030fe
359     U+03260 - U+0327f
360     U+0fb46 - U+0fbb1
361     U+10450 - U+1049d
362 ph10 286
363 ph10 279 12. The -o option (show only the matching part of a line) for pcregrep was not
364     compatible with GNU grep in that, if there was more than one match in a
365     line, it showed only the first of them. It now behaves in the same way as
366     GNU grep.
367 ph10 286
368     13. If the -o and -v options were combined for pcregrep, it printed a blank
369     line for every non-matching line. GNU grep prints nothing, and pcregrep now
370     does the same. The return code can be used to tell if there were any
371     non-matching lines.
372    
373 ph10 289 14. Added --file-offsets and --line-offsets to pcregrep.
374    
375     15. The pattern (?=something)(?R) was not being diagnosed as a potentially
376 ph10 282 infinitely looping recursion. The bug was that positive lookaheads were not
377 ph10 286 being skipped when checking for a possible empty match (negative lookaheads
378     and both kinds of lookbehind were skipped).
379    
380 ph10 289 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
381 ph10 284 inclusion of <windows.h> to before rather than after the definition of
382     INVALID_FILE_ATTRIBUTES (patch from David Byron).
383 ph10 263
384 ph10 289 17. Specifying a possessive quantifier with a specific limit for a Unicode
385 ph10 286 character property caused pcre_compile() to compile bad code, which led at
386     runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
387     are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
388     caused the error; without that there was no problem.
389 ph10 263
390 ph10 289 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
391 ph10 286
392 ph10 289 19. Added --enable-pcretest-libreadline.
393 ph10 286
394 ph10 289 20. In pcrecpp.cc, the variable 'count' was incremented twice in
395 ph10 288 RE::GlobalReplace(). As a result, the number of replacements returned was
396 ph10 298 double what it should be. I removed one of the increments, but Craig sent a
397     later patch that removed the other one (the right fix) and added unit tests
398     that check the return values (which was not done before).
399 ph10 292
400 ph10 291 21. Several CMake things:
401 ph10 287
402 ph10 291 (1) Arranged that, when cmake is used on Unix, the libraries end up with
403 ph10 292 the names libpcre and libpcreposix, not just pcre and pcreposix.
404 ph10 288
405 ph10 292 (2) The above change means that pcretest and pcregrep are now correctly
406     linked with the newly-built libraries, not previously installed ones.
407 ph10 291
408 ph10 292 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
409 ph10 298
410     22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
411     crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
412     UTF-8 newline character). The key issue is that the pattern starts .*;
413     this means that the match must be either at the beginning, or after a
414     newline. The bug was in the code for advancing after a failed match and
415     checking that the new position followed a newline. It was not taking
416 ph10 294 account of UTF-8 characters correctly.
417 ph10 298
418     23. PCRE was behaving differently from Perl in the way it recognized POSIX
419     character classes. PCRE was not treating the sequence [:...:] as a
420     character class unless the ... were all letters. Perl, however, seems to
421     allow any characters between [: and :], though of course it rejects as
422     unknown any "names" that contain non-letters, because all the known class
423     names consist only of letters. Thus, Perl gives an error for [[:1234:]],
424     for example, whereas PCRE did not - it did not recognize a POSIX character
425     class. This seemed a bit dangerous, so the code has been changed to be
426     closer to Perl. The behaviour is not identical to Perl, because PCRE will
427     diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
428     treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
429     Perl does, and where it didn't before.
430    
431 ph10 296 24. Rewrite so as to remove the single use of %n from pcregrep because in some
432     Windows environments %n is disabled by default.
433 ph10 292
434    
435 ph10 260 Version 7.4 21-Sep-07
436 ph10 230 ---------------------
437    
438     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
439 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
440     LF". That isn't really right - the whole point of the change was to try to
441     help when there was an actual mention of one of the two characters. So now
442     the change happens only if \r or \n (or a literal CR or LF) character is
443 ph10 230 encountered.
444 ph10 231
445     2. The 32-bit options word was also used for 6 internal flags, but the numbers
446     of both had grown to the point where there were only 3 bits left.
447     Fortunately, there was spare space in the data structure, and so I have
448     moved the internal flags into a new 16-bit field to free up more option
449 ph10 230 bits.
450 ph10 231
451     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
452     but did not set the internal JCHANGED flag - either of these is enough to
453     control the way the "get" function works - but the PCRE_INFO_JCHANGED
454     facility is supposed to tell if (?J) was ever used, so now (?J) at the
455 ph10 230 start sets both bits.
456    
457 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
458     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
459 ph10 230
460 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
461    
462     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
463 ph10 233 compatibility, even though it is no longer used.
464 ph10 243
465     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
466 ph10 254 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
467 ph10 259 windows.h file is present (where different names are used). [This was
468 ph10 257 reversed later after testing - see 16 below.]
469 ph10 243
470     8. Changed all #include <config.h> to #include "config.h". There were also
471 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
472 ph10 243
473 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
474 ph10 243 sequence off the lines that it output.
475    
476     10. It was pointed out to me that arrays of string pointers cause lots of
477     relocations when a shared library is dynamically loaded. A technique of
478     using a single long string with a table of offsets can drastically reduce
479     these. I have refactored PCRE in four places to do this. The result is
480 ph10 240 dramatic:
481 ph10 243
482 ph10 240 Originally: 290
483     After changing UCP table: 187
484 ph10 243 After changing error message table: 43
485 ph10 240 After changing table of "verbs" 36
486     After changing table of Posix names 22
487 ph10 243
488 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
489 ph10 247
490 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
491 ph10 247 unicode-properties was also set.
492    
493 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
494 ph10 231
495 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
496     checked only for CRLF.
497 ph10 233
498 ph10 259 14. Added casts to pcretest.c to avoid compiler warnings.
499    
500 ph10 256 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
501 ph10 253
502 ph10 257 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
503 ph10 259 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
504     entirely. This removes changes made in 7 above.
505 ph10 256
506 ph10 261 17. The CMake files have been updated, and there is now more information about
507     building with CMake in the NON-UNIX-USE document.
508 ph10 257
509 ph10 261
510 ph10 228 Version 7.3 28-Aug-07
511 ph10 157 ---------------------
512    
513 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
514     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
515     brackets there is not right, since it causes compilers to look for an
516     installed pcre.h, not the version that is in the source that is being
517     compiled (which of course may be different). I have changed it back to:
518 ph10 197
519 ph10 189 #include "pcre.h"
520 ph10 197
521     I have a vague recollection that the change was concerned with compiling in
522     different directories, but in the new build system, that is taken care of
523     by the VPATH setting the Makefile.
524    
525 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
526     when the subject happened to end in the byte 0x85 (e.g. if the last
527     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
528     characters but of course it shouldn't be taken as a newline when it is part
529     of another character. The bug was that, for an unlimited repeat of . in
530     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
531     characters when looking for a newline.
532 ph10 197
533     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
534    
535     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
536     in debug output.
537    
538 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
539     long printing in the pcrecpp unittest when running under MinGW.
540 ph10 189
541 ph10 197 6. ESC_K was left out of the EBCDIC table.
542 ph10 189
543 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
544     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
545     limit also applies to "virtual nesting" when a pattern is recursive, and in
546     this case 1000 isn't so big. I have been able to remove this limit at the
547     expense of backing off one optimization in certain circumstances. Normally,
548     when pcre_exec() would call its internal match() function recursively and
549     immediately return the result unconditionally, it uses a "tail recursion"
550     feature to save stack. However, when a subpattern that can match an empty
551     string has an unlimited repetition quantifier, it no longer makes this
552     optimization. That gives it a stack frame in which to save the data for
553     checking that an empty string has been matched. Previously this was taken
554     from the 1000-entry workspace that had been reserved. So now there is no
555     explicit limit, but more stack is used.
556 ph10 189
557 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
558     syntax that is required for Windows, and which was going wrong for the
559     pcreposix and pcrecpp parts of the library. These were overlooked when this
560     problem was solved for the main library.
561 ph10 197
562 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
563     the size of patterns that contain repeated groups with explicit upper
564     limits. As the maximum quantifier is 65535, the maximum group length was
565     set at 30,000 so that the product of these two numbers did not overflow a
566     32-bit integer. However, it turns out that people want to use groups that
567     are longer than 30,000 bytes (though not repeat them that many times).
568     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
569     made it possible to implement the integer overflow checks in a much more
570     dynamic way, which I have now done. The artificial limitation on group
571     length has been removed - we now have only the limit on the total length of
572     the compiled pattern, which depends on the LINK_SIZE setting.
573 ph10 208
574     10. Fixed a bug in the documentation for get/copy named substring when
575     duplicate names are permitted. If none of the named substrings are set, the
576     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
577     empty string.
578    
579     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
580     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
581     because the ] is interpreted as the first data character and the
582     terminating ] is not found. PCRE has been made compatible with Perl in this
583     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
584     cause memory overwriting.
585    
586 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
587     string has been matched (to stop an infinite loop). It was not recognizing
588 ph10 208 a conditional subpattern that could match an empty string if that
589 ph10 206 subpattern was within another subpattern. For example, it looped when
590 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
591 ph10 206 condition was not nested. This bug has been fixed.
592 ph10 208
593 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
594     past the start of the subject in the presence of bytes with the top bit
595     set, for example "\x8aBCD".
596 ph10 211
597 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
598     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
599 ph10 200
600 ph10 211 14. Optimized (?!) to (*FAIL).
601 ph10 202
602 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
603     This restricts code points to be within the range 0 to 0x10FFFF, excluding
604     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
605     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
606     does: it's just the validity check that is more restrictive.
607 ph10 220
608     16. Inserted checks for integer overflows during escape sequence (backslash)
609     processing, and also fixed erroneous offset values for syntax errors during
610     backslash processing.
611    
612 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
613 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
614    
615 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
616     caused an overrun.
617 ph10 220
618     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
619     something other than just ASCII characters) inside a group that had an
620     unlimited repeat caused a loop at compile time (while checking to see
621     whether the group could match an empty string).
622    
623 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
624     [\P{Any}] did so. (Error in the code for printing property names.)
625 ph10 210
626 ph10 220 21. An orphan \E inside a character class could cause a crash.
627    
628     22. A repeated capturing bracket such as (A)? could cause a wild memory
629 ph10 218 reference during compilation.
630 ph10 220
631     23. There are several functions in pcre_compile() that scan along a compiled
632     expression for various reasons (e.g. to see if it's fixed length for look
633 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
634 ph10 220 present in the pattern. These operators have additional parameters compared
635     with \d, etc, and these were not being taken into account when moving along
636 ph10 218 the compiled data. Specifically:
637 ph10 220
638     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
639     length.
640    
641     (b) An item such as \pL+ within a repeated group could cause crashes or
642 ph10 218 loops.
643 ph10 220
644 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
645 ph10 220 "reference to non-existent subpattern" error.
646    
647 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
648    
649 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
650 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
651 ph10 211
652 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
653 ph10 218
654 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
655 ph10 223 character were causing crashes (broken optimization).
656 ph10 227
657 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
658 ph10 227 \p or \P) caused a compile-time loop.
659    
660 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
661     break. For example, the unstudied pattern [\r\n]A does not match the string
662     "\r\nA" because change 7.0/46 below moves the current point on by two
663     characters after failing to match at the start. However, the pattern \nA
664 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
665     the same is true. There doesn't seem any very clean way out of this, but
666     what I have chosen to do makes the common cases work: PCRE now takes note
667     of whether there can be an explicit match for \r or \n anywhere in the
668     pattern, and if so, 7.0/46 no longer applies. As part of this change,
669     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
670     pattern has explicit CR or LF references.
671 ph10 222
672 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
673 ph10 223
674 ph10 227
675 ph10 189 Version 7.2 19-Jun-07
676     ---------------------
677    
678 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
679     which is apparently normally available under Windows.
680 ph10 159
681     2. Re-jig the pcregrep tests with different newline settings in an attempt
682     to make them independent of the local environment's newline setting.
683 ph10 157
684 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
685 ph10 165
686 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
687     was not 2, because the output contained actual offsets. The recent new
688 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
689     usable with all link sizes.
690    
691 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
692     stack recursion. This gives a massive performance boost under BSD, but just
693 ph10 165 a small improvement under Linux. However, it saves one field in the frame
694 ph10 164 in all cases.
695 ph10 172
696 ph10 166 6. Added more features from the forthcoming Perl 5.10:
697 ph10 172
698 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
699     recursion call. It refers to the nth most recently opened parentheses.
700 ph10 172
701 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
702 ph10 172 to be opened parentheses.
703    
704     (c) Conditions that refer to capturing parentheses can be specified
705 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
706 ph10 172
707 ph10 168 (d) \K resets the start of the current match so that everything before
708 ph10 172 is not part of it.
709    
710 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
711 ph10 172
712 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
713 ph10 172 reference syntax.
714 ph10 182
715 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
716 ph10 182 alternative starts with the same number.
717 ph10 172
718 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
719    
720 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
721     PCRE_INFO_JCHANGED.
722    
723     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
724     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
725 ph10 170 for detecting groups that can match an empty string.
726 ph10 159
727 ph10 172 9. A pattern with a very large number of alternatives (more than several
728     hundred) was running out of internal workspace during the pre-compile
729     phase, where pcre_compile() figures out how much memory will be needed. A
730     bit of new cunning has reduced the workspace needed for groups with
731     alternatives. The 1000-alternative test pattern now uses 12 bytes of
732     workspace instead of running out of the 4096 that are available.
733 ph10 182
734 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
735 ph10 172
736 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
737     The report of the bug said:
738 ph10 182
739 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
740     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
741 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
742 ph10 185
743 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
744 ph10 185 it matched the wrong number of bytes.
745 ph10 172
746 ph10 179
747 ph10 155 Version 7.1 24-Apr-07
748 ph10 98 ---------------------
749    
750 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
751 ph10 122 that is more "standard", making use of automake and other Autotools. There
752 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
753     on this.
754 ph10 111
755     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
756     for recursive directory scanning broke on some systems because the files
757     are not scanned in any specific order and on different systems the order
758     was different. A call to "sort" has been inserted into RunGrepTest for the
759     approprate test as a short-term fix. In the longer term there may be an
760 ph10 100 alternative.
761 ph10 111
762 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
763 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
764     people do their own conversions for various reasons). The problems
765     concerned the use of low-level troff macros .br and .in. I have therefore
766     removed all such uses from the man pages (some were redundant, some could
767 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
768     HTML has been updated to handle .nf/.fi and to complain if it encounters
769     .br or .in.
770 ph10 111
771 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
772 ph10 123 arranged for config.h to be included in the distribution, with the name
773 ph10 111 config.h.generic, for the benefit of those who have to compile without
774     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
775    
776     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
777     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
778 ph10 127 makevp.bat for the current PCRE, using the additional files
779 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
780 ph10 111
781     6. A Windows user reported a minor discrepancy with test 2, which turned out
782     to be caused by a trailing space on an input line that had got lost in his
783 ph10 102 copy. The trailing space was an accident, so I've just removed it.
784 ph10 111
785 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
786 ph10 111 that is needed.
787    
788 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
789     as "const" (a) because they are and (b) because it helps the PHP
790     maintainers who have recently made a script to detect big data structures
791 ph10 111 in the php code that should be moved to the .rodata section. I remembered
792     to update Builducptable as well, so it won't revert if ucptable.h is ever
793 ph10 105 re-created.
794 ph10 111
795     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
796 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
797 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
798     support is not required. This saves 1.5-2K of code, which is important in
799 ph10 107 some applications.
800 ph10 111
801 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
802 ph10 111 so as not to refer to the tables, even though these functions will never be
803     called when UTF-8 support is disabled. Otherwise there are problems with a
804     shared library.
805    
806 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
807    
808     (a) It was defining its arguments as char * instead of void *.
809    
810 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
811     a long time ago when I wrote it, but is no longer the case.
812    
813 ph10 118 The emulated memove() is provided for those environments that have neither
814 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
815 ph10 118 is clearly not the case, as these two bugs were recently reported.
816 ph10 123
817 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
818 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
819     pages, and it removes trailing spaces from listed files. It also creates
820     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
821     case, it wraps all the #defines with #ifndefs. This script should be run
822 ph10 111 before "make dist".
823 ph10 123
824 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
825     with Unicode property support.
826 ph10 123
827     (a) For a maximizing quantifier, if the two different cases of the
828     character were of different lengths in their UTF-8 codings (there are
829     some cases like this - I found 11), and the matching function had to
830 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
831     were both the same length.
832 ph10 123
833     (b) When PCRE was configured to use the heap rather than the stack for
834     recursion during matching, it was not correctly preserving the data for
835     the other case of a UTF-8 character when checking ahead for a match
836     while processing a minimizing repeat. If the check also involved
837 ph10 115 matching a wide character, but failed, corruption could cause an
838     erroneous result when trying to check for a repeat of the original
839     character.
840 ph10 123
841 ph10 116 13. Some tidying changes to the testing mechanism:
842 ph10 98
843 ph10 116 (a) The RunTest script now detects the internal link size and whether there
844     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
845 ph10 123 values substituted by "configure". (The RunGrepTest script already did
846     this for UTF-8.) The configure.ac script no longer substitutes the
847     relevant variables.
848    
849 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
850     with length and offset values. This means that the output is different
851     for different internal link sizes. Test 2 is skipped for link sizes
852     other than 2 because of this, bypassing the problem. Unfortunately,
853     there was also a test in test 3 (the locale tests) that used /B and
854 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
855     I have added a new /Z option to pcretest that replaces the length and
856     offset values with spaces. This is now used to make test 3 independent
857 ph10 122 of link size. (Test 2 will be tidied up later.)
858 ph10 123
859     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
860 ph10 122 segmentation fault instead of returning the appropriate error message.
861 ph10 134
862 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
863 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
864     This doesn't seem right; it now treats the CRLF combination as the line
865     ending, and so does not match in that case. It's only a pattern such as ^$
866     that would hit this one: something like ^ABC$ would have failed after \r
867     and then tried again after \r\n.
868    
869 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
870 ph10 134 in an attempt to make files that differ only in their line terminators
871     compare equal. This works on Linux.
872 ph10 142
873 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
874     as it exited. This is now fixed, thanks to valgrind.
875 ph10 142
876 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
877 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
878     was because its rules for how to advance for /g after matching an empty
879 ph10 143 string at the end of a line did not allow for this case. They now check for
880     it specially.
881 ph10 150
882     20. pcretest is supposed to handle patterns and data of any length, by
883     extending its buffers when necessary. It was getting this wrong when the
884 ph10 147 buffer for a data line had to be extended.
885 ph10 150
886 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
887 ph10 150 CRLF as a newline sequence.
888 ph10 152
889 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
890 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
891     I have nevertheless tidied it up.
892 ph10 111
893 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
894 ph10 123
895 ph10 153 24. Added a man page for pcre-config.
896 ph10 152
897 ph10 153
898 nigel 93 Version 7.0 19-Dec-06
899     ---------------------
900    
901     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
902     moving to gcc 4.1.1.
903    
904     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
905     sys/time.h, which is documented as needed for this function. It doesn't
906     seem to matter on Linux, but it showed up on some releases of OS X.
907    
908     3. It seems that there are systems where bytes whose values are greater than
909     127 match isprint() in the "C" locale. The "C" locale should be the
910     default when a C program starts up. In most systems, only ASCII printing
911     characters match isprint(). This difference caused the output from pcretest
912     to vary, making some of the tests fail. I have changed pcretest so that:
913    
914     (a) When it is outputting text in the compiled version of a pattern, bytes
915     other than 32-126 are always shown as hex escapes.
916    
917     (b) When it is outputting text that is a matched part of a subject string,
918     it does the same, unless a different locale has been set for the match
919     (using the /L modifier). In this case, it uses isprint() to decide.
920    
921     4. Fixed a major bug that caused incorrect computation of the amount of memory
922     required for a compiled pattern when options that changed within the
923     pattern affected the logic of the preliminary scan that determines the
924     length. The relevant options are -x, and -i in UTF-8 mode. The result was
925     that the computed length was too small. The symptoms of this bug were
926     either the PCRE error "internal error: code overflow" from pcre_compile(),
927     or a glibc crash with a message such as "pcretest: free(): invalid next
928     size (fast)". Examples of patterns that provoked this bug (shown in
929     pcretest format) are:
930    
931     /(?-x: )/x
932     /(?x)(?-x: \s*#\s*)/
933     /((?i)[\x{c0}])/8
934     /(?i:[\x{c0}])/8
935    
936     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
937     is now done differently.
938    
939     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
940     wrapper classes; (b) implement a new function in the C++ scanner that is
941     more efficient than the old way of doing things because it avoids levels of
942     recursion in the regex matching; (c) add a paragraph to the documentation
943     for the FullMatch() function.
944    
945     6. The escape sequence \n was being treated as whatever was defined as
946     "newline". Not only was this contrary to the documentation, which states
947     that \n is character 10 (hex 0A), but it also went horribly wrong when
948     "newline" was defined as CRLF. This has been fixed.
949    
950     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
951     was being set to -1 for the "end of line" case (supposedly a value that no
952     character can have). Though this value is never used (the check for end of
953     line is "zero bytes in current character"), it caused compiler complaints.
954     I've changed it to 0xffffffff.
955    
956     8. In pcre_version.c, the version string was being built by a sequence of
957     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
958     string (as it is for production releases) called a macro with an empty
959     argument. The C standard says the result of this is undefined. The gcc
960     compiler treats it as an empty string (which was what was wanted) but it is
961     reported that Visual C gives an error. The source has been hacked around to
962     avoid this problem.
963    
964     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
965     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
966     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
967     of them did).
968    
969     10. Originally, pcretest opened its input and output without "b"; then I was
970     told that "b" was needed in some environments, so it was added for release
971     5.0 to both the input and output. (It makes no difference on Unix-like
972     systems.) Later I was told that it is wrong for the input on Windows. I've
973     now abstracted the modes into two macros, to make it easier to fiddle with
974     them, and removed "b" from the input mode under Windows.
975    
976     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
977    
978     12. Added -help and --help to pcretest as an official way of being reminded
979     of the options.
980    
981     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
982     and pcrecpp.cc because they annoy compilers at high warning levels.
983    
984     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
985    
986     15. Fixed an occurrence of == in configure.ac that should have been = (shell
987     scripts are not C programs :-) and which was not noticed because it works
988     on Linux.
989    
990     16. pcretest is supposed to handle any length of pattern and data line (as one
991     line or as a continued sequence of lines) by extending its input buffer if
992     necessary. This feature was broken for very long pattern lines, leading to
993     a string of junk being passed to pcre_compile() if the pattern was longer
994     than about 50K.
995    
996     17. I have done a major re-factoring of the way pcre_compile() computes the
997     amount of memory needed for a compiled pattern. Previously, there was code
998     that made a preliminary scan of the pattern in order to do this. That was
999     OK when PCRE was new, but as the facilities have expanded, it has become
1000     harder and harder to keep it in step with the real compile phase, and there
1001     have been a number of bugs (see for example, 4 above). I have now found a
1002     cunning way of running the real compile function in a "fake" mode that
1003     enables it to compute how much memory it would need, while actually only
1004     ever using a few hundred bytes of working memory and without too many
1005     tests of the mode. This should make future maintenance and development
1006     easier. A side effect of this work is that the limit of 200 on the nesting
1007     depth of parentheses has been removed (though this was never a serious
1008     limitation, I suspect). However, there is a downside: pcre_compile() now
1009     runs more slowly than before (30% or more, depending on the pattern). I
1010     hope this isn't a big issue. There is no effect on runtime performance.
1011    
1012     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1013     newline (only possible for the last line of a file) and it was a
1014     pattern that set a locale (followed by /Lsomething), pcretest crashed.
1015    
1016     19. Added additional timing features to pcretest. (1) The -tm option now times
1017     matching only, not compiling. (2) Both -t and -tm can be followed, as a
1018     separate command line item, by a number that specifies the number of
1019     repeats to use when timing. The default is 50000; this gives better
1020     precision, but takes uncomfortably long for very large patterns.
1021    
1022     20. Extended pcre_study() to be more clever in cases where a branch of a
1023     subpattern has no definite first character. For example, (a*|b*)[cd] would
1024     previously give no result from pcre_study(). Now it recognizes that the
1025     first character must be a, b, c, or d.
1026    
1027     21. There was an incorrect error "recursive call could loop indefinitely" if
1028     a subpattern (or the entire pattern) that was being tested for matching an
1029     empty string contained only one non-empty item after a nested subpattern.
1030     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1031     incorrectly, because the \d was being skipped in the check.
1032    
1033     22. The pcretest program now has a new pattern option /B and a command line
1034     option -b, which is equivalent to adding /B to every pattern. This causes
1035     it to show the compiled bytecode, without the additional information that
1036     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1037     is the same as /B/I).
1038    
1039     23. A new optimization is now able automatically to treat some sequences such
1040     as a*b as a*+b. More specifically, if something simple (such as a character
1041     or a simple class like \d) has an unlimited quantifier, and is followed by
1042     something that cannot possibly match the quantified thing, the quantifier
1043     is automatically "possessified".
1044    
1045     24. A recursive reference to a subpattern whose number was greater than 39
1046     went wrong under certain circumstances in UTF-8 mode. This bug could also
1047     have affected the operation of pcre_study().
1048    
1049     25. Realized that a little bit of performance could be had by replacing
1050     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1051    
1052     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1053    
1054     27. Possessive quantifiers such as a++ were previously implemented by turning
1055     them into atomic groups such as ($>a+). Now they have their own opcodes,
1056     which improves performance. This includes the automatically created ones
1057     from 23 above.
1058    
1059     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1060     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1061     the first matched character to be a colon. This applied both to named and
1062     numbered groups.
1063    
1064     29. The ucpinternal.h header file was missing its idempotency #ifdef.
1065    
1066     30. I was sent a "project" file called libpcre.a.dev which I understand makes
1067     building PCRE on Windows easier, so I have included it in the distribution.
1068    
1069     31. There is now a check in pcretest against a ridiculously large number being
1070     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1071     loop, the loop is abandoned.
1072    
1073     32. Forward references to subpatterns in conditions such as (?(2)...) where
1074     subpattern 2 is defined later cause pcre_compile() to search forwards in
1075     the pattern for the relevant set of parentheses. This search went wrong
1076     when there were unescaped parentheses in a character class, parentheses
1077     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1078    
1079     33. "Subroutine" calls and backreferences were previously restricted to
1080     referencing subpatterns earlier in the regex. This restriction has now
1081     been removed.
1082    
1083     34. Added a number of extra features that are going to be in Perl 5.10. On the
1084     whole, these are just syntactic alternatives for features that PCRE had
1085     previously implemented using the Python syntax or my own invention. The
1086     other formats are all retained for compatibility.
1087    
1088     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1089     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1090     also .NET compatible.
1091    
1092     (b) A recursion or subroutine call to a named group can now be defined as
1093     (?&name) as well as (?P>name).
1094    
1095     (c) A backreference to a named group can now be defined as \k<name> or
1096     \k'name' as well as (?P=name). The new forms, as well as being in Perl
1097     5.10, are also .NET compatible.
1098    
1099     (d) A conditional reference to a named group can now use the syntax
1100     (?(<name>) or (?('name') as well as (?(name).
1101    
1102     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1103     groups (named and numbered) that are never evaluated inline, but can be
1104     called as "subroutines" from elsewhere. In effect, the DEFINE condition
1105     is always false. There may be only one alternative in such a group.
1106    
1107     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1108     as the simple (?(R). The condition is true only if the most recent
1109     recursion is that of the given number or name. It does not search out
1110     through the entire recursion stack.
1111    
1112     (g) The escape \gN or \g{N} has been added, where N is a positive or
1113     negative number, specifying an absolute or relative reference.
1114    
1115     35. Tidied to get rid of some further signed/unsigned compiler warnings and
1116     some "unreachable code" warnings.
1117    
1118     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1119     things, this adds five new scripts.
1120    
1121     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1122     There were also incompatibilities regarding the handling of \Q..\E inside
1123     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1124     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1125    
1126     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1127     matches an empty string, and forcibly breaks the loop. There were bugs in
1128     this code in non-simple cases. For a pattern such as ^(a()*)* matched
1129     against aaaa the result was just "a" rather than "aaaa", for example. Two
1130     separate and independent bugs (that affected different cases) have been
1131     fixed.
1132    
1133     39. Refactored the code to abolish the use of different opcodes for small
1134     capturing bracket numbers. This is a tidy that I avoided doing when I
1135     removed the limit on the number of capturing brackets for 3.5 back in 2001.
1136     The new approach is not only tidier, it makes it possible to reduce the
1137     memory needed to fix the previous bug (38).
1138    
1139     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1140     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1141     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1142     mode.
1143    
1144     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1145     report.
1146    
1147     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1148     copy construction and assignment in the C++ wrapper.
1149    
1150     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1151     couple of bugs that could have given wrong results in the "--newline=crlf"
1152     case.
1153    
1154     44. Added a number of casts and did some reorganization of signed/unsigned int
1155     variables following suggestions from Dair Grant. Also renamed the variable
1156     "this" as "item" because it is a C++ keyword.
1157    
1158     45. Arranged for dftables to add
1159    
1160     #include "pcre_internal.h"
1161    
1162     to pcre_chartables.c because without it, gcc 4.x may remove the array
1163     definition from the final binary if PCRE is built into a static library and
1164     dead code stripping is activated.
1165    
1166     46. For an unanchored pattern, if a match attempt fails at the start of a
1167     newline sequence, and the newline setting is CRLF or ANY, and the next two
1168     characters are CRLF, advance by two characters instead of one.
1169    
1170    
1171 nigel 91 Version 6.7 04-Jul-06
1172     ---------------------
1173    
1174     1. In order to handle tests when input lines are enormously long, pcretest has
1175     been re-factored so that it automatically extends its buffers when
1176     necessary. The code is crude, but this _is_ just a test program. The
1177     default size has been increased from 32K to 50K.
1178    
1179     2. The code in pcre_study() was using the value of the re argument before
1180     testing it for NULL. (Of course, in any sensible call of the function, it
1181     won't be NULL.)
1182    
1183     3. The memmove() emulation function in pcre_internal.h, which is used on
1184     systems that lack both memmove() and bcopy() - that is, hardly ever -
1185     was missing a "static" storage class specifier.
1186    
1187     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1188     containing an extended class (one that cannot be represented by a bitmap
1189     because it contains high-valued characters or Unicode property items, e.g.
1190     [\pZ]). Almost always one would set UTF-8 mode when processing such a
1191     pattern, but PCRE should not loop if you do not (it no longer does).
1192     [Detail: two cases were found: (a) a repeated subpattern containing an
1193     extended class; (b) a recursive reference to a subpattern that followed a
1194     previous extended class. It wasn't skipping over the extended class
1195     correctly when UTF-8 mode was not set.]
1196    
1197     5. A negated single-character class was not being recognized as fixed-length
1198     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1199     compile error "lookbehind assertion is not fixed length".
1200    
1201     6. The RunPerlTest auxiliary script was showing an unexpected difference
1202     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1203     write a Perl script that can interpret lines of an input file either as
1204     byte characters or as UTF-8, which is what "perltest" was being required to
1205     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1206     can't do is switch easily at run time between having the "use utf8;" pragma
1207     or not. In the end, I fudged it by using the RunPerlTest script to insert
1208     "use utf8;" explicitly for the UTF-8 tests.
1209    
1210     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1211     the end of the subject string, contrary to the documentation and to what
1212     Perl does. This was true of both matching functions. Now it matches only at
1213     the start of the subject and immediately after *internal* newlines.
1214    
1215     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1216     a pointer to an int instead of a pointer to an unsigned long int. This
1217     caused problems on 64-bit systems.
1218    
1219     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1220     instance of the 'standard' template library not being so standard".
1221    
1222     10. There was no check on the number of named subpatterns nor the maximum
1223     length of a subpattern name. The product of these values is used to compute
1224     the size of the memory block for a compiled pattern. By supplying a very
1225     long subpattern name and a large number of named subpatterns, the size
1226     computation could be caused to overflow. This is now prevented by limiting
1227     the length of names to 32 characters, and the number of named subpatterns
1228     to 10,000.
1229    
1230     11. Subpatterns that are repeated with specific counts have to be replicated in
1231     the compiled pattern. The size of memory for this was computed from the
1232     length of the subpattern and the repeat count. The latter is limited to
1233     65535, but there was no limit on the former, meaning that integer overflow
1234     could in principle occur. The compiled length of a repeated subpattern is
1235     now limited to 30,000 bytes in order to prevent this.
1236    
1237     12. Added the optional facility to have named substrings with the same name.
1238    
1239     13. Added the ability to use a named substring as a condition, using the
1240     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1241     are numbers (not recommended). Forward references are permitted.
1242    
1243     14. Added forward references in named backreferences (if you see what I mean).
1244    
1245     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1246     pattern could run off the end of the subject. For example, the pattern
1247     "(?s)(.{1,5})"8 did this with the subject "ab".
1248    
1249     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1250     PCRE_CASELESS was set when matching characters that were quantified with ?
1251     or *.
1252    
1253     17. A character class other than a single negated character that had a minimum
1254     but no maximum quantifier - for example [ab]{6,} - was not handled
1255     correctly by pce_dfa_exec(). It would match only one character.
1256    
1257     18. A valid (though odd) pattern that looked like a POSIX character
1258     class but used an invalid character after [ (for example [[,abc,]]) caused
1259     pcre_compile() to give the error "Failed: internal error: code overflow" or
1260     in some cases to crash with a glibc free() error. This could even happen if
1261     the pattern terminated after [[ but there just happened to be a sequence of
1262     letters, a binary zero, and a closing ] in the memory that followed.
1263    
1264     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1265     over the years. Originally (before any Unicode support), just the bottom 8
1266     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1267     output from "man perlunicode" includes this:
1268    
1269     The regular expression compiler produces polymorphic opcodes. That
1270     is, the pattern adapts to the data and automatically switches to
1271     the Unicode character scheme when presented with Unicode data--or
1272     instead uses a traditional byte scheme when presented with byte
1273     data.
1274    
1275     Sadly, a wide octal escape does not cause a switch, and in a string with
1276     no other multibyte characters, these octal escapes are treated as before.
1277     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1278     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1279     Unicode string.
1280    
1281     I have not perpetrated such confusion in PCRE. Up till now, it took just
1282     the bottom 8 bits, as in old Perl. I have now made octal escapes with
1283     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1284     translate to the appropriate multibyte character.
1285    
1286     29. Applied some refactoring to reduce the number of warnings from Microsoft
1287     and Borland compilers. This has included removing the fudge introduced
1288     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1289     a warning about an unused variable.
1290    
1291     21. PCRE has not included VT (character 0x0b) in the set of whitespace
1292     characters since release 4.0, because Perl (from release 5.004) does not.
1293     [Or at least, is documented not to: some releases seem to be in conflict
1294     with the documentation.] However, when a pattern was studied with
1295     pcre_study() and all its branches started with \s, PCRE still included VT
1296     as a possible starting character. Of course, this did no harm; it just
1297     caused an unnecessary match attempt.
1298    
1299     22. Removed a now-redundant internal flag bit that recorded the fact that case
1300     dependency changed within the pattern. This was once needed for "required
1301     byte" processing, but is no longer used. This recovers a now-scarce options
1302     bit. Also moved the least significant internal flag bit to the most-
1303     significant bit of the word, which was not previously used (hangover from
1304     the days when it was an int rather than a uint) to free up another bit for
1305     the future.
1306    
1307     23. Added support for CRLF line endings as well as CR and LF. As well as the
1308     default being selectable at build time, it can now be changed at runtime
1309     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1310     specify that it is scanning data with non-default line endings.
1311    
1312     24. Changed the definition of CXXLINK to make it agree with the definition of
1313     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1314    
1315     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1316     recursions. This makes a big different to stack usage for some patterns.
1317    
1318     26. If a subpattern containing a named recursion or subroutine reference such
1319     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1320     the space required for the compiled pattern went wrong and gave too small a
1321     value. Depending on the environment, this could lead to "Failed: internal
1322     error: code overflow at offset 49" or "glibc detected double free or
1323     corruption" errors.
1324    
1325     27. Applied patches from Google (a) to support the new newline modes and (b) to
1326     advance over multibyte UTF-8 characters in GlobalReplace.
1327    
1328     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1329     difference for some implementation of PCRE in some Windows version.
1330    
1331     29. Added some extra testing facilities to pcretest:
1332    
1333     \q<number> in a data line sets the "match limit" value
1334     \Q<number> in a data line sets the "match recursion limt" value
1335     -S <number> sets the stack size, where <number> is in megabytes
1336    
1337     The -S option isn't available for Windows.
1338    
1339    
1340 nigel 89 Version 6.6 06-Feb-06
1341     ---------------------
1342    
1343     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1344     in pcreposix.h. I have copied the definition from pcre.h.
1345    
1346     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1347     because pcre.h is no longer a built file.
1348    
1349     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1350     not normally included in the compiled code.
1351    
1352    
1353 nigel 87 Version 6.5 01-Feb-06
1354     ---------------------
1355    
1356     1. When using the partial match feature with pcre_dfa_exec(), it was not
1357     anchoring the second and subsequent partial matches at the new starting
1358     point. This could lead to incorrect results. For example, with the pattern
1359     /1234/, partially matching against "123" and then "a4" gave a match.
1360    
1361     2. Changes to pcregrep:
1362    
1363     (a) All non-match returns from pcre_exec() were being treated as failures
1364     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1365     error message is output. Some extra information is given for the
1366     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1367     probably the only errors that are likely to be caused by users (by
1368     specifying a regex that has nested indefinite repeats, for instance).
1369     If there are more than 20 of these errors, pcregrep is abandoned.
1370    
1371     (b) A binary zero was treated as data while matching, but terminated the
1372     output line if it was written out. This has been fixed: binary zeroes
1373     are now no different to any other data bytes.
1374    
1375     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1376     used to set a locale for matching. The --locale=xxxx long option has
1377     been added (no short equivalent) to specify a locale explicitly on the
1378     pcregrep command, overriding the environment variables.
1379    
1380     (d) When -B was used with -n, some line numbers in the output were one less
1381     than they should have been.
1382    
1383     (e) Added the -o (--only-matching) option.
1384    
1385     (f) If -A or -C was used with -c (count only), some lines of context were
1386     accidentally printed for the final match.
1387    
1388     (g) Added the -H (--with-filename) option.
1389    
1390     (h) The combination of options -rh failed to suppress file names for files
1391     that were found from directory arguments.
1392    
1393     (i) Added the -D (--devices) and -d (--directories) options.
1394    
1395     (j) Added the -F (--fixed-strings) option.
1396    
1397     (k) Allow "-" to be used as a file name for -f as well as for a data file.
1398    
1399     (l) Added the --colo(u)r option.
1400    
1401     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1402     is not present by default.
1403    
1404     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1405     items such as (?R) or (?1), when the recursion could match a number of
1406     alternatives. If it matched one of the alternatives, but subsequently,
1407     outside the recursion, there was a failure, the code tried to back up into
1408     the recursion. However, because of the way PCRE is implemented, this is not
1409     possible, and the result was an incorrect result from the match.
1410    
1411     In order to prevent this happening, the specification of recursion has
1412     been changed so that all such subpatterns are automatically treated as
1413     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1414    
1415     4. I had overlooked the fact that, in some locales, there are characters for
1416     which isalpha() is true but neither isupper() nor islower() are true. In
1417     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1418     and ordfeminine) are like this. This affected the treatment of \w and \W
1419     when they appeared in character classes, but not when they appeared outside
1420     a character class. The bit map for "word" characters is now created
1421     separately from the results of isalnum() instead of just taking it from the
1422     upper, lower, and digit maps. (Plus the underscore character, of course.)
1423    
1424     5. The above bug also affected the handling of POSIX character classes such as
1425     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1426     permanent tables. Instead, the bit maps for such a class were previously
1427     created as the appropriate unions of the upper, lower, and digit bitmaps.
1428     Now they are created by subtraction from the [[:word:]] class, which has
1429     its own bitmap.
1430    
1431     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1432     It is created by subtracting the vertical space characters (\x09, \x0a,
1433     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1434     subtraction was done in the overall bitmap for a character class, meaning
1435     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1436     be recognized. This bug has been fixed.
1437    
1438     7. Patches from the folks at Google:
1439    
1440     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1441     real life, but is still worth protecting against".
1442    
1443     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1444     regular expressions".
1445    
1446     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1447     have it.
1448    
1449     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1450     "configure" and the latter not, in order to fix a problem somebody had
1451     with compiling the Arg class on HP-UX.
1452    
1453     (e) Improve the error-handling of the C++ wrapper a little bit.
1454    
1455     (f) New tests for checking recursion limiting.
1456    
1457     8. The pcre_memmove() function, which is used only if the environment does not
1458     have a standard memmove() function (and is therefore rarely compiled),
1459     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1460     returning a result (though PCRE never actually uses the result).
1461    
1462     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1463     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1464     returned instead of calling malloc() with an overflowing number that would
1465     most likely cause subsequent chaos.
1466    
1467     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1468    
1469     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1470     with this option is matched, the nmatch and pmatch options of regexec() are
1471     ignored.
1472    
1473     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1474     provided in case anyone wants to the the POSIX interface with UTF-8
1475     strings.
1476    
1477     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1478     C++ linking (needed for some HP-UX environments).
1479    
1480     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1481     (unused parameter) and in the pcre_printint() function (omitted "default"
1482     switch label when the default is to do nothing).
1483    
1484     15. Added some code to make it possible, when PCRE is compiled as a C++
1485     library, to replace subject pointers for pcre_exec() with a smart pointer
1486     class, thus making it possible to process discontinuous strings.
1487    
1488     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1489     much the same function. They were added by different people who were trying
1490     to make PCRE easy to compile on non-Unix systems. It has been suggested
1491     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1492     for compiling on Windows systems. I have therefore replaced it with
1493     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1494     defaults to "extern" for C or "extern C" for C++, which works fine on
1495     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1496     SCOPE with something explicit in config.h. In addition:
1497    
1498     (a) pcreposix.h still had just "extern" instead of either of these macros;
1499     I have replaced it with PCRE_DATA_SCOPE.
1500    
1501     (b) Functions such as _pcre_xclass(), which are internal to the library,
1502     but external in the C sense, all had PCRE_EXPORT in their definitions.
1503     This is apparently wrong for the Windows case, so I have removed it.
1504     (It makes no difference on Unix-like systems.)
1505    
1506     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1507     of recursive calls to match(). This is different to MATCH_LIMIT because
1508     that limits the total number of calls to match(), not all of which increase
1509     the depth of recursion. Limiting the recursion depth limits the amount of
1510     stack (or heap if NO_RECURSE is set) that is used. The default can be set
1511     when PCRE is compiled, and changed at run time. A patch from Google adds
1512     this functionality to the C++ interface.
1513    
1514     18. Changes to the handling of Unicode character properties:
1515    
1516     (a) Updated the table to Unicode 4.1.0.
1517    
1518     (b) Recognize characters that are not in the table as "Cn" (undefined).
1519    
1520     (c) I revised the way the table is implemented to a much improved format
1521     which includes recognition of ranges. It now supports the ranges that
1522     are defined in UnicodeData.txt, and it also amalgamates other
1523     characters into ranges. This has reduced the number of entries in the
1524     table from around 16,000 to around 3,000, thus reducing its size
1525     considerably. I realized I did not need to use a tree structure after
1526     all - a binary chop search is just as efficient. Having reduced the
1527     number of entries, I extended their size from 6 bytes to 8 bytes to
1528     allow for more data.
1529    
1530     (d) Added support for Unicode script names via properties such as \p{Han}.
1531    
1532     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1533     matching that character.
1534    
1535     20. When matching a repeated Unicode property with a minimum greater than zero,
1536     (for example \pL{2,}), PCRE could look past the end of the subject if it
1537     reached it while seeking the minimum number of characters. This could
1538     happen only if some of the characters were more than one byte long, because
1539     there is a check for at least the minimum number of bytes.
1540    
1541     21. Refactored the implementation of \p and \P so as to be more general, to
1542     allow for more different types of property in future. This has changed the
1543     compiled form incompatibly. Anybody with saved compiled patterns that use
1544     \p or \P will have to recompile them.
1545    
1546     22. Added "Any" and "L&" to the supported property types.
1547    
1548     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1549     but give a compile time error if the value is greater than 0xff.
1550    
1551     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1552     accidentally not being installed or uninstalled.
1553    
1554     25. The pcre.h file was built from pcre.h.in, but the only changes that were
1555     made were to insert the current release number. This seemed silly, because
1556     it made things harder for people building PCRE on systems that don't run
1557     "configure". I have turned pcre.h into a distributed file, no longer built
1558     by "configure", with the version identification directly included. There is
1559     no longer a pcre.h.in file.
1560    
1561     However, this change necessitated a change to the pcre-config script as
1562     well. It is built from pcre-config.in, and one of the substitutions was the
1563     release number. I have updated configure.ac so that ./configure now finds
1564     the release number by grepping pcre.h.
1565    
1566     26. Added the ability to run the tests under valgrind.
1567    
1568    
1569 nigel 85 Version 6.4 05-Sep-05
1570     ---------------------
1571    
1572     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1573     "--" to be printed when multiple files were scanned, even when none of the
1574     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1575     consider it to be a bug, and have restored the previous behaviour.
1576    
1577     2. A couple of code tidies to get rid of compiler warnings.
1578    
1579     3. The pcretest program used to cheat by referring to symbols in the library
1580     whose names begin with _pcre_. These are internal symbols that are not
1581     really supposed to be visible externally, and in some environments it is
1582     possible to suppress them. The cheating is now confined to including
1583     certain files from the library's source, which is a bit cleaner.
1584    
1585     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1586     file's purpose clearer.
1587    
1588     5. Reorganized pcre_ucp_findchar().
1589    
1590    
1591 nigel 83 Version 6.3 15-Aug-05
1592     ---------------------
1593    
1594     1. The file libpcre.pc.in did not have general read permission in the tarball.
1595    
1596     2. There were some problems when building without C++ support:
1597    
1598     (a) If C++ support was not built, "make install" and "make test" still
1599     tried to test it.
1600    
1601     (b) There were problems when the value of CXX was explicitly set. Some
1602     changes have been made to try to fix these, and ...
1603    
1604     (c) --disable-cpp can now be used to explicitly disable C++ support.
1605    
1606     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1607     backslash in a target when C++ was disabled. This confuses some
1608     versions of "make", apparently. Using an intermediate variable solves
1609     this. (Same for CPP_LOBJ.)
1610    
1611     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1612     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1613     necessary on certain architectures.
1614    
1615     4. Added a setting of -export-symbols-regex to the link command to remove
1616     those symbols that are exported in the C sense, but actually are local
1617     within the library, and not documented. Their names all begin with
1618     "_pcre_". This is not a perfect job, because (a) we have to except some
1619     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1620     available (and never for static libraries). I have made a note to try to
1621     find a way round (a) in the future.
1622    
1623    
1624 nigel 81 Version 6.2 01-Aug-05
1625     ---------------------
1626    
1627     1. There was no test for integer overflow of quantifier values. A construction
1628     such as {1111111111111111} would give undefined results. What is worse, if
1629     a minimum quantifier for a parenthesized subpattern overflowed and became
1630     negative, the calculation of the memory size went wrong. This could have
1631     led to memory overwriting.
1632    
1633     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1634    
1635     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1636     operating environments where this matters.
1637    
1638     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1639     PCRE options from within the C++ wrapper.
1640    
1641     5. Named capturing subpatterns were not being correctly counted when a pattern
1642     was compiled. This caused two problems: (a) If there were more than 100
1643     such subpatterns, the calculation of the memory needed for the whole
1644     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1645     back references of the form \12, where the number was greater than 9, were
1646     not recognized as back references, even though there were sufficient
1647     previous subpatterns.
1648    
1649     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1650     versions of gcc, e.g. 2.95.4.
1651    
1652    
1653 nigel 79 Version 6.1 21-Jun-05
1654     ---------------------
1655    
1656     1. There was one reference to the variable "posix" in pcretest.c that was not
1657     surrounded by "#if !defined NOPOSIX".
1658    
1659     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1660     the cross-check on the old pcre_info() function, for the benefit of the
1661     cut-down version of PCRE that is currently imported into Exim.
1662    
1663     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1664     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1665     patterns that start (?i) but I don't think that matters. The use of (?i) is
1666     just an example; this all applies to the other options as well.
1667    
1668     4. Since libtool seems to echo the compile commands it is issuing, the output
1669     from "make" can be reduced a bit by putting "@" in front of each libtool
1670     compile command.
1671    
1672     5. Patch from the folks at Google for configure.in to be a bit more thorough
1673     in checking for a suitable C++ installation before trying to compile the
1674     C++ stuff. This should fix a reported problem when a compiler was present,
1675     but no suitable headers.
1676    
1677     6. The man pages all had just "PCRE" as their title. I have changed them to
1678     be the relevant file name. I have also arranged that these names are
1679     retained in the file doc/pcre.txt, which is a concatenation in text format
1680     of all the man pages except the little individual ones for each function.
1681    
1682     7. The NON-UNIX-USE file had not been updated for the different set of source
1683     files that come with release 6. I also added a few comments about the C++
1684     wrapper.
1685    
1686    
1687 nigel 77 Version 6.0 07-Jun-05
1688     ---------------------
1689    
1690     1. Some minor internal re-organization to help with my DFA experiments.
1691    
1692     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1693     didn't matter for the library itself when fully configured, but did matter
1694     when compiling without UCP support, or within Exim, where the ucp files are
1695     not imported.
1696    
1697     3. Refactoring of the library code to split up the various functions into
1698     different source modules. The addition of the new DFA matching code (see
1699     below) to a single monolithic source would have made it really too
1700     unwieldy, quite apart from causing all the code to be include in a
1701     statically linked application, when only some functions are used. This is
1702     relevant even without the DFA addition now that patterns can be compiled in
1703     one application and matched in another.
1704    
1705     The downside of splitting up is that there have to be some external
1706     functions and data tables that are used internally in different modules of
1707     the library but which are not part of the API. These have all had their
1708     names changed to start with "_pcre_" so that they are unlikely to clash
1709     with other external names.
1710    
1711     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1712     a different (DFA) algorithm. Although it is slower than the original
1713     function, it does have some advantages for certain types of matching
1714     problem.
1715    
1716     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1717     including restarting after a partial match.
1718    
1719     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1720     defined when compiling for Windows was sent to me. I have put it into the
1721     code, though I have no means of testing or verifying it.
1722    
1723     7. Added the pcre_refcount() auxiliary function.
1724    
1725     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1726     match before or at the first newline in the subject string. In pcretest,
1727     the /f option on a pattern can be used to set this.
1728    
1729     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1730     would behave wrongly. This has been present in PCRE since release 4.0.
1731    
1732     10. A number of changes to the pcregrep command:
1733    
1734     (a) Refactored how -x works; insert ^(...)$ instead of setting
1735     PCRE_ANCHORED and checking the length, in preparation for adding
1736     something similar for -w.
1737    
1738     (b) Added the -w (match as a word) option.
1739    
1740     (c) Refactored the way lines are read and buffered so as to have more
1741     than one at a time available.
1742    
1743     (d) Implemented a pcregrep test script.
1744    
1745     (e) Added the -M (multiline match) option. This allows patterns to match
1746     over several lines of the subject. The buffering ensures that at least
1747     8K, or the rest of the document (whichever is the shorter) is available
1748     for matching (and similarly the previous 8K for lookbehind assertions).
1749    
1750     (f) Changed the --help output so that it now says
1751    
1752     -w, --word-regex(p)
1753    
1754     instead of two lines, one with "regex" and the other with "regexp"
1755     because that confused at least one person since the short forms are the
1756     same. (This required a bit of code, as the output is generated
1757     automatically from a table. It wasn't just a text change.)
1758    
1759     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1760     option but starts with a hyphen. Could be a pattern or a path name
1761     starting with a hyphen, for instance.
1762    
1763     (h) "-" can be given as a file name to represent stdin.
1764    
1765     (i) When file names are being printed, "(standard input)" is used for
1766     the standard input, for compatibility with GNU grep. Previously
1767     "<stdin>" was used.
1768    
1769     (j) The option --label=xxx can be used to supply a name to be used for
1770     stdin when file names are being printed. There is no short form.
1771    
1772     (k) Re-factored the options decoding logic because we are going to add
1773     two more options that take data. Such options can now be given in four
1774     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1775    
1776     (l) Added the -A, -B, and -C options for requesting that lines of context
1777     around matches be printed.
1778    
1779     (m) Added the -L option to print the names of files that do not contain
1780     any matching lines, that is, the complement of -l.
1781    
1782     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1783     continue to scan other files.
1784    
1785     (o) The -s option was incorrectly implemented. For compatibility with other
1786     greps, it now suppresses the error message for a non-existent or non-
1787     accessible file (but not the return code). There is a new option called
1788     -q that suppresses the output of matching lines, which was what -s was
1789     previously doing.
1790    
1791     (p) Added --include and --exclude options to specify files for inclusion
1792     and exclusion when recursing.
1793    
1794     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1795     Hopefully, it now does.
1796    
1797     12. Missing cast in pcre_study().
1798    
1799     13. Added an "uninstall" target to the makefile.
1800    
1801     14. Replaced "extern" in the function prototypes in Makefile.in with
1802     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1803     world, but is set differently for Windows.
1804    
1805     15. Added a second compiling function called pcre_compile2(). The only
1806     difference is that it has an extra argument, which is a pointer to an
1807     integer error code. When there is a compile-time failure, this is set
1808     non-zero, in addition to the error test pointer being set to point to an
1809     error message. The new argument may be NULL if no error number is required
1810     (but then you may as well call pcre_compile(), which is now just a
1811     wrapper). This facility is provided because some applications need a
1812     numeric error indication, but it has also enabled me to tidy up the way
1813     compile-time errors are handled in the POSIX wrapper.
1814    
1815     16. Added VPATH=.libs to the makefile; this should help when building with one
1816     prefix path and installing with another. (Or so I'm told by someone who
1817     knows more about this stuff than I do.)
1818    
1819     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1820     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1821     match everything, including newlines. This is not POSIX-compatible, but
1822     somebody wanted the feature. From pcretest it can be activated by using
1823     both the P and the s flags.
1824    
1825     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1826    
1827     19. libpcre.pc was being incorrectly installed as executable.
1828    
1829     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1830     it now also looks for '\r' so that it will work unmodified on Windows.
1831    
1832     21. Added Google's contributed C++ wrapper to the distribution.
1833    
1834     22. Added some untidy missing memory free() calls in pcretest, to keep
1835     Electric Fence happy when testing.
1836    
1837    
1838    
1839 nigel 75 Version 5.0 13-Sep-04
1840     ---------------------
1841    
1842     1. Internal change: literal characters are no longer packed up into items
1843     containing multiple characters in a single byte-string. Each character
1844     is now matched using a separate opcode. However, there may be more than one
1845     byte in the character in UTF-8 mode.
1846    
1847     2. The pcre_callout_block structure has two new fields: pattern_position and
1848     next_item_length. These contain the offset in the pattern to the next match
1849     item, and its length, respectively.
1850    
1851     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1852     insertion of callouts before each pattern item. Added the /C option to
1853     pcretest to make use of this.
1854    
1855     4. On the advice of a Windows user, the lines
1856    
1857     #if defined(_WIN32) || defined(WIN32)
1858     _setmode( _fileno( stdout ), 0x8000 );
1859     #endif /* defined(_WIN32) || defined(WIN32) */
1860    
1861     have been added to the source of pcretest. This apparently does useful
1862     magic in relation to line terminators.
1863    
1864     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1865     for the benefit of those environments where the "b" makes a difference.
1866    
1867     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1868     to know about it. I have put a hack into configure.in that adds in code
1869     to set GCC=yes if CC=icc. This seems to end up at a point in the
1870     generated configure script that is early enough to affect the setting of
1871     compiler options, which is what is needed, but I have no means of testing
1872     whether it really works. (The user who reported this had patched the
1873     generated configure script, which of course I cannot do.)
1874    
1875     LATER: After change 22 below (new libtool files), the configure script
1876     seems to know about icc (and also ecc). Therefore, I have commented out
1877     this hack in configure.in.
1878    
1879     7. Added support for pkg-config (2 patches were sent in).
1880    
1881     8. Negated POSIX character classes that used a combination of internal tables
1882     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1883     [[:^ascii]]. Typically, they would match almost any characters. The other
1884     POSIX classes were not broken in this way.
1885    
1886     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1887     to find the match, as PCRE was deluded into thinking that the match had to
1888     start at the start point or following a newline. The same bug applied to
1889     patterns with negative forward assertions or any backward assertions
1890     preceding ".*" at the start, unless the pattern required a fixed first
1891     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1892    
1893     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1894     starting at the last subject character, bytes beyond the end of the subject
1895     string were read.
1896    
1897     11. Renamed the variable "class" as "classbits" to make life easier for C++
1898     users. (Previously there was a macro definition, but it apparently wasn't
1899     enough.)
1900    
1901     12. Added the new field "tables" to the extra data so that tables can be passed
1902     in at exec time, or the internal tables can be re-selected. This allows
1903     a compiled regex to be saved and re-used at a later time by a different
1904     program that might have everything at different addresses.
1905    
1906     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1907     -R library as well as a -L library.
1908    
1909     14. The debugging options of pcretest (-d on the command line or D on a
1910     pattern) showed incorrect output for anything following an extended class
1911     that contained multibyte characters and which was followed by a quantifier.
1912    
1913     15. Added optional support for general category Unicode character properties
1914     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1915     support. It adds about 90K to the size of the library. The meanings of the
1916     inbuilt class escapes such as \d and \s have NOT been changed.
1917    
1918     16. Updated pcredemo.c to include calls to free() to release the memory for the
1919     compiled pattern.
1920    
1921     17. The generated file chartables.c was being created in the source directory
1922     instead of in the building directory. This caused the build to fail if the
1923     source directory was different from the building directory, and was
1924     read-only.
1925    
1926     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1927     file. No doubt somebody will tell me if they don't make sense... Also added
1928     Dan Mooney's comments about building on OpenVMS.
1929    
1930     19. Added support for partial matching via the PCRE_PARTIAL option for
1931     pcre_exec() and the \P data escape in pcretest.
1932    
1933     20. Extended pcretest with 3 new pattern features:
1934    
1935     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1936     write the compiled pattern to the file whose name is "rest-of-line".
1937     This is a straight binary dump of the data, with the saved pointer to
1938     the character tables forced to be NULL. The study data, if any, is
1939     written too. After writing, pcretest reads a new pattern.
1940    
1941     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1942     compiled pattern from the given file. There must not be any
1943     occurrences of "<" in the file name (pretty unlikely); if there are,
1944     pcretest will instead treat the initial "<" as a pattern delimiter.
1945     After reading in the pattern, pcretest goes on to read data lines as
1946     usual.
1947    
1948     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1949     and 16-bit fields in a compiled pattern, to simulate a pattern that
1950     was compiled on a host of opposite endianness.
1951    
1952     21. The pcre-exec() function can now cope with patterns that were compiled on
1953     hosts of opposite endianness, with this restriction:
1954    
1955     As for any compiled expression that is saved and used later, the tables
1956     pointer field cannot be preserved; the extra_data field in the arguments
1957     to pcre_exec() should be used to pass in a tables address if a value
1958     other than the default internal tables were used at compile time.
1959    
1960     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1961     now diagnosed as an error. Previously, most of the time, a negative number
1962     would have been treated as zero, but if in addition "ovector" was passed as
1963     NULL, a crash could occur.
1964    
1965     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1966     new versions from the libtool 1.5 distribution (the last one is a copy of
1967     a file called libtool.m4). This seems to have fixed the need to patch
1968     "configure" to support Darwin 1.3 (which I used to do). However, I still
1969     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1970     workstation).
1971    
1972     24. Changed the PCRE licence to be the more standard "BSD" licence.
1973    
1974    
1975 nigel 73 Version 4.5 01-Dec-03
1976     ---------------------
1977    
1978     1. There has been some re-arrangement of the code for the match() function so
1979     that it can be compiled in a version that does not call itself recursively.
1980     Instead, it keeps those local variables that need separate instances for
1981     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1982     needs to "recurse". Keeping track of where control must go is done by means
1983     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1984     hide most of the details from the main code, and operates only if
1985     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1986     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1987     operating.
1988    
1989     To make it easier for callers to provide specially tailored get/free
1990     functions for this usage, two new functions, pcre_stack_malloc, and
1991     pcre_stack_free, are used. They are always called in strict stacking order,
1992     and the size of block requested is always the same.
1993    
1994     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1995     PCRE has been compiled to use the stack or the heap for recursion. The
1996     -C option of pcretest uses this to show which version is compiled.
1997    
1998     A new data escape \S, is added to pcretest; it causes the amounts of store
1999     obtained and freed by both kinds of malloc/free at match time to be added
2000     to the output.
2001    
2002     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2003     what's available on my current Linux desktop machine.
2004    
2005     3. When matching a UTF-8 string, the test for a valid string at the start has
2006     been extended. If start_offset is not zero, PCRE now checks that it points
2007     to a byte that is the start of a UTF-8 character. If not, it returns
2008     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2009     this is necessary because there may be backward assertions in the pattern.
2010     When matching the same subject several times, it may save resources to use
2011     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2012    
2013     4. The code for checking the validity of UTF-8 strings has been tightened so
2014     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2015     containing "overlong sequences".
2016    
2017     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2018     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2019     should have been "&", but it just so happened that all the cases this let
2020     through by mistake were picked up later in the function.
2021    
2022     6. I had used a variable called "isblank" - this is a C99 function, causing
2023     some compilers to warn. To avoid this, I renamed it (as "blankclass").
2024    
2025     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2026     prompting; (b) run "./pcretest /dev/null" at the start of the test script
2027     so the version is shown; (c) stop "make test" echoing "./RunTest".
2028    
2029     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2030    
2031     9. The prototype for memmove() for systems that don't have it was using
2032     size_t, but the inclusion of the header that defines size_t was later. I've
2033     moved the #includes for the C headers earlier to avoid this.
2034    
2035     10. Added some adjustments to the code to make it easier to compiler on certain
2036     special systems:
2037    
2038     (a) Some "const" qualifiers were missing.
2039     (b) Added the macro EXPORT before all exported functions; by default this
2040     is defined to be empty.
2041     (c) Changed the dftables auxiliary program (that builds chartables.c) so
2042     that it reads its output file name as an argument instead of writing
2043     to the standard output and assuming this can be redirected.
2044    
2045     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2046     class containing characters with values greater than 255, PCRE compilation
2047     went into a loop.
2048    
2049     12. A recursive reference to a subpattern that was within another subpattern
2050     that had a minimum quantifier of zero caused PCRE to crash. For example,
2051     (x(y(?2))z)? provoked this bug with a subject that got as far as the
2052     recursion. If the recursively-called subpattern itself had a zero repeat,
2053     that was OK.
2054    
2055     13. In pcretest, the buffer for reading a data line was set at 30K, but the
2056     buffer into which it was copied (for escape processing) was still set at
2057     1024, so long lines caused crashes.
2058    
2059     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2060     "internal error: code overflow...". This applied to any character class
2061     that was followed by a possessive quantifier.
2062    
2063     15. Modified the Makefile to add libpcre.la as a prerequisite for
2064     libpcreposix.la because I was told this is needed for a parallel build to
2065     work.
2066    
2067     16. If a pattern that contained .* following optional items at the start was
2068     studied, the wrong optimizing data was generated, leading to matching
2069     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2070     matching string must start with a or b or c. The correct conclusion for
2071     this pattern is that a match can start with any character.
2072    
2073    
2074 nigel 71 Version 4.4 13-Aug-03
2075     ---------------------
2076    
2077     1. In UTF-8 mode, a character class containing characters with values between
2078     127 and 255 was not handled correctly if the compiled pattern was studied.
2079     In fixing this, I have also improved the studying algorithm for such
2080     classes (slightly).
2081    
2082     2. Three internal functions had redundant arguments passed to them. Removal
2083     might give a very teeny performance improvement.
2084    
2085     3. Documentation bug: the value of the capture_top field in a callout is *one
2086     more than* the number of the hightest numbered captured substring.
2087    
2088     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2089     in incorrectly linking with a previously installed version. They now link
2090     explicitly with libpcre.la.
2091    
2092     5. configure.in no longer needs to recognize Cygwin specially.
2093    
2094     6. A problem in pcre.in for Windows platforms is fixed.
2095    
2096     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2097     pcretest, it used to include the size of the study block as part of its
2098     output. Unfortunately, the structure contains a field that has a different
2099     size on different hardware architectures. This meant that the tests that
2100     showed this size failed. As the block is currently always of a fixed size,
2101     this information isn't actually particularly useful in pcretest output, so
2102     I have just removed it.
2103    
2104     8. Three pre-processor statements accidentally did not start in column 1.
2105     Sadly, there are *still* compilers around that complain, even though
2106     standard C has not required this for well over a decade. Sigh.
2107    
2108     9. In pcretest, the code for checking callouts passed small integers in the
2109     callout_data field, which is a void * field. However, some picky compilers
2110     complained about the casts involved for this on 64-bit systems. Now
2111     pcretest passes the address of the small integer instead, which should get
2112     rid of the warnings.
2113    
2114     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2115     both compile and run time, and gives an error if an invalid UTF-8 sequence
2116     is found. There is a option for disabling this check in cases where the
2117     string is known to be correct and/or the maximum performance is wanted.
2118    
2119     11. In response to a bug report, I changed one line in Makefile.in from
2120    
2121     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2122     to
2123     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2124    
2125     to look similar to other lines, but I have no way of telling whether this
2126     is the right thing to do, as I do not use Windows. No doubt I'll get told
2127     if it's wrong...
2128    
2129    
2130 nigel 69 Version 4.3 21-May-03
2131     ---------------------
2132    
2133     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2134     Makefile.
2135    
2136     2. Some refactoring to improve the quality of the code:
2137    
2138     (i) The utf8_table... variables are now declared "const".
2139    
2140     (ii) The code for \cx, which used the "case flipping" table to upper case
2141     lower case letters, now just substracts 32. This is ASCII-specific,
2142     but the whole concept of \cx is ASCII-specific, so it seems
2143     reasonable.
2144    
2145     (iii) PCRE was using its character types table to recognize decimal and
2146     hexadecimal digits in the pattern. This is silly, because it handles
2147     only 0-9, a-f, and A-F, but the character types table is locale-
2148     specific, which means strange things might happen. A private
2149     table is now used for this - though it costs 256 bytes, a table is
2150     much faster than multiple explicit tests. Of course, the standard
2151     character types table is still used for matching digits in subject
2152     strings against \d.
2153    
2154     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2155     ending in _t are). So I've renamed it as ESC_tee.
2156    
2157     3. The first argument for regexec() in the POSIX wrapper should have been
2158     defined as "const".
2159    
2160     4. Changed pcretest to use malloc() for its buffers so that they can be
2161     Electric Fenced for debugging.
2162    
2163     5. There were several places in the code where, in UTF-8 mode, PCRE would try
2164     to read one or more bytes before the start of the subject string. Often this
2165     had no effect on PCRE's behaviour, but in some circumstances it could
2166     provoke a segmentation fault.
2167    
2168     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2169     to try to read one or more bytes before the start of the subject string.
2170    
2171     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2172     UTF-8 support could misbehave in various ways if the subject string
2173     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2174     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2175     back over UTF-8 characters.)
2176    
2177    
2178 nigel 67 Version 4.2 14-Apr-03
2179     ---------------------
2180    
2181     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2182    
2183     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2184     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2185     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2186     [WIN_PREFIX]: new variable, "cyg" for Cygwin
2187     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2188     and BUILD_EXEEXT
2189     Note: automatic setting of the BUILD variables is not yet working
2190     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2191     compile-time but not at link-time
2192     [LINK]: use for linking executables only
2193     make different versions for Windows and non-Windows
2194     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2195     libraries
2196     [LINK_FOR_BUILD]: new variable
2197     [OBJEXT]: use throughout
2198     [EXEEXT]: use throughout
2199     <winshared>: new target
2200     <wininstall>: new target
2201     <dftables.o>: use native compiler
2202     <dftables>: use native linker
2203     <install>: handle Windows platform correctly
2204     <clean>: ditto
2205     <check>: ditto
2206     copy DLL to top builddir before testing
2207    
2208     As part of these changes, -no-undefined was removed again. This was reported
2209     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2210     in any case.
2211    
2212     3. Some tidies to get rid of compiler warnings:
2213    
2214     . In the match_data structure, match_limit was an unsigned long int, whereas
2215     match_call_count was an int. I've made them both unsigned long ints.
2216    
2217     . In pcretest the fact that a const uschar * doesn't automatically cast to
2218     a void * provoked a warning.
2219    
2220     . Turning on some more compiler warnings threw up some "shadow" variables
2221     and a few more missing casts.
2222    
2223     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2224     option, a class that contained a single character with a value between 128
2225     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2226    
2227     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2228     option, a class that contained several characters, but with at least one
2229     whose value was between 128 and 255 caused PCRE to crash.
2230    
2231    
2232 nigel 65 Version 4.1 12-Mar-03
2233     ---------------------
2234 nigel 63
2235 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
2236     needed, and a string in dftables.c that was longer than standard compilers are
2237     required to support.
2238    
2239     2. Compiling with Sun's compiler found a few more places where the code could
2240     be tidied up in order to avoid warnings.
2241    
2242     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2243     first of these names is deprecated in the latest Autoconf in favour of the name
2244     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2245     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2246     analogy I have changed it to CFLAGS_FOR_BUILD.
2247    
2248     4. Added -no-undefined to the linking command in the Makefile, because this is
2249     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2250     linking step for the pcreposix library.
2251    
2252     5. PCRE was failing to diagnose the case of two named groups with the same
2253     name.
2254    
2255     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2256     literal character that is needed in the subject for a match, and scans along to
2257     ensure that it is present before embarking on the full matching process. This
2258     saves time in cases of nested unlimited repeats that are never going to match.
2259     Problem: the scan can take a lot of time if the subject is very long (e.g.
2260     megabytes), thus penalizing straightforward matches. It is now done only if the
2261     amount of subject to be scanned is less than 1000 bytes.
2262    
2263     7. A lesser problem with the same optimization is that it was recording the
2264     first character of an anchored pattern as "needed", thus provoking a search
2265     right along the subject, even when the first match of the pattern was going to
2266     fail. The "needed" character is now not set for anchored patterns, unless it
2267     follows something in the pattern that is of non-fixed length. Thus, it still
2268     fulfils its original purpose of finding quick non-matches in cases of nested
2269     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2270    
2271    
2272     Version 4.0 17-Feb-03
2273     ---------------------
2274    
2275 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
2276     extended to the end of string, PCRE compiled incorrect data. This could lead to
2277     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2278    
2279     2. Moved to autoconf 2.53 and libtool 1.4.2.
2280    
2281     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2282     the special perltest8 script is no longer needed - all the tests can be run
2283     from a single perltest script.
2284    
2285     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2286     by \s. It has now been removed in PCRE. This means it isn't recognized as
2287     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2288     class [:space:] *does* include VT, thereby creating a mess.
2289    
2290     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2291     space and tab.
2292    
2293     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2294     its new features into the main test script, reducing the number of scripts.
2295    
2296     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2297     were backward compatible, and made the (?i) apply to the whole pattern, as if
2298     /i were given. Now it behaves more logically, and applies the option setting
2299     only to what follows. PCRE has been changed to follow suit. However, if it
2300     finds options settings right at the start of the pattern, it extracts them into
2301     the global options, as before. Thus, they show up in the info data.
2302    
2303     8. Added support for the \Q...\E escape sequence. Characters in between are
2304     treated as literals. This is slightly different from Perl in that $ and @ are
2305     also handled as literals inside the quotes. In Perl, they will cause variable
2306     interpolation. Note the following examples:
2307    
2308     Pattern PCRE matches Perl matches
2309    
2310     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2311     \Qabc\$xyz\E abc\$xyz abc\$xyz
2312     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2313    
2314     For compatibility with Perl, \Q...\E sequences are recognized inside character
2315     classes as well as outside them.
2316    
2317     9. Re-organized 3 code statements in pcretest to avoid "overflow in
2318     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2319     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2320     signed/unsigned warnings.
2321    
2322     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2323     option for pcretest, so I've replaced it by a simple function that does just
2324     that job.
2325    
2326     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2327     "pcregrep -".
2328    
2329     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2330     Java package. This provides some syntactic sugar for simple cases of what my
2331     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2332     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2333     item, you can use this simplified notation. Note that only makes sense with
2334     greedy quantifiers. Consequently, the use of the possessive quantifier forces
2335     greediness, whatever the setting of the PCRE_UNGREEDY option.
2336    
2337     13. A change of greediness default within a pattern was not taking effect at
2338     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2339     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2340     was abstracted outside.
2341    
2342     14. PCRE now supports the \G assertion. It is true when the current matching
2343     position is at the start point of the match. This differs from \A when the
2344     starting offset is non-zero. Used with the /g option of pcretest (or similar
2345     code), it works in the same way as it does for Perl's /g option. If all
2346     alternatives of a regex begin with \G, the expression is anchored to the start
2347     match position, and the "anchored" flag is set in the compiled expression.
2348    
2349     15. Some bugs concerning the handling of certain option changes within patterns
2350     have been fixed. These applied to options other than (?ims). For example,
2351     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2352     been the other way round. Some of this was related to change 7 above.
2353    
2354     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2355     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2356     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2357     POSIX classes only within a class (e.g. /[[:alpha:]]/).
2358    
2359     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2360     mode. Unlike ".", it always matches newline, whatever the setting of
2361     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2362     assertions. Perl allows it, but it doesn't (in general) work because it can't
2363     calculate the length of the lookbehind. At least, that's the case for Perl
2364     5.8.0 - I've been told they are going to document that it doesn't work in
2365     future.
2366    
2367     18. Added an error diagnosis for escapes that PCRE does not support: these are
2368     \L, \l, \N, \P, \p, \U, \u, and \X.
2369    
2370     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2371     reading past the end of the pattern in cases such as /[abcd/.
2372    
2373     20. PCRE was getting more memory than necessary for patterns with classes that
2374     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2375    
2376     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2377     compiling PCRE for use with Virtual Pascal.
2378    
2379     22. Small fix to the Makefile to make it work properly if the build is done
2380     outside the source tree.
2381    
2382     23. Added a new extension: a condition to go with recursion. If a conditional
2383     subpattern starts with (?(R) the "true" branch is used if recursion has
2384     happened, whereas the "false" branch is used only at the top level.
2385    
2386     24. When there was a very long string of literal characters (over 255 bytes
2387     without UTF support, over 250 bytes with UTF support), the computation of how
2388     much memory was required could be incorrect, leading to segfaults or other
2389     strange effects.
2390    
2391     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2392     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2393     there was a subsequent back reference to those brackets. This meant that, for
2394     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2395     possible to check for precisely this case. All we can do is abandon the
2396     optimization if .* occurs inside capturing brackets when there are any back
2397     references whatsoever. (See below for a better fix that came later.)
2398    
2399     26. The handling of the optimization for finding the first character of a
2400     non-anchored pattern, and for finding a character that is required later in the
2401     match were failing in some cases. This didn't break the matching; it just
2402     failed to optimize when it could. The way this is done has been re-implemented.
2403    
2404     27. Fixed typo in error message for invalid (?R item (it said "(?p").
2405    
2406     28. Added a new feature that provides some of the functionality that Perl
2407     provides with (?{...}). The facility is termed a "callout". The way it is done
2408     in PCRE is for the caller to provide an optional function, by setting
2409     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2410     global variable. By default it is unset, which disables all calling out. To get
2411     the function called, the regex must include (?C) at appropriate points. This
2412     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2413     This provides a means of identifying different callout points. When PCRE
2414     reaches such a point in the regex, if pcre_callout has been set, the external
2415     function is called. It is provided with data in a structure called
2416     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2417     matching continues; if it returns a non-zero value, the match at the current
2418     point fails. However, backtracking will occur if possible. [This was changed
2419     later and other features added - see item 49 below.]
2420    
2421     29. pcretest is upgraded to test the callout functionality. It provides a
2422     callout function that displays information. By default, it shows the start of
2423     the match and the current position in the text. There are some new data escapes
2424     to vary what happens:
2425    
2426     \C+ in addition, show current contents of captured substrings
2427     \C- do not supply a callout function
2428     \C!n return 1 when callout number n is reached
2429     \C!n!m return 1 when callout number n is reached for the mth time
2430    
2431     30. If pcregrep was called with the -l option and just a single file name, it
2432     output "<stdin>" if a match was found, instead of the file name.
2433    
2434     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2435     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2436     pcre_exec(). This saves a malloc/free per call. The default value of
2437     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2438     when configuring.
2439    
2440     32. The default maximum size of a compiled pattern is 64K. There have been a
2441     few cases of people hitting this limit. The code now uses macros to handle the
2442     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2443     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2444     configuring. Tests 2 and 5 work only with 2-byte links because they output
2445     debugging information about compiled patterns.
2446    
2447     33. Internal code re-arrangements:
2448    
2449     (a) Moved the debugging function for printing out a compiled regex into
2450     its own source file (printint.c) and used #include to pull it into
2451     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2452     separate copies.
2453    
2454     (b) Defined the list of op-code names for debugging as a macro in
2455     internal.h so that it is next to the definition of the opcodes.
2456    
2457     (c) Defined a table of op-code lengths for simpler skipping along compiled
2458     code. This is again a macro in internal.h so that it is next to the
2459     definition of the opcodes.
2460    
2461     34. Added support for recursive calls to individual subpatterns, along the
2462     lines of Robin Houston's patch (but implemented somewhat differently).
2463    
2464     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2465     allow it to read and process whole directories in Win32. This code was
2466     contributed by Lionel Fourquaux; it has not been tested by me.
2467    
2468     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2469     used to name a group. Names consist of alphanumerics and underscores, and must
2470     be unique. Back references use the syntax (?P=name) and recursive calls use
2471     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2472     numbers. The function pcre_fullinfo() can be used after compilation to extract
2473     a name/number map. There are three relevant calls:
2474    
2475     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2476     PCRE_INFO_NAMECOUNT yields the number of entries
2477     PCRE_INFO_NAMETABLE yields a pointer to the map.
2478    
2479     The map is a vector of fixed-size entries. The size of each entry depends on
2480     the length of the longest name used. The first two bytes of each entry are the
2481     group number, most significant byte first. There follows the corresponding
2482     name, zero terminated. The names are in alphabetical order.
2483    
2484     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2485     case instead of 255. Making it the same both with and without UTF-8 support
2486     means that the same test output works with both.
2487    
2488     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2489     calling malloc() with a zero argument.
2490    
2491     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2492     optimization. I've improved things by keeping a bitmap of backreferences with
2493     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2494     fact referenced, the optimization can be applied. It is unlikely that a
2495     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2496     the match to follow \n) will appear inside brackets with a number greater than
2497     31, but if it does, any back reference > 31 suppresses the optimization.
2498    
2499     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2500     of disabling numbered capturing parentheses. Any opening parenthesis that is
2501     not followed by ? behaves as if it were followed by ?: but named parentheses
2502     can still be used for capturing (and they will acquire numbers in the usual
2503     way).
2504    
2505     41. Redesigned the return codes from the match() function into yes/no/error so
2506     that errors can be passed back from deep inside the nested calls. A malloc
2507     failure while inside a recursive subpattern call now causes the
2508     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2509    
2510     42. It is now possible to set a limit on the number of times the match()
2511     function is called in a call to pcre_exec(). This facility makes it possible to
2512     limit the amount of recursion and backtracking, though not in a directly
2513     obvious way, because the match() function is used in a number of different
2514     circumstances. The count starts from zero for each position in the subject
2515     string (for non-anchored patterns). The default limit is, for compatibility, a
2516     large number, namely 10 000 000. You can change this in two ways:
2517    
2518     (a) When configuring PCRE before making, you can use --with-match-limit=n
2519     to set a default value for the compiled library.
2520    
2521     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2522     a different value is set. See 45 below.
2523    
2524     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2525    
2526     43. Added a new function pcre_config(int, void *) to enable run-time extraction
2527     of things that can be changed at compile time. The first argument specifies
2528     what is wanted and the second points to where the information is to be placed.
2529     The current list of available information is:
2530    
2531     PCRE_CONFIG_UTF8
2532    
2533     The output is an integer that is set to one if UTF-8 support is available;
2534     otherwise it is set to zero.
2535    
2536     PCRE_CONFIG_NEWLINE
2537    
2538     The output is an integer that it set to the value of the code that is used for
2539     newline. It is either LF (10) or CR (13).
2540    
2541     PCRE_CONFIG_LINK_SIZE
2542    
2543     The output is an integer that contains the number of bytes used for internal
2544     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2545    
2546     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2547    
2548     The output is an integer that contains the threshold above which the POSIX
2549     interface uses malloc() for output vectors. See item 31 above.
2550    
2551     PCRE_CONFIG_MATCH_LIMIT
2552    
2553     The output is an unsigned integer that contains the default limit of the number
2554     of match() calls in a pcre_exec() execution. See 42 above.
2555    
2556     44. pcretest has been upgraded by the addition of the -C option. This causes it
2557     to extract all the available output from the new pcre_config() function, and to
2558     output it. The program then exits immediately.
2559    
2560     45. A need has arisen to pass over additional data with calls to pcre_exec() in
2561     order to support additional features. One way would have been to define
2562     pcre_exec2() (for example) with extra arguments, but this would not have been
2563     extensible, and would also have required all calls to the original function to
2564     be mapped to the new one. Instead, I have chosen to extend the mechanism that
2565     is used for passing in "extra" data from pcre_study().
2566    
2567     The pcre_extra structure is now exposed and defined in pcre.h. It currently
2568     contains the following fields:
2569    
2570     flags a bitmap indicating which of the following fields are set
2571     study_data opaque data from pcre_study()
2572     match_limit a way of specifying a limit on match() calls for a specific
2573     call to pcre_exec()
2574     callout_data data for callouts (see 49 below)
2575    
2576     The flag bits are also defined in pcre.h, and are
2577    
2578     PCRE_EXTRA_STUDY_DATA
2579     PCRE_EXTRA_MATCH_LIMIT
2580     PCRE_EXTRA_CALLOUT_DATA
2581    
2582     The pcre_study() function now returns one of these new pcre_extra blocks, with
2583     the actual study data pointed to by the study_data field, and the
2584     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2585     before. That is, this change is entirely upwards-compatible and requires no
2586     change to existing code.
2587    
2588     If you want to pass in additional data to pcre_exec(), you can either place it
2589     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2590     block.
2591    
2592     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2593     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2594     times with different match limits, until it finds the minimum value needed for
2595     pcre_exec() to complete. The value is then output. This can be instructive; for
2596     most simple matches the number is quite small, but for pathological cases it
2597     gets very large very quickly.
2598    
2599     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2600     returns the size of the data block pointed to by the study_data field in a
2601     pcre_extra block, that is, the value that was passed as the argument to
2602     pcre_malloc() when PCRE was getting memory in which to place the information
2603     created by pcre_study(). The fourth argument should point to a size_t variable.
2604     pcretest has been extended so that this information is shown after a successful
2605     pcre_study() call when information about the compiled regex is being displayed.
2606    
2607     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2608     because what follows is always an absolute path. (Later: it turns out that this
2609     is more than cosmetic for MinGW, because it doesn't like empty path
2610     components.)
2611    
2612     49. Some changes have been made to the callout feature (see 28 above):
2613    
2614     (i) A callout function now has three choices for what it returns:
2615    
2616     0 => success, carry on matching
2617     > 0 => failure at this point, but backtrack if possible
2618     < 0 => serious error, return this value from pcre_exec()
2619    
2620     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2621     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2622     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2623     use by callout functions. It will never be used by PCRE itself.
2624    
2625     (ii) The pcre_extra structure (see 45 above) has a void * field called
2626     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2627     pcre_callout_block structure has a field of the same name. The contents of
2628     the field passed in the pcre_extra structure are passed to the callout
2629     function in the corresponding field in the callout block. This makes it
2630     easier to use the same callout-containing regex from multiple threads. For
2631     testing, the pcretest program has a new data escape
2632    
2633     \C*n pass the number n (may be negative) as callout_data
2634    
2635     If the callout function in pcretest receives a non-zero value as
2636     callout_data, it returns that value.
2637    
2638     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2639     there were some redundant $(CFLAGS) in commands that are now specified as
2640     $(LINK), which already includes $(CFLAGS).
2641    
2642     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2643     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2644     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2645     one-byte characters throughout. Note that case-insensitive matching applies
2646     only to characters whose values are less than 256. PCRE doesn't support the
2647     notion of cases for higher-valued characters.
2648    
2649     (i) A character class whose characters are all within 0-255 is handled as
2650     a bit map, and the map is inverted for negative classes. Previously, a
2651     character > 255 always failed to match such a class; however it should
2652     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2653    
2654     (ii) A negated character class with a single character < 255 is coded as
2655     "not this character" (OP_NOT). This wasn't working properly when the test
2656     character was multibyte, either singly or repeated.
2657    
2658     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2659     mode, for example: \x{100}{2,3}.
2660    
2661     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2662     singly or repeated) now correctly test multibyte characters. However,
2663     PCRE doesn't recognize any characters with values greater than 255 as
2664     digits, spaces, or word characters. Such characters always match \D, \S,
2665     and \W, and never match \d, \s, or \w.
2666    
2667     (v) Classes may now contain characters and character ranges with values
2668     greater than 255. For example: [ab\x{100}-\x{400}].
2669    
2670     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2671     PCRE in UTF-8 mode.
2672    
2673     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2674     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2675     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2676     value.)
2677    
2678     53. The single man page has become too large. I have therefore split it up into
2679     a number of separate man pages. These also give rise to individual HTML pages;
2680     these are now put in a separate directory, and there is an index.html page that
2681     lists them all. Some hyperlinking between the pages has been installed.
2682    
2683     54. Added convenience functions for handling named capturing parentheses.
2684    
2685     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2686     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2687     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2688     are faulted.
2689    
2690     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2691     calling configure. These values are used when compiling the dftables.c program
2692     which is run to generate the source of the default character tables. They
2693     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2694     you will need to set these values.
2695    
2696     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2697    
2698    
2699     Version 3.9 02-Jan-02
2700 nigel 61 ---------------------
2701    
2702     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2703    
2704     2. If --disable-static was given, the building process failed when trying to
2705     build pcretest and pcregrep. (For some reason it was using libtool to compile
2706     them, which is not right, as they aren't part of the library.)
2707    
2708    
2709 nigel 59 Version 3.8 18-Dec-01
2710     ---------------------
2711    
2712     1. The experimental UTF-8 code was completely screwed up. It was packing the
2713     bytes in the wrong order. How dumb can you get?
2714    
2715    
2716 nigel 57 Version 3.7 29-Oct-01
2717     ---------------------
2718    
2719     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2720     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2721     this didn't happen under Solaris 8, where I normally test things.
2722    
2723 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2724     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2725     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2726     this fix an hour or so after the initial 3.7 release.)
2727 nigel 57
2728 nigel 59
2729 nigel 55 Version 3.6 23-Oct-01
2730     ---------------------
2731    
2732     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2733     offsets passed as NULL with zero offset count.
2734    
2735     2. The config.guess and config.sub files had not been updated when I moved to
2736     the latest autoconf.
2737    
2738    
2739 nigel 53 Version 3.5 15-Aug-01
2740     ---------------------
2741 nigel 5
2742 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2743     had been forgotten.
2744    
2745     2. By using declared but undefined structures, we can avoid using "void"
2746     definitions in pcre.h while keeping the internal definitions of the structures
2747     private.
2748    
2749     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2750     user point of view, this means that both static and shared libraries are built
2751     by default, but this can be individually controlled. More of the work of
2752     handling this static/shared cases is now inside libtool instead of PCRE's make
2753     file.
2754    
2755     4. The pcretest utility is now installed along with pcregrep because it is
2756     useful for users (to test regexs) and by doing this, it automatically gets
2757     relinked by libtool. The documentation has been turned into a man page, so
2758     there are now .1, .txt, and .html versions in /doc.
2759    
2760     5. Upgrades to pcregrep:
2761     (i) Added long-form option names like gnu grep.
2762     (ii) Added --help to list all options with an explanatory phrase.
2763     (iii) Added -r, --recursive to recurse into sub-directories.
2764     (iv) Added -f, --file to read patterns from a file.
2765    
2766     6. pcre_exec() was referring to its "code" argument before testing that
2767     argument for NULL (and giving an error if it was NULL).
2768    
2769     7. Upgraded Makefile.in to allow for compiling in a different directory from
2770     the source directory.
2771    
2772     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2773     options bits, the pointer it was passed was to an int instead of to an unsigned
2774     long int. This mattered only on 64-bit systems.
2775    
2776     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2777     generated) instead of pcre.in, which it its source. Also made the same change
2778     in several of the .c files.
2779    
2780     10. A new release of gcc defines printf() as a macro, which broke pcretest
2781     because it had an ifdef in the middle of a string argument for printf(). Fixed
2782     by using separate calls to printf().
2783    
2784     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2785     script, to force use of CR or LF instead of \n in the source. On non-Unix
2786     systems, the value can be set in config.h.
2787    
2788     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2789     absolute limit. Changed the text of the error message to make this clear, and
2790     likewise updated the man page.
2791    
2792     13. The limit of 99 on the number of capturing subpatterns has been removed.
2793     The new limit is 65535, which I hope will not be a "real" limit.
2794    
2795    
2796 nigel 51 Version 3.4 22-Aug-00
2797     ---------------------
2798    
2799     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2800    
2801     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2802    
2803    
2804 nigel 49 Version 3.3 01-Aug-00
2805     ---------------------
2806    
2807     1. If an octal character was given, but the value was greater than \377, it
2808     was not getting masked to the least significant bits, as documented. This could
2809     lead to crashes in some systems.
2810    
2811     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2812     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2813    
2814     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2815     These just pass their arguments on to (pcre_free)(), but they are provided
2816     because some uses of PCRE bind it to non-C systems that can call its functions,
2817     but cannot call free() or pcre_free() directly.
2818    
2819     4. Add "make test" as a synonym for "make check". Corrected some comments in
2820     the Makefile.
2821    
2822     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2823     Makefile.
2824    
2825     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2826     command called pgrep for grepping around the active processes.
2827    
2828     7. Added the beginnings of support for UTF-8 character strings.
2829    
2830     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2831     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2832     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2833     out for the ar command.)
2834    
2835    
2836 nigel 47 Version 3.2 12-May-00
2837     ---------------------
2838    
2839     This is purely a bug fixing release.
2840    
2841     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2842     of ZA. This was just one example of several cases that could provoke this bug,
2843     which was introduced by change 9 of version 2.00. The code for breaking
2844     infinite loops after an iteration that matches an empty string was't working
2845     correctly.
2846    
2847     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2848     when matched against abbab (for example). After matching an empty string, it
2849     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2850     caused it to match further down the string than it should.
2851    
2852     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2853     was there because it doesn't seem to be needed, and it causes trouble on some
2854     systems, as it is not a Standard C header. It has been removed.
2855    
2856     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2857     were reported on the Macintosh. The changes were from
2858    
2859     while ((c = *(++ptr)) != 0 && c != '\n');
2860     to
2861     while ((c = *(++ptr)) != 0 && c != '\n') ;
2862    
2863     Totally extraordinary, but if that's what it takes...
2864    
2865     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2866     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2867     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2868     assumes the way PCRE uses memmove() (always moving upwards).
2869    
2870     6. PCRE is being used in one environment where strchr() is not available. There
2871     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2872     faster code anyway.
2873    
2874    
2875 nigel 45 Version 3.1 09-Feb-00
2876     ---------------------
2877    
2878     The only change in this release is the fixing of some bugs in Makefile.in for
2879     the "install" target:
2880    
2881     (1) It was failing to install pcreposix.h.
2882    
2883     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2884    
2885    
2886 nigel 43 Version 3.0 01-Feb-00
2887     ---------------------
2888 nigel 41
2889     1. Add support for the /+ modifier to perltest (to output $` like it does in
2890     pcretest).
2891    
2892     2. Add support for the /g modifier to perltest.
2893    
2894     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2895     matches null strings.
2896    
2897     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2898     pattern. Perl treats empty patterns specially - it reuses the most recent
2899     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2900     effect.
2901    
2902     5. The POSIX interface was broken in that it was just handing over the POSIX
2903     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2904     required a bigger vector, with some working space on the end. This means that
2905     the POSIX wrapper now has to get and free some memory, and copy the results.
2906    
2907 nigel 43 6. Added some simple autoconf support, placing the test data and the
2908     documentation in separate directories, re-organizing some of the
2909     information files, and making it build pcre-config (a GNU standard). Also added
2910     libtool support for building PCRE as a shared library, which is now the
2911     default.
2912 nigel 41
2913 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2914     09 are not valid octal constants. Single digits will be used for minor values
2915     less than 10.
2916    
2917     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2918     existing programs that set these in the POSIX interface can use PCRE without
2919     modification.
2920    
2921     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2922     return all that pcre_info() returns, plus additional data. The pcre_info()
2923     function is retained for compatibility, but is considered to be obsolete.
2924    
2925     10. Added experimental recursion feature (?R) to handle one common case that
2926     Perl 5.6 will be able to do with (?p{...}).
2927    
2928     11. Added support for POSIX character classes like [:alpha:], which Perl is
2929     adopting.
2930    
2931    
2932 nigel 39 Version 2.08 31-Aug-99
2933     ----------------------
2934    
2935     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2936     trying to match at the startoffset position, but instead was moving forward to
2937     the next newline as if a previous match had failed.
2938    
2939     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2940     and could get into a loop if a null string was matched other than at the start
2941     of the subject.
2942    
2943     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2944     be distinguished at compile time, and for completeness also added PCRE_DATE.
2945    
2946     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2947     in GnuWin32 environments.
2948    
2949    
2950 nigel 37 Version 2.07 29-Jul-99
2951     ----------------------
2952    
2953     1. The documentation is now supplied in plain text form and HTML as well as in
2954     the form of man page sources.
2955    
2956     2. C++ compilers don't like assigning (void *) values to other pointer types.
2957     In particular this affects malloc(). Although there is no problem in Standard
2958     C, I've put in casts to keep C++ compilers happy.
2959    
2960     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2961     should be (const char *).
2962    
2963     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2964     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2965     However, I haven't made this a standard facility. The documentation doesn't
2966     mention it, and the Makefile doesn't support it.
2967    
2968     5. The Makefile now contains an "install" target, with editable destinations at
2969     the top of the file. The pcretest program is not installed.
2970    
2971     6. pgrep -V now gives the PCRE version number and date.
2972    
2973     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2974     causing the entire string to be ignored, instead of just the last character.
2975    
2976     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2977     non-matching string, it can take a very, very long time, even for strings of
2978     quite modest length, because of the nested recursion. PCRE now does better in
2979     some of these cases. It does this by remembering the last required literal
2980     character in the pattern, and pre-searching the subject to ensure it is present
2981     before running the real match. In other words, it applies a heuristic to detect
2982     some types of certain failure quickly, and in the above example, if presented
2983     with a string that has no trailing " it gives "no match" very quickly.
2984    
2985     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2986     other alternatives are tried instead.
2987    
2988    
2989 nigel 35 Version 2.06 09-Jun-99
2990     ----------------------
2991    
2992     1. Change pcretest's output for amount of store used to show just the code
2993     space, because the remainder (the data block) varies in size between 32-bit and
2994     64-bit systems.
2995    
2996     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2997     start matching at. This allows lookbehinds to work when searching for multiple
2998     occurrences in a string.
2999    
3000     3. Added additional options to pcretest for testing multiple occurrences:
3001    
3002     /+ outputs the rest of the string that follows a match
3003     /g loops for multiple occurrences, using the new startoffset argument
3004     /G loops for multiple occurrences by passing an incremented pointer
3005    
3006     4. PCRE wasn't doing the "first character" optimization for patterns starting
3007     with \b or \B, though it was doing it for other lookbehind assertions. That is,
3008     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3009     the letter 'x'. On long subject strings, this gives a significant speed-up.
3010    
3011    
3012 nigel 33 Version 2.05 21-Apr-99
3013     ----------------------
3014    
3015     1. Changed the type of magic_number from int to long int so that it works
3016     properly on 16-bit systems.
3017    
3018     2. Fixed a bug which caused patterns starting with .* not to work correctly
3019     when the subject string contained newline characters. PCRE was assuming
3020     anchoring for such patterns in all cases, which is not correct because .* will
3021     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3022     DOTALL is set at top level; otherwise it knows that patterns starting with .*
3023     must be retried after every newline in the subject.
3024    
3025    
3026 nigel 31 Version 2.04 18-Feb-99
3027     ----------------------
3028    
3029     1. For parenthesized subpatterns with repeats whose minimum was zero, the
3030     computation of the store needed to hold the pattern was incorrect (too large).
3031     If such patterns were nested a few deep, this could multiply and become a real
3032     problem.
3033    
3034     2. Added /M option to pcretest to show the memory requirement of a specific
3035     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3036    
3037     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3038     compiled in such a way that the backtracking after subsequent failure was
3039     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3040     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3041    
3042    
3043 nigel 29 Version 2.03 02-Feb-99
3044     ----------------------
3045    
3046     1. Fixed typo and small mistake in man page.
3047    
3048 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
3049     LICENCE file containing the conditions.
3050 nigel 29
3051     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3052     Perl, that is the internal \ allows the delimiter to be included in the
3053     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3054     the final delimiter, add \ to the end of the pattern (to test the error).
3055    
3056     4. Added the convenience functions for extracting substrings after a successful
3057     match. Updated pcretest to make it able to test these functions.
3058    
3059    
3060 nigel 27 Version 2.02 14-Jan-99
3061     ----------------------
3062    
3063     1. Initialized the working variables associated with each extraction so that
3064     their saving and restoring doesn't refer to uninitialized store.
3065    
3066     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3067     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3068     fix the problem.
3069    
3070     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3071     calls, and wasn't printing the correct value for compiling calls. Increased the
3072     default value of LOOPREPEAT, and the number of significant figures in the
3073     times.
3074    
3075     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3076    
3077     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3078     a building problem on Windows NT with a FAT file system.
3079    
3080    
3081 nigel 25 Version 2.01 21-Oct-98
3082     ----------------------
3083    
3084     1. Changed the API for pcre_compile() to allow for the provision of a pointer
3085     to character tables built by pcre_maketables() in the current locale. If NULL
3086     is passed, the default tables are used.
3087    
3088    
3089 nigel 23 Version 2.00 24-Sep-98
3090 nigel 21 ----------------------
3091    
3092 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3093     it any more.
3094    
3095     2. Allow quantification of (?>) groups, and make it work correctly.
3096    
3097     3. The first character computation wasn't working for (?>) groups.
3098    
3099     4. Correct the implementation of \Z (it is permitted to match on the \n at the
3100     end of the subject) and add 5.005's \z, which really does match only at the
3101     very end of the subject.
3102    
3103     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3104    
3105     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3106     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3107     localized options. All options to pcre_study() were also removed.
3108    
3109     7. Add other new features from 5.005:
3110    
3111     $(?<= positive lookbehind
3112     $(?<! negative lookbehind
3113     (?imsx-imsx) added the unsetting capability
3114     such a setting is global if at outer level; local otherwise
3115     (?imsx-imsx:) non-capturing groups with option setting
3116     (?(cond)re|re) conditional pattern matching
3117    
3118     A backreference to itself in a repeated group matches the previous
3119     captured string.
3120    
3121     8. General tidying up of studying (both automatic and via "study")
3122     consequential on the addition of new assertions.
3123    
3124     9. As in 5.005, unlimited repeated groups that could match an empty substring
3125     are no longer faulted at compile time. Instead, the loop is forcibly broken at
3126     runtime if any iteration does actually match an empty substring.
3127    
3128     10. Include the RunTest script in the distribution.
3129    
3130     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3131     discrepancies, some of which were old and were also with respect to 5.004. They
3132     have now been fixed.
3133    
3134    
3135     Version 1.09 28-Apr-98
3136     ----------------------
3137    
3138 nigel 21 1. A negated single character class followed by a quantifier with a minimum
3139     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3140     program crashes, or just wrong answers. This did not apply to negated classes
3141     containing more than one character, or to minima other than one.
3142    
3143    
3144 nigel 19 Version 1.08 27-Mar-98
3145     ----------------------
3146    
3147     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3148    
3149     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3150     latter must appear before anything that relies on it in the pattern.
3151    
3152    
3153 nigel 17 Version 1.07 16-Feb-98
3154     ----------------------
3155    
3156     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3157     repeat of a potentially empty string).
3158    
3159    
3160 nigel 15 Version 1.06 23-Jan-98
3161     ----------------------
3162    
3163     1. Added Markus Oberhumer's little patches for C++.
3164    
3165     2. Literal strings longer than 255 characters were broken.
3166    
3167    
3168 nigel 13 Version 1.05 23-Dec-97
3169     ----------------------
3170    
3171     1. Negated character classes containing more than one character were failing if
3172     PCRE_CASELESS was set at run time.
3173    
3174    
3175 nigel 11 Version 1.04 19-Dec-97
3176     ----------------------
3177    
3178     1. Corrected the man page, where some "const" qualifiers had been omitted.
3179    
3180     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3181     input syntax.
3182    
3183     3. Fixed memory leak which occurred when a regex with back references was
3184     matched with an offsets vector that wasn't big enough. The temporary memory
3185     that is used in this case wasn't being freed if the match failed.
3186    
3187     4. Tidied pcretest to ensure it frees memory that it gets.
3188    
3189     5. Temporary memory was being obtained in the case where the passed offsets
3190     vector was exactly big enough.
3191    
3192     6. Corrected definition of offsetof() from change 5 below.
3193    
3194     7. I had screwed up change 6 below and broken the rules for the use of
3195     setjmp(). Now fixed.
3196    
3197    
3198 nigel 9 Version 1.03 18-Dec-97
3199     ----------------------
3200    
3201     1. A erroneous regex with a missing opening parenthesis was correctly
3202     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3203     on some systems.
3204    
3205     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3206     it was reported that one broken compiler failed on the former because "code" is
3207     also an independent variable.
3208    
3209     3. The erroneous regex a[]b caused an array overrun reference.
3210    
3211     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3212     fail on data ending with that character. (It was going on too far, and checking
3213     the next character, typically a binary zero.) This was specific to the
3214     optimized code for single-character negative classes.
3215    
3216     5. Added a contributed patch from the TIN world which does the following:
3217    
3218     + Add an undef for memmove, in case the the system defines a macro for it.
3219    
3220     + Add a definition of offsetof(), in case there isn't one. (I don't know
3221     the reason behind this - offsetof() is part of the ANSI standard - but
3222     it does no harm).
3223    
3224     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3225     most of the places where whitespace preceded '#'. I have given up and
3226     allowed the remaining 2 cases to be at the margin.
3227    
3228     + Rename some variables in pcre to eliminate shadowing. This seems very
3229     pedantic, but does no harm, of course.
3230    
3231     6. Moved the call to setjmp() into its own function, to get rid of warnings
3232     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3233    
3234     7. Constructs such as \d{8,} were compiling into the equivalent of
3235 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3236 nigel 9 outcome, but in this particular case used more store than had been allocated,
3237     which caused the bug to be discovered because it threw up an internal error.
3238    
3239     8. The debugging code in both pcre and pcretest for outputting the compiled
3240     form of a regex was going wrong in the case of back references followed by
3241     curly-bracketed repeats.
3242    
3243    
3244 nigel 7 Version 1.02 12-Dec-97
3245     ----------------------
3246    
3247     1. Typos in pcre.3 and comments in the source fixed.
3248    
3249     2. Applied a contributed patch to get rid of places where it used to remove
3250     'const' from variables, and fixed some signed/unsigned and uninitialized
3251     variable warnings.
3252    
3253     3. Added the "runtest" target to Makefile.
3254    
3255     4. Set default compiler flag to -O2 rather than just -O.
3256    
3257    
3258 nigel 5 Version 1.01 19-Nov-97
3259     ----------------------
3260    
3261     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3262     like /([ab]*)*/, that is, for classes with more than one character in them.
3263    
3264     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3265     as /((?>a*))*/ (a PCRE_EXTRA facility).
3266    
3267    
3268     Version 1.00 18-Nov-97
3269     ----------------------
3270    
3271     1. Added compile-time macros to support systems such as SunOS4 which don't have
3272     memmove() or strerror() but have other things that can be used instead.
3273    
3274     2. Arranged that "make clean" removes the executables.
3275    
3276    
3277 nigel 3 Version 0.99 27-Oct-97
3278     ----------------------
3279    
3280     1. Fixed bug in code for optimizing classes with only one character. It was
3281     initializing a 32-byte map regardless, which could cause it to run off the end
3282     of the memory it had got.
3283    
3284     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3285    
3286    
3287     Version 0.98 22-Oct-97
3288     ----------------------
3289    
3290     1. Fixed bug in code for handling temporary memory usage when there are more
3291     back references than supplied space in the ovector. This could cause segfaults.
3292    
3293    
3294     Version 0.97 21-Oct-97
3295     ----------------------
3296    
3297     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3298    
3299     2. Optimized negated single characters not to use a bit map.
3300    
3301     3. Brought error texts together as macro definitions; clarified some of them;
3302     fixed one that was wrong - it said "range out of order" when it meant "invalid
3303     escape sequence".
3304    
3305     4. Changed some char * arguments to const char *.
3306    
3307     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3308    
3309     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3310     pcretest.
3311    
3312    
3313     Version 0.96 16-Oct-97
3314     ----------------------
3315    
3316     1. Added a simple "pgrep" utility to the distribution.
3317    
3318     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3319     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3320     where "ddd" means "one or more decimal digits".
3321    
3322     3. Fixed serious bug. If a pattern had a back reference, but the call to
3323     pcre_exec() didn't supply a large enough ovector to record the related
3324     identifying subpattern, the match always failed. PCRE now remembers the number
3325     of the largest back reference, and gets some temporary memory in which to save
3326     the offsets during matching if necessary, in order to ensure that
3327     backreferences always work.
3328    
3329     4. Increased the compatibility with Perl in a number of ways:
3330    
3331     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3332     to request this handling. The option can be set at compile or exec time.
3333    
3334     (b) $ matches before a terminating newline by default; an option
3335     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3336     mode). The option can be set at compile or exec time.
3337    
3338     (c) The handling of \ followed by a digit other than 0 is now supposed to be
3339     the same as Perl's. If the decimal number it represents is less than 10
3340     or there aren't that many previous left capturing parentheses, an octal
3341     escape is read. Inside a character class, it's always an octal escape,
3342     even if it is a single digit.
3343    
3344     (d) An escaped but undefined alphabetic character is taken as a literal,
3345     unless PCRE_EXTRA is set. Currently this just reserves the remaining
3346     escapes.
3347    
3348     (e) {0} is now permitted. (The previous item is removed from the compiled
3349     pattern).
3350    
3351     5. Changed all the names of code files so that the basic parts are no longer
3352     than 10 characters, and abolished the teeny "globals.c" file.
3353    
3354     6. Changed the handling of character classes; they are now done with a 32-byte
3355     bit map always.
3356    
3357     7. Added the -d and /D options to pcretest to make it possible to look at the
3358     internals of compilation without having to recompile pcre.
3359    
3360    
3361     Version 0.95 23-Sep-97
3362     ----------------------
3363    
3364     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3365     \x20 at the start of a run of normal characters. These were being treated as
3366     real characters, instead of the source characters being re-checked.
3367    
3368    
3369     Version 0.94 18-Sep-97
3370     ----------------------
3371    
3372     1. The functions are now thread-safe, with the caveat that the global variables
3373     containing pointers to malloc() and free() or alternative functions are the
3374     same for all threads.
3375    
3376     2. Get pcre_study() to generate a bitmap of initial characters for non-
3377     anchored patterns when this is possible, and use it if passed to pcre_exec().
3378    
3379    
3380     Version 0.93 15-Sep-97
3381     ----------------------
3382    
3383     1. /(b)|(:+)/ was computing an incorrect first character.
3384    
3385     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3386     but not actually doing anything yet.
3387    
3388     3. Treat "-" characters in classes that cannot be part of ranges as literals,
3389     as Perl does (e.g. [-az] or [az-]).
3390    
3391     4. Set the anchored flag if a branch starts with .* or .*? because that tests
3392     all possible positions.
3393    
3394     5. Split up into different modules to avoid including unneeded functions in a
3395     compiled binary. However, compile and exec are still in one module. The "study"
3396     function is split off.
3397    
3398     6. The character tables are now in a separate module whose source is generated
3399     by an auxiliary program - but can then be edited by hand if required. There are
3400     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3401     toupper() in the code.
3402    
3403     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3404     make them global. Abolish the function for setting them, as the caller can now
3405     set them directly.
3406    
3407    
3408     Version 0.92 11-Sep-97
3409     ----------------------
3410    
3411     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3412     (e.g. /a{1,3}/) was broken (I mis-optimized it).
3413    
3414     2. Caseless matching was not working in character classes if the characters in
3415     the pattern were in upper case.
3416    
3417     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3418    
3419     4. Make PCRE_ANCHORED public and accept as a compile option.
3420    
3421     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3422     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3423     pass them.
3424    
3425     6. Give an error if bad option bits passed at compile or run time.
3426    
3427     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3428     pcretest to cause it to pass that flag.
3429    
3430     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3431     options, and the first character, if set.
3432    
3433     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3434    
3435    
3436     Version 0.91 10-Sep-97
3437     ----------------------
3438    
3439     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3440     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3441    
3442     2. PCRE was looping on encountering an indefinitely repeated back reference to
3443     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3444     Perl does - treats the match as successful.
3445    
3446     ****

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