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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 354 - (hide annotations) (download)
Mon Jul 7 16:30:33 2008 UTC (6 years, 3 months ago) by ph10
File size: 157108 byte(s)
Fix caseless backreferences for non-ASCII characters.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 349 Version 8.0 02 Jul-08
5     ---------------------
6    
7     1. Replaced UCP searching code with optimized version as implemented for Ad
8     Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
9 ph10 351 stage table and inline lookup instead of a function, giving speed ups of 2
10     to 5 times on some simple patterns that I tested. Permission was given to
11     distribute the MultiStage2.py script that generates the tables (it's not in
12     the tarball, but is in the Subversion repository).
13 ph10 349
14 ph10 351 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
15     scripts.
16 ph10 353
17     3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
18     a group with a zero qualifier. The result of the study could be incorrect,
19     or the function might crash, depending on the pattern.
20 ph10 354
21     4. Caseless matching was not working for non-ASCII characters in back
22     references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
23     It now works when Unicode Property Support is available.
24 ph10 349
25 ph10 351
26 ph10 347 Version 7.7 07-May-08
27 ph10 321 ---------------------
28    
29     1. Applied Craig's patch to sort out a long long problem: "If we can't convert
30 ph10 345 a string to a long long, pretend we don't even have a long long." This is
31 ph10 321 done by checking for the strtoq, strtoll, and _strtoi64 functions.
32 ph10 345
33 ph10 322 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
34     pre-7.6 versions, which defined a global no_arg variable instead of putting
35 ph10 332 it in the RE class. (See also #8 below.)
36 ph10 345
37     3. Remove a line of dead code, identified by coverity and reported by Nuno
38     Lopes.
39    
40 ph10 324 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
41 ph10 321
42 ph10 324 (1) The include/exclude patterns were being applied to the whole pathnames
43 ph10 345 of files, instead of just to the final components.
44    
45 ph10 324 (2) If there was more than one level of directory, the subdirectories were
46     skipped unless they satisfied the include/exclude conditions. This is
47     inconsistent with GNU grep (and could even be seen as contrary to the
48     pcregrep specification - which I improved to make it absolutely clear).
49     The action now is always to scan all levels of directory, and just
50     apply the include/exclude patterns to regular files.
51 ph10 345
52 ph10 325 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
53 ph10 345 --exclude_dir in the tests to avoid scanning .svn directories.
54    
55 ph10 326 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
56 ph10 345 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
57 ph10 326 doesn't support NULs in patterns.
58 ph10 345
59     7. Added some missing "const"s to declarations of static tables in
60     pcre_compile.c and pcre_dfa_exec.c.
61    
62 ph10 328 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
63 ph10 329 caused by fix #2 above. (Subsequently also a second patch to fix the
64 ph10 345 first patch. And a third patch - this was a messy problem.)
65 ph10 321
66 ph10 345 9. Applied Craig's patch to remove the use of push_back().
67    
68 ph10 332 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
69     matching function regexec().
70 ph10 345
71     11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
72     which, however, unlike Perl's \g{...}, are subroutine calls, not back
73 ph10 333 references. PCRE supports relative numbers with this syntax (I don't think
74     Oniguruma does).
75 ph10 345
76     12. Previously, a group with a zero repeat such as (...){0} was completely
77 ph10 335 omitted from the compiled regex. However, this means that if the group
78     was called as a subroutine from elsewhere in the pattern, things went wrong
79 ph10 345 (an internal error was given). Such groups are now left in the compiled
80     pattern, with a new opcode that causes them to be skipped at execution
81 ph10 335 time.
82 ph10 345
83 ph10 341 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
84     to the way PCRE behaves:
85 ph10 345
86 ph10 336 (a) A lone ] character is dis-allowed (Perl treats it as data).
87 ph10 345
88     (b) A back reference to an unmatched subpattern matches an empty string
89 ph10 336 (Perl fails the current match path).
90 ph10 345
91 ph10 341 (c) A data ] in a character class must be notated as \] because if the
92 ph10 345 first data character in a class is ], it defines an empty class. (In
93 ph10 341 Perl it is not possible to have an empty class.) The empty class []
94     never matches; it forces failure and is equivalent to (*FAIL) or (?!).
95     The negative empty class [^] matches any one character, independently
96     of the DOTALL setting.
97 ph10 345
98     14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
99 ph10 340 non-existent subpattern following a character class starting with ']' and
100     containing () gave an internal compiling error instead of "reference to
101     non-existent subpattern". Fortunately, when the pattern did exist, the
102 ph10 345 compiled code was correct. (When scanning forwards to check for the
103 ph10 340 existencd of the subpattern, it was treating the data ']' as terminating
104 ph10 345 the class, so got the count wrong. When actually compiling, the reference
105 ph10 340 was subsequently set up correctly.)
106 ph10 345
107 ph10 341 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
108 ph10 345 it was being rejected as not supported by pcre_dfa_exec(), even though
109     other assertions are supported. I have made pcre_dfa_exec() support
110     (*FAIL).
111    
112 ph10 342 16. The implementation of 13c above involved the invention of a new opcode,
113 ph10 345 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
114     cannot be changed at match time, I realized I could make a small
115     improvement to matching performance by compiling OP_ALLANY instead of
116     OP_ANY for "." when DOTALL was set, and then removing the runtime tests
117     on the OP_ANY path.
118    
119     17. Compiling pcretest on Windows with readline support failed without the
120 ph10 343 following two fixes: (1) Make the unistd.h include conditional on
121 ph10 345 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
122    
123 ph10 344 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
124 ph10 345 ncurses library to be included for pcretest when ReadLine support is
125     requested, but also to allow for it to be overridden. This patch came from
126     Daniel Bergström.
127 ph10 347
128 ph10 346 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
129 ph10 347 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
130     any errors with the current Unicode tables. Thanks to Peter Kankowski for
131     spotting this.
132 ph10 324
133 ph10 332
134 ph10 319 Version 7.6 28-Jan-08
135 ph10 300 ---------------------
136    
137 ph10 302 1. A character class containing a very large number of characters with
138     codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
139     overflow.
140 ph10 309
141     2. Patch to cut out the "long long" test in pcrecpp_unittest when
142     HAVE_LONG_LONG is not defined.
143    
144 ph10 303 3. Applied Christian Ehrlicher's patch to update the CMake build files to
145 ph10 304 bring them up to date and include new features. This patch includes:
146 ph10 309
147 ph10 304 - Fixed PH's badly added libz and libbz2 support.
148     - Fixed a problem with static linking.
149 ph10 312 - Added pcredemo. [But later removed - see 7 below.]
150 ph10 304 - Fixed dftables problem and added an option.
151     - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
152     HAVE_LONG_LONG.
153     - Added readline support for pcretest.
154 ph10 309 - Added an listing of the option settings after cmake has run.
155    
156 ph10 314 4. A user submitted a patch to Makefile that makes it easy to create
157     "pcre.dll" under mingw when using Configure/Make. I added stuff to
158     Makefile.am that cause it to include this special target, without
159     affecting anything else. Note that the same mingw target plus all
160     the other distribution libraries and programs are now supported
161     when configuring with CMake (see 6 below) instead of with
162     Configure/Make.
163 ph10 309
164 ph10 308 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
165     This is an attempt to solve the reported problem "pcrecpp::no_arg is not
166 ph10 309 exported in the Windows port". It has not yet been confirmed that the patch
167     solves the problem, but it does no harm.
168 ph10 313
169 ph10 311 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
170 ph10 319 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
171 ph10 314 with CMake, and also correct the comment about stack recursion.
172 ph10 313
173 ph10 312 7. Remove the automatic building of pcredemo from the ./configure system and
174 ph10 313 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
175     of a program that users should build themselves after PCRE is installed, so
176     building it automatically is not really right. What is more, it gave
177 ph10 312 trouble in some build environments.
178 ph10 300
179 ph10 319 8. Further tidies to CMakeLists.txt from Sheri and Christian.
180 ph10 308
181 ph10 319
182 ph10 298 Version 7.5 10-Jan-08
183 ph10 263 ---------------------
184    
185     1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
186     values in parens when parsing an RE using the C++ wrapper."
187 ph10 286
188 ph10 264 2. Negative specials like \S did not work in character classes in UTF-8 mode.
189     Characters greater than 255 were excluded from the class instead of being
190     included.
191 ph10 286
192     3. The same bug as (2) above applied to negated POSIX classes such as
193 ph10 265 [:^space:].
194 ph10 286
195 ph10 267 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
196     defined or documented. It seems to have been a typo for PCRE_STATIC, so
197 ph10 286 I have changed it.
198    
199 ph10 268 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
200 ph10 286 first named subpattern) and a construct such as (?&a) would reference the
201     first named subpattern whose name started with "a" (in other words, the
202 ph10 272 length check was missing). Both these problems are fixed. "Subpattern name
203     expected" is now given for (?&) (a zero-length name), and this patch also
204     makes it give the same error for \k'' (previously it complained that that
205     was a reference to a non-existent subpattern).
206 ph10 286
207 ph10 269 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
208 ph10 286 this is right because (?- can be followed by option settings as well as by
209 ph10 269 digits. I have, however, made the messages clearer.
210 ph10 286
211 ph10 270 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
212     than the number used in the conditional) now cause a compile-time error.
213     This is actually not compatible with Perl, which accepts such patterns, but
214     treats the conditional as always being FALSE (as PCRE used to), but it
215     seems to me that giving a diagnostic is better.
216 ph10 286
217 ph10 274 8. Change "alphameric" to the more common word "alphanumeric" in comments
218 ph10 275 and messages.
219 ph10 286
220     9. Fix two occurrences of "backslash" in comments that should have been
221     "backspace".
222    
223     10. Remove two redundant lines of code that can never be obeyed (their function
224     was moved elsewhere).
225    
226     11. The program that makes PCRE's Unicode character property table had a bug
227     which caused it to generate incorrect table entries for sequences of
228 ph10 277 characters that have the same character type, but are in different scripts.
229 ph10 286 It amalgamated them into a single range, with the script of the first of
230 ph10 277 them. In other words, some characters were in the wrong script. There were
231     thirteen such cases, affecting characters in the following ranges:
232 ph10 286
233 ph10 277 U+002b0 - U+002c1
234     U+0060c - U+0060d
235 ph10 286 U+0061e - U+00612
236 ph10 277 U+0064b - U+0065e
237     U+0074d - U+0076d
238     U+01800 - U+01805
239     U+01d00 - U+01d77
240     U+01d9b - U+01dbf
241     U+0200b - U+0200f
242     U+030fc - U+030fe
243     U+03260 - U+0327f
244     U+0fb46 - U+0fbb1
245     U+10450 - U+1049d
246 ph10 286
247 ph10 279 12. The -o option (show only the matching part of a line) for pcregrep was not
248     compatible with GNU grep in that, if there was more than one match in a
249     line, it showed only the first of them. It now behaves in the same way as
250     GNU grep.
251 ph10 286
252     13. If the -o and -v options were combined for pcregrep, it printed a blank
253     line for every non-matching line. GNU grep prints nothing, and pcregrep now
254     does the same. The return code can be used to tell if there were any
255     non-matching lines.
256    
257 ph10 289 14. Added --file-offsets and --line-offsets to pcregrep.
258    
259     15. The pattern (?=something)(?R) was not being diagnosed as a potentially
260 ph10 282 infinitely looping recursion. The bug was that positive lookaheads were not
261 ph10 286 being skipped when checking for a possible empty match (negative lookaheads
262     and both kinds of lookbehind were skipped).
263    
264 ph10 289 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
265 ph10 284 inclusion of <windows.h> to before rather than after the definition of
266     INVALID_FILE_ATTRIBUTES (patch from David Byron).
267 ph10 263
268 ph10 289 17. Specifying a possessive quantifier with a specific limit for a Unicode
269 ph10 286 character property caused pcre_compile() to compile bad code, which led at
270     runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
271     are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
272     caused the error; without that there was no problem.
273 ph10 263
274 ph10 289 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
275 ph10 286
276 ph10 289 19. Added --enable-pcretest-libreadline.
277 ph10 286
278 ph10 289 20. In pcrecpp.cc, the variable 'count' was incremented twice in
279 ph10 288 RE::GlobalReplace(). As a result, the number of replacements returned was
280 ph10 298 double what it should be. I removed one of the increments, but Craig sent a
281     later patch that removed the other one (the right fix) and added unit tests
282     that check the return values (which was not done before).
283 ph10 292
284 ph10 291 21. Several CMake things:
285 ph10 287
286 ph10 291 (1) Arranged that, when cmake is used on Unix, the libraries end up with
287 ph10 292 the names libpcre and libpcreposix, not just pcre and pcreposix.
288 ph10 288
289 ph10 292 (2) The above change means that pcretest and pcregrep are now correctly
290     linked with the newly-built libraries, not previously installed ones.
291 ph10 291
292 ph10 292 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
293 ph10 298
294     22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
295     crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
296     UTF-8 newline character). The key issue is that the pattern starts .*;
297     this means that the match must be either at the beginning, or after a
298     newline. The bug was in the code for advancing after a failed match and
299     checking that the new position followed a newline. It was not taking
300 ph10 294 account of UTF-8 characters correctly.
301 ph10 298
302     23. PCRE was behaving differently from Perl in the way it recognized POSIX
303     character classes. PCRE was not treating the sequence [:...:] as a
304     character class unless the ... were all letters. Perl, however, seems to
305     allow any characters between [: and :], though of course it rejects as
306     unknown any "names" that contain non-letters, because all the known class
307     names consist only of letters. Thus, Perl gives an error for [[:1234:]],
308     for example, whereas PCRE did not - it did not recognize a POSIX character
309     class. This seemed a bit dangerous, so the code has been changed to be
310     closer to Perl. The behaviour is not identical to Perl, because PCRE will
311     diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
312     treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
313     Perl does, and where it didn't before.
314    
315 ph10 296 24. Rewrite so as to remove the single use of %n from pcregrep because in some
316     Windows environments %n is disabled by default.
317 ph10 292
318    
319 ph10 260 Version 7.4 21-Sep-07
320 ph10 230 ---------------------
321    
322     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
323 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
324     LF". That isn't really right - the whole point of the change was to try to
325     help when there was an actual mention of one of the two characters. So now
326     the change happens only if \r or \n (or a literal CR or LF) character is
327 ph10 230 encountered.
328 ph10 231
329     2. The 32-bit options word was also used for 6 internal flags, but the numbers
330     of both had grown to the point where there were only 3 bits left.
331     Fortunately, there was spare space in the data structure, and so I have
332     moved the internal flags into a new 16-bit field to free up more option
333 ph10 230 bits.
334 ph10 231
335     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
336     but did not set the internal JCHANGED flag - either of these is enough to
337     control the way the "get" function works - but the PCRE_INFO_JCHANGED
338     facility is supposed to tell if (?J) was ever used, so now (?J) at the
339 ph10 230 start sets both bits.
340    
341 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
342     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
343 ph10 230
344 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
345    
346     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
347 ph10 233 compatibility, even though it is no longer used.
348 ph10 243
349     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
350 ph10 254 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
351 ph10 259 windows.h file is present (where different names are used). [This was
352 ph10 257 reversed later after testing - see 16 below.]
353 ph10 243
354     8. Changed all #include <config.h> to #include "config.h". There were also
355 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
356 ph10 243
357 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
358 ph10 243 sequence off the lines that it output.
359    
360     10. It was pointed out to me that arrays of string pointers cause lots of
361     relocations when a shared library is dynamically loaded. A technique of
362     using a single long string with a table of offsets can drastically reduce
363     these. I have refactored PCRE in four places to do this. The result is
364 ph10 240 dramatic:
365 ph10 243
366 ph10 240 Originally: 290
367     After changing UCP table: 187
368 ph10 243 After changing error message table: 43
369 ph10 240 After changing table of "verbs" 36
370     After changing table of Posix names 22
371 ph10 243
372 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
373 ph10 247
374 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
375 ph10 247 unicode-properties was also set.
376    
377 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
378 ph10 231
379 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
380     checked only for CRLF.
381 ph10 233
382 ph10 259 14. Added casts to pcretest.c to avoid compiler warnings.
383    
384 ph10 256 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
385 ph10 253
386 ph10 257 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
387 ph10 259 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
388     entirely. This removes changes made in 7 above.
389 ph10 256
390 ph10 261 17. The CMake files have been updated, and there is now more information about
391     building with CMake in the NON-UNIX-USE document.
392 ph10 257
393 ph10 261
394 ph10 228 Version 7.3 28-Aug-07
395 ph10 157 ---------------------
396    
397 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
398     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
399     brackets there is not right, since it causes compilers to look for an
400     installed pcre.h, not the version that is in the source that is being
401     compiled (which of course may be different). I have changed it back to:
402 ph10 197
403 ph10 189 #include "pcre.h"
404 ph10 197
405     I have a vague recollection that the change was concerned with compiling in
406     different directories, but in the new build system, that is taken care of
407     by the VPATH setting the Makefile.
408    
409 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
410     when the subject happened to end in the byte 0x85 (e.g. if the last
411     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
412     characters but of course it shouldn't be taken as a newline when it is part
413     of another character. The bug was that, for an unlimited repeat of . in
414     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
415     characters when looking for a newline.
416 ph10 197
417     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
418    
419     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
420     in debug output.
421    
422 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
423     long printing in the pcrecpp unittest when running under MinGW.
424 ph10 189
425 ph10 197 6. ESC_K was left out of the EBCDIC table.
426 ph10 189
427 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
428     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
429     limit also applies to "virtual nesting" when a pattern is recursive, and in
430     this case 1000 isn't so big. I have been able to remove this limit at the
431     expense of backing off one optimization in certain circumstances. Normally,
432     when pcre_exec() would call its internal match() function recursively and
433     immediately return the result unconditionally, it uses a "tail recursion"
434     feature to save stack. However, when a subpattern that can match an empty
435     string has an unlimited repetition quantifier, it no longer makes this
436     optimization. That gives it a stack frame in which to save the data for
437     checking that an empty string has been matched. Previously this was taken
438     from the 1000-entry workspace that had been reserved. So now there is no
439     explicit limit, but more stack is used.
440 ph10 189
441 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
442     syntax that is required for Windows, and which was going wrong for the
443     pcreposix and pcrecpp parts of the library. These were overlooked when this
444     problem was solved for the main library.
445 ph10 197
446 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
447     the size of patterns that contain repeated groups with explicit upper
448     limits. As the maximum quantifier is 65535, the maximum group length was
449     set at 30,000 so that the product of these two numbers did not overflow a
450     32-bit integer. However, it turns out that people want to use groups that
451     are longer than 30,000 bytes (though not repeat them that many times).
452     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
453     made it possible to implement the integer overflow checks in a much more
454     dynamic way, which I have now done. The artificial limitation on group
455     length has been removed - we now have only the limit on the total length of
456     the compiled pattern, which depends on the LINK_SIZE setting.
457 ph10 208
458     10. Fixed a bug in the documentation for get/copy named substring when
459     duplicate names are permitted. If none of the named substrings are set, the
460     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
461     empty string.
462    
463     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
464     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
465     because the ] is interpreted as the first data character and the
466     terminating ] is not found. PCRE has been made compatible with Perl in this
467     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
468     cause memory overwriting.
469    
470 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
471     string has been matched (to stop an infinite loop). It was not recognizing
472 ph10 208 a conditional subpattern that could match an empty string if that
473 ph10 206 subpattern was within another subpattern. For example, it looped when
474 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
475 ph10 206 condition was not nested. This bug has been fixed.
476 ph10 208
477 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
478     past the start of the subject in the presence of bytes with the top bit
479     set, for example "\x8aBCD".
480 ph10 211
481 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
482     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
483 ph10 200
484 ph10 211 14. Optimized (?!) to (*FAIL).
485 ph10 202
486 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
487     This restricts code points to be within the range 0 to 0x10FFFF, excluding
488     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
489     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
490     does: it's just the validity check that is more restrictive.
491 ph10 220
492     16. Inserted checks for integer overflows during escape sequence (backslash)
493     processing, and also fixed erroneous offset values for syntax errors during
494     backslash processing.
495    
496 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
497 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
498    
499 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
500     caused an overrun.
501 ph10 220
502     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
503     something other than just ASCII characters) inside a group that had an
504     unlimited repeat caused a loop at compile time (while checking to see
505     whether the group could match an empty string).
506    
507 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
508     [\P{Any}] did so. (Error in the code for printing property names.)
509 ph10 210
510 ph10 220 21. An orphan \E inside a character class could cause a crash.
511    
512     22. A repeated capturing bracket such as (A)? could cause a wild memory
513 ph10 218 reference during compilation.
514 ph10 220
515     23. There are several functions in pcre_compile() that scan along a compiled
516     expression for various reasons (e.g. to see if it's fixed length for look
517 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
518 ph10 220 present in the pattern. These operators have additional parameters compared
519     with \d, etc, and these were not being taken into account when moving along
520 ph10 218 the compiled data. Specifically:
521 ph10 220
522     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
523     length.
524    
525     (b) An item such as \pL+ within a repeated group could cause crashes or
526 ph10 218 loops.
527 ph10 220
528 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
529 ph10 220 "reference to non-existent subpattern" error.
530    
531 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
532    
533 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
534 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
535 ph10 211
536 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
537 ph10 218
538 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
539 ph10 223 character were causing crashes (broken optimization).
540 ph10 227
541 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
542 ph10 227 \p or \P) caused a compile-time loop.
543    
544 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
545     break. For example, the unstudied pattern [\r\n]A does not match the string
546     "\r\nA" because change 7.0/46 below moves the current point on by two
547     characters after failing to match at the start. However, the pattern \nA
548 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
549     the same is true. There doesn't seem any very clean way out of this, but
550     what I have chosen to do makes the common cases work: PCRE now takes note
551     of whether there can be an explicit match for \r or \n anywhere in the
552     pattern, and if so, 7.0/46 no longer applies. As part of this change,
553     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
554     pattern has explicit CR or LF references.
555 ph10 222
556 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
557 ph10 223
558 ph10 227
559 ph10 189 Version 7.2 19-Jun-07
560     ---------------------
561    
562 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
563     which is apparently normally available under Windows.
564 ph10 159
565     2. Re-jig the pcregrep tests with different newline settings in an attempt
566     to make them independent of the local environment's newline setting.
567 ph10 157
568 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
569 ph10 165
570 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
571     was not 2, because the output contained actual offsets. The recent new
572 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
573     usable with all link sizes.
574    
575 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
576     stack recursion. This gives a massive performance boost under BSD, but just
577 ph10 165 a small improvement under Linux. However, it saves one field in the frame
578 ph10 164 in all cases.
579 ph10 172
580 ph10 166 6. Added more features from the forthcoming Perl 5.10:
581 ph10 172
582 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
583     recursion call. It refers to the nth most recently opened parentheses.
584 ph10 172
585 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
586 ph10 172 to be opened parentheses.
587    
588     (c) Conditions that refer to capturing parentheses can be specified
589 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
590 ph10 172
591 ph10 168 (d) \K resets the start of the current match so that everything before
592 ph10 172 is not part of it.
593    
594 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
595 ph10 172
596 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
597 ph10 172 reference syntax.
598 ph10 182
599 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
600 ph10 182 alternative starts with the same number.
601 ph10 172
602 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
603    
604 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
605     PCRE_INFO_JCHANGED.
606    
607     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
608     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
609 ph10 170 for detecting groups that can match an empty string.
610 ph10 159
611 ph10 172 9. A pattern with a very large number of alternatives (more than several
612     hundred) was running out of internal workspace during the pre-compile
613     phase, where pcre_compile() figures out how much memory will be needed. A
614     bit of new cunning has reduced the workspace needed for groups with
615     alternatives. The 1000-alternative test pattern now uses 12 bytes of
616     workspace instead of running out of the 4096 that are available.
617 ph10 182
618 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
619 ph10 172
620 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
621     The report of the bug said:
622 ph10 182
623 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
624     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
625 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
626 ph10 185
627 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
628 ph10 185 it matched the wrong number of bytes.
629 ph10 172
630 ph10 179
631 ph10 155 Version 7.1 24-Apr-07
632 ph10 98 ---------------------
633    
634 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
635 ph10 122 that is more "standard", making use of automake and other Autotools. There
636 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
637     on this.
638 ph10 111
639     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
640     for recursive directory scanning broke on some systems because the files
641     are not scanned in any specific order and on different systems the order
642     was different. A call to "sort" has been inserted into RunGrepTest for the
643     approprate test as a short-term fix. In the longer term there may be an
644 ph10 100 alternative.
645 ph10 111
646 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
647 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
648     people do their own conversions for various reasons). The problems
649     concerned the use of low-level troff macros .br and .in. I have therefore
650     removed all such uses from the man pages (some were redundant, some could
651 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
652     HTML has been updated to handle .nf/.fi and to complain if it encounters
653     .br or .in.
654 ph10 111
655 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
656 ph10 123 arranged for config.h to be included in the distribution, with the name
657 ph10 111 config.h.generic, for the benefit of those who have to compile without
658     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
659    
660     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
661     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
662 ph10 127 makevp.bat for the current PCRE, using the additional files
663 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
664 ph10 111
665     6. A Windows user reported a minor discrepancy with test 2, which turned out
666     to be caused by a trailing space on an input line that had got lost in his
667 ph10 102 copy. The trailing space was an accident, so I've just removed it.
668 ph10 111
669 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
670 ph10 111 that is needed.
671    
672 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
673     as "const" (a) because they are and (b) because it helps the PHP
674     maintainers who have recently made a script to detect big data structures
675 ph10 111 in the php code that should be moved to the .rodata section. I remembered
676     to update Builducptable as well, so it won't revert if ucptable.h is ever
677 ph10 105 re-created.
678 ph10 111
679     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
680 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
681 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
682     support is not required. This saves 1.5-2K of code, which is important in
683 ph10 107 some applications.
684 ph10 111
685 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
686 ph10 111 so as not to refer to the tables, even though these functions will never be
687     called when UTF-8 support is disabled. Otherwise there are problems with a
688     shared library.
689    
690 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
691    
692     (a) It was defining its arguments as char * instead of void *.
693    
694 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
695     a long time ago when I wrote it, but is no longer the case.
696    
697 ph10 118 The emulated memove() is provided for those environments that have neither
698 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
699 ph10 118 is clearly not the case, as these two bugs were recently reported.
700 ph10 123
701 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
702 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
703     pages, and it removes trailing spaces from listed files. It also creates
704     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
705     case, it wraps all the #defines with #ifndefs. This script should be run
706 ph10 111 before "make dist".
707 ph10 123
708 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
709     with Unicode property support.
710 ph10 123
711     (a) For a maximizing quantifier, if the two different cases of the
712     character were of different lengths in their UTF-8 codings (there are
713     some cases like this - I found 11), and the matching function had to
714 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
715     were both the same length.
716 ph10 123
717     (b) When PCRE was configured to use the heap rather than the stack for
718     recursion during matching, it was not correctly preserving the data for
719     the other case of a UTF-8 character when checking ahead for a match
720     while processing a minimizing repeat. If the check also involved
721 ph10 115 matching a wide character, but failed, corruption could cause an
722     erroneous result when trying to check for a repeat of the original
723     character.
724 ph10 123
725 ph10 116 13. Some tidying changes to the testing mechanism:
726 ph10 98
727 ph10 116 (a) The RunTest script now detects the internal link size and whether there
728     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
729 ph10 123 values substituted by "configure". (The RunGrepTest script already did
730     this for UTF-8.) The configure.ac script no longer substitutes the
731     relevant variables.
732    
733 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
734     with length and offset values. This means that the output is different
735     for different internal link sizes. Test 2 is skipped for link sizes
736     other than 2 because of this, bypassing the problem. Unfortunately,
737     there was also a test in test 3 (the locale tests) that used /B and
738 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
739     I have added a new /Z option to pcretest that replaces the length and
740     offset values with spaces. This is now used to make test 3 independent
741 ph10 122 of link size. (Test 2 will be tidied up later.)
742 ph10 123
743     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
744 ph10 122 segmentation fault instead of returning the appropriate error message.
745 ph10 134
746 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
747 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
748     This doesn't seem right; it now treats the CRLF combination as the line
749     ending, and so does not match in that case. It's only a pattern such as ^$
750     that would hit this one: something like ^ABC$ would have failed after \r
751     and then tried again after \r\n.
752    
753 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
754 ph10 134 in an attempt to make files that differ only in their line terminators
755     compare equal. This works on Linux.
756 ph10 142
757 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
758     as it exited. This is now fixed, thanks to valgrind.
759 ph10 142
760 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
761 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
762     was because its rules for how to advance for /g after matching an empty
763 ph10 143 string at the end of a line did not allow for this case. They now check for
764     it specially.
765 ph10 150
766     20. pcretest is supposed to handle patterns and data of any length, by
767     extending its buffers when necessary. It was getting this wrong when the
768 ph10 147 buffer for a data line had to be extended.
769 ph10 150
770 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
771 ph10 150 CRLF as a newline sequence.
772 ph10 152
773 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
774 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
775     I have nevertheless tidied it up.
776 ph10 111
777 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
778 ph10 123
779 ph10 153 24. Added a man page for pcre-config.
780 ph10 152
781 ph10 153
782 nigel 93 Version 7.0 19-Dec-06
783     ---------------------
784    
785     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
786     moving to gcc 4.1.1.
787    
788     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
789     sys/time.h, which is documented as needed for this function. It doesn't
790     seem to matter on Linux, but it showed up on some releases of OS X.
791    
792     3. It seems that there are systems where bytes whose values are greater than
793     127 match isprint() in the "C" locale. The "C" locale should be the
794     default when a C program starts up. In most systems, only ASCII printing
795     characters match isprint(). This difference caused the output from pcretest
796     to vary, making some of the tests fail. I have changed pcretest so that:
797    
798     (a) When it is outputting text in the compiled version of a pattern, bytes
799     other than 32-126 are always shown as hex escapes.
800    
801     (b) When it is outputting text that is a matched part of a subject string,
802     it does the same, unless a different locale has been set for the match
803     (using the /L modifier). In this case, it uses isprint() to decide.
804    
805     4. Fixed a major bug that caused incorrect computation of the amount of memory
806     required for a compiled pattern when options that changed within the
807     pattern affected the logic of the preliminary scan that determines the
808     length. The relevant options are -x, and -i in UTF-8 mode. The result was
809     that the computed length was too small. The symptoms of this bug were
810     either the PCRE error "internal error: code overflow" from pcre_compile(),
811     or a glibc crash with a message such as "pcretest: free(): invalid next
812     size (fast)". Examples of patterns that provoked this bug (shown in
813     pcretest format) are:
814    
815     /(?-x: )/x
816     /(?x)(?-x: \s*#\s*)/
817     /((?i)[\x{c0}])/8
818     /(?i:[\x{c0}])/8
819    
820     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
821     is now done differently.
822    
823     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
824     wrapper classes; (b) implement a new function in the C++ scanner that is
825     more efficient than the old way of doing things because it avoids levels of
826     recursion in the regex matching; (c) add a paragraph to the documentation
827     for the FullMatch() function.
828    
829     6. The escape sequence \n was being treated as whatever was defined as
830     "newline". Not only was this contrary to the documentation, which states
831     that \n is character 10 (hex 0A), but it also went horribly wrong when
832     "newline" was defined as CRLF. This has been fixed.
833    
834     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
835     was being set to -1 for the "end of line" case (supposedly a value that no
836     character can have). Though this value is never used (the check for end of
837     line is "zero bytes in current character"), it caused compiler complaints.
838     I've changed it to 0xffffffff.
839    
840     8. In pcre_version.c, the version string was being built by a sequence of
841     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
842     string (as it is for production releases) called a macro with an empty
843     argument. The C standard says the result of this is undefined. The gcc
844     compiler treats it as an empty string (which was what was wanted) but it is
845     reported that Visual C gives an error. The source has been hacked around to
846     avoid this problem.
847    
848     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
849     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
850     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
851     of them did).
852    
853     10. Originally, pcretest opened its input and output without "b"; then I was
854     told that "b" was needed in some environments, so it was added for release
855     5.0 to both the input and output. (It makes no difference on Unix-like
856     systems.) Later I was told that it is wrong for the input on Windows. I've
857     now abstracted the modes into two macros, to make it easier to fiddle with
858     them, and removed "b" from the input mode under Windows.
859    
860     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
861    
862     12. Added -help and --help to pcretest as an official way of being reminded
863     of the options.
864    
865     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
866     and pcrecpp.cc because they annoy compilers at high warning levels.
867    
868     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
869    
870     15. Fixed an occurrence of == in configure.ac that should have been = (shell
871     scripts are not C programs :-) and which was not noticed because it works
872     on Linux.
873    
874     16. pcretest is supposed to handle any length of pattern and data line (as one
875     line or as a continued sequence of lines) by extending its input buffer if
876     necessary. This feature was broken for very long pattern lines, leading to
877     a string of junk being passed to pcre_compile() if the pattern was longer
878     than about 50K.
879    
880     17. I have done a major re-factoring of the way pcre_compile() computes the
881     amount of memory needed for a compiled pattern. Previously, there was code
882     that made a preliminary scan of the pattern in order to do this. That was
883     OK when PCRE was new, but as the facilities have expanded, it has become
884     harder and harder to keep it in step with the real compile phase, and there
885     have been a number of bugs (see for example, 4 above). I have now found a
886     cunning way of running the real compile function in a "fake" mode that
887     enables it to compute how much memory it would need, while actually only
888     ever using a few hundred bytes of working memory and without too many
889     tests of the mode. This should make future maintenance and development
890     easier. A side effect of this work is that the limit of 200 on the nesting
891     depth of parentheses has been removed (though this was never a serious
892     limitation, I suspect). However, there is a downside: pcre_compile() now
893     runs more slowly than before (30% or more, depending on the pattern). I
894     hope this isn't a big issue. There is no effect on runtime performance.
895    
896     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
897     newline (only possible for the last line of a file) and it was a
898     pattern that set a locale (followed by /Lsomething), pcretest crashed.
899    
900     19. Added additional timing features to pcretest. (1) The -tm option now times
901     matching only, not compiling. (2) Both -t and -tm can be followed, as a
902     separate command line item, by a number that specifies the number of
903     repeats to use when timing. The default is 50000; this gives better
904     precision, but takes uncomfortably long for very large patterns.
905    
906     20. Extended pcre_study() to be more clever in cases where a branch of a
907     subpattern has no definite first character. For example, (a*|b*)[cd] would
908     previously give no result from pcre_study(). Now it recognizes that the
909     first character must be a, b, c, or d.
910    
911     21. There was an incorrect error "recursive call could loop indefinitely" if
912     a subpattern (or the entire pattern) that was being tested for matching an
913     empty string contained only one non-empty item after a nested subpattern.
914     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
915     incorrectly, because the \d was being skipped in the check.
916    
917     22. The pcretest program now has a new pattern option /B and a command line
918     option -b, which is equivalent to adding /B to every pattern. This causes
919     it to show the compiled bytecode, without the additional information that
920     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
921     is the same as /B/I).
922    
923     23. A new optimization is now able automatically to treat some sequences such
924     as a*b as a*+b. More specifically, if something simple (such as a character
925     or a simple class like \d) has an unlimited quantifier, and is followed by
926     something that cannot possibly match the quantified thing, the quantifier
927     is automatically "possessified".
928    
929     24. A recursive reference to a subpattern whose number was greater than 39
930     went wrong under certain circumstances in UTF-8 mode. This bug could also
931     have affected the operation of pcre_study().
932    
933     25. Realized that a little bit of performance could be had by replacing
934     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
935    
936     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
937    
938     27. Possessive quantifiers such as a++ were previously implemented by turning
939     them into atomic groups such as ($>a+). Now they have their own opcodes,
940     which improves performance. This includes the automatically created ones
941     from 23 above.
942    
943     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
944     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
945     the first matched character to be a colon. This applied both to named and
946     numbered groups.
947    
948     29. The ucpinternal.h header file was missing its idempotency #ifdef.
949    
950     30. I was sent a "project" file called libpcre.a.dev which I understand makes
951     building PCRE on Windows easier, so I have included it in the distribution.
952    
953     31. There is now a check in pcretest against a ridiculously large number being
954     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
955     loop, the loop is abandoned.
956    
957     32. Forward references to subpatterns in conditions such as (?(2)...) where
958     subpattern 2 is defined later cause pcre_compile() to search forwards in
959     the pattern for the relevant set of parentheses. This search went wrong
960     when there were unescaped parentheses in a character class, parentheses
961     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
962    
963     33. "Subroutine" calls and backreferences were previously restricted to
964     referencing subpatterns earlier in the regex. This restriction has now
965     been removed.
966    
967     34. Added a number of extra features that are going to be in Perl 5.10. On the
968     whole, these are just syntactic alternatives for features that PCRE had
969     previously implemented using the Python syntax or my own invention. The
970     other formats are all retained for compatibility.
971    
972     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
973     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
974     also .NET compatible.
975    
976     (b) A recursion or subroutine call to a named group can now be defined as
977     (?&name) as well as (?P>name).
978    
979     (c) A backreference to a named group can now be defined as \k<name> or
980     \k'name' as well as (?P=name). The new forms, as well as being in Perl
981     5.10, are also .NET compatible.
982    
983     (d) A conditional reference to a named group can now use the syntax
984     (?(<name>) or (?('name') as well as (?(name).
985    
986     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
987     groups (named and numbered) that are never evaluated inline, but can be
988     called as "subroutines" from elsewhere. In effect, the DEFINE condition
989     is always false. There may be only one alternative in such a group.
990    
991     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
992     as the simple (?(R). The condition is true only if the most recent
993     recursion is that of the given number or name. It does not search out
994     through the entire recursion stack.
995    
996     (g) The escape \gN or \g{N} has been added, where N is a positive or
997     negative number, specifying an absolute or relative reference.
998    
999     35. Tidied to get rid of some further signed/unsigned compiler warnings and
1000     some "unreachable code" warnings.
1001    
1002     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1003     things, this adds five new scripts.
1004    
1005     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1006     There were also incompatibilities regarding the handling of \Q..\E inside
1007     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1008     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1009    
1010     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1011     matches an empty string, and forcibly breaks the loop. There were bugs in
1012     this code in non-simple cases. For a pattern such as ^(a()*)* matched
1013     against aaaa the result was just "a" rather than "aaaa", for example. Two
1014     separate and independent bugs (that affected different cases) have been
1015     fixed.
1016    
1017     39. Refactored the code to abolish the use of different opcodes for small
1018     capturing bracket numbers. This is a tidy that I avoided doing when I
1019     removed the limit on the number of capturing brackets for 3.5 back in 2001.
1020     The new approach is not only tidier, it makes it possible to reduce the
1021     memory needed to fix the previous bug (38).
1022    
1023     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1024     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1025     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1026     mode.
1027    
1028     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1029     report.
1030    
1031     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1032     copy construction and assignment in the C++ wrapper.
1033    
1034     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1035     couple of bugs that could have given wrong results in the "--newline=crlf"
1036     case.
1037    
1038     44. Added a number of casts and did some reorganization of signed/unsigned int
1039     variables following suggestions from Dair Grant. Also renamed the variable
1040     "this" as "item" because it is a C++ keyword.
1041    
1042     45. Arranged for dftables to add
1043    
1044     #include "pcre_internal.h"
1045    
1046     to pcre_chartables.c because without it, gcc 4.x may remove the array
1047     definition from the final binary if PCRE is built into a static library and
1048     dead code stripping is activated.
1049    
1050     46. For an unanchored pattern, if a match attempt fails at the start of a
1051     newline sequence, and the newline setting is CRLF or ANY, and the next two
1052     characters are CRLF, advance by two characters instead of one.
1053    
1054    
1055 nigel 91 Version 6.7 04-Jul-06
1056     ---------------------
1057    
1058     1. In order to handle tests when input lines are enormously long, pcretest has
1059     been re-factored so that it automatically extends its buffers when
1060     necessary. The code is crude, but this _is_ just a test program. The
1061     default size has been increased from 32K to 50K.
1062    
1063     2. The code in pcre_study() was using the value of the re argument before
1064     testing it for NULL. (Of course, in any sensible call of the function, it
1065     won't be NULL.)
1066    
1067     3. The memmove() emulation function in pcre_internal.h, which is used on
1068     systems that lack both memmove() and bcopy() - that is, hardly ever -
1069     was missing a "static" storage class specifier.
1070    
1071     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1072     containing an extended class (one that cannot be represented by a bitmap
1073     because it contains high-valued characters or Unicode property items, e.g.
1074     [\pZ]). Almost always one would set UTF-8 mode when processing such a
1075     pattern, but PCRE should not loop if you do not (it no longer does).
1076     [Detail: two cases were found: (a) a repeated subpattern containing an
1077     extended class; (b) a recursive reference to a subpattern that followed a
1078     previous extended class. It wasn't skipping over the extended class
1079     correctly when UTF-8 mode was not set.]
1080    
1081     5. A negated single-character class was not being recognized as fixed-length
1082     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1083     compile error "lookbehind assertion is not fixed length".
1084    
1085     6. The RunPerlTest auxiliary script was showing an unexpected difference
1086     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1087     write a Perl script that can interpret lines of an input file either as
1088     byte characters or as UTF-8, which is what "perltest" was being required to
1089     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1090     can't do is switch easily at run time between having the "use utf8;" pragma
1091     or not. In the end, I fudged it by using the RunPerlTest script to insert
1092     "use utf8;" explicitly for the UTF-8 tests.
1093    
1094     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1095     the end of the subject string, contrary to the documentation and to what
1096     Perl does. This was true of both matching functions. Now it matches only at
1097     the start of the subject and immediately after *internal* newlines.
1098    
1099     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1100     a pointer to an int instead of a pointer to an unsigned long int. This
1101     caused problems on 64-bit systems.
1102    
1103     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1104     instance of the 'standard' template library not being so standard".
1105    
1106     10. There was no check on the number of named subpatterns nor the maximum
1107     length of a subpattern name. The product of these values is used to compute
1108     the size of the memory block for a compiled pattern. By supplying a very
1109     long subpattern name and a large number of named subpatterns, the size
1110     computation could be caused to overflow. This is now prevented by limiting
1111     the length of names to 32 characters, and the number of named subpatterns
1112     to 10,000.
1113    
1114     11. Subpatterns that are repeated with specific counts have to be replicated in
1115     the compiled pattern. The size of memory for this was computed from the
1116     length of the subpattern and the repeat count. The latter is limited to
1117     65535, but there was no limit on the former, meaning that integer overflow
1118     could in principle occur. The compiled length of a repeated subpattern is
1119     now limited to 30,000 bytes in order to prevent this.
1120    
1121     12. Added the optional facility to have named substrings with the same name.
1122    
1123     13. Added the ability to use a named substring as a condition, using the
1124     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1125     are numbers (not recommended). Forward references are permitted.
1126    
1127     14. Added forward references in named backreferences (if you see what I mean).
1128    
1129     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1130     pattern could run off the end of the subject. For example, the pattern
1131     "(?s)(.{1,5})"8 did this with the subject "ab".
1132    
1133     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1134     PCRE_CASELESS was set when matching characters that were quantified with ?
1135     or *.
1136    
1137     17. A character class other than a single negated character that had a minimum
1138     but no maximum quantifier - for example [ab]{6,} - was not handled
1139     correctly by pce_dfa_exec(). It would match only one character.
1140    
1141     18. A valid (though odd) pattern that looked like a POSIX character
1142     class but used an invalid character after [ (for example [[,abc,]]) caused
1143     pcre_compile() to give the error "Failed: internal error: code overflow" or
1144     in some cases to crash with a glibc free() error. This could even happen if
1145     the pattern terminated after [[ but there just happened to be a sequence of
1146     letters, a binary zero, and a closing ] in the memory that followed.
1147    
1148     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1149     over the years. Originally (before any Unicode support), just the bottom 8
1150     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1151     output from "man perlunicode" includes this:
1152    
1153     The regular expression compiler produces polymorphic opcodes. That
1154     is, the pattern adapts to the data and automatically switches to
1155     the Unicode character scheme when presented with Unicode data--or
1156     instead uses a traditional byte scheme when presented with byte
1157     data.
1158    
1159     Sadly, a wide octal escape does not cause a switch, and in a string with
1160     no other multibyte characters, these octal escapes are treated as before.
1161     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1162     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1163     Unicode string.
1164    
1165     I have not perpetrated such confusion in PCRE. Up till now, it took just
1166     the bottom 8 bits, as in old Perl. I have now made octal escapes with
1167     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1168     translate to the appropriate multibyte character.
1169    
1170     29. Applied some refactoring to reduce the number of warnings from Microsoft
1171     and Borland compilers. This has included removing the fudge introduced
1172     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1173     a warning about an unused variable.
1174    
1175     21. PCRE has not included VT (character 0x0b) in the set of whitespace
1176     characters since release 4.0, because Perl (from release 5.004) does not.
1177     [Or at least, is documented not to: some releases seem to be in conflict
1178     with the documentation.] However, when a pattern was studied with
1179     pcre_study() and all its branches started with \s, PCRE still included VT
1180     as a possible starting character. Of course, this did no harm; it just
1181     caused an unnecessary match attempt.
1182    
1183     22. Removed a now-redundant internal flag bit that recorded the fact that case
1184     dependency changed within the pattern. This was once needed for "required
1185     byte" processing, but is no longer used. This recovers a now-scarce options
1186     bit. Also moved the least significant internal flag bit to the most-
1187     significant bit of the word, which was not previously used (hangover from
1188     the days when it was an int rather than a uint) to free up another bit for
1189     the future.
1190    
1191     23. Added support for CRLF line endings as well as CR and LF. As well as the
1192     default being selectable at build time, it can now be changed at runtime
1193     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1194     specify that it is scanning data with non-default line endings.
1195    
1196     24. Changed the definition of CXXLINK to make it agree with the definition of
1197     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1198    
1199     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1200     recursions. This makes a big different to stack usage for some patterns.
1201    
1202     26. If a subpattern containing a named recursion or subroutine reference such
1203     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1204     the space required for the compiled pattern went wrong and gave too small a
1205     value. Depending on the environment, this could lead to "Failed: internal
1206     error: code overflow at offset 49" or "glibc detected double free or
1207     corruption" errors.
1208    
1209     27. Applied patches from Google (a) to support the new newline modes and (b) to
1210     advance over multibyte UTF-8 characters in GlobalReplace.
1211    
1212     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1213     difference for some implementation of PCRE in some Windows version.
1214    
1215     29. Added some extra testing facilities to pcretest:
1216    
1217     \q<number> in a data line sets the "match limit" value
1218     \Q<number> in a data line sets the "match recursion limt" value
1219     -S <number> sets the stack size, where <number> is in megabytes
1220    
1221     The -S option isn't available for Windows.
1222    
1223    
1224 nigel 89 Version 6.6 06-Feb-06
1225     ---------------------
1226    
1227     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1228     in pcreposix.h. I have copied the definition from pcre.h.
1229    
1230     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1231     because pcre.h is no longer a built file.
1232    
1233     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1234     not normally included in the compiled code.
1235    
1236    
1237 nigel 87 Version 6.5 01-Feb-06
1238     ---------------------
1239    
1240     1. When using the partial match feature with pcre_dfa_exec(), it was not
1241     anchoring the second and subsequent partial matches at the new starting
1242     point. This could lead to incorrect results. For example, with the pattern
1243     /1234/, partially matching against "123" and then "a4" gave a match.
1244    
1245     2. Changes to pcregrep:
1246    
1247     (a) All non-match returns from pcre_exec() were being treated as failures
1248     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1249     error message is output. Some extra information is given for the
1250     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1251     probably the only errors that are likely to be caused by users (by
1252     specifying a regex that has nested indefinite repeats, for instance).
1253     If there are more than 20 of these errors, pcregrep is abandoned.
1254    
1255     (b) A binary zero was treated as data while matching, but terminated the
1256     output line if it was written out. This has been fixed: binary zeroes
1257     are now no different to any other data bytes.
1258    
1259     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1260     used to set a locale for matching. The --locale=xxxx long option has
1261     been added (no short equivalent) to specify a locale explicitly on the
1262     pcregrep command, overriding the environment variables.
1263    
1264     (d) When -B was used with -n, some line numbers in the output were one less
1265     than they should have been.
1266    
1267     (e) Added the -o (--only-matching) option.
1268    
1269     (f) If -A or -C was used with -c (count only), some lines of context were
1270     accidentally printed for the final match.
1271    
1272     (g) Added the -H (--with-filename) option.
1273    
1274     (h) The combination of options -rh failed to suppress file names for files
1275     that were found from directory arguments.
1276    
1277     (i) Added the -D (--devices) and -d (--directories) options.
1278    
1279     (j) Added the -F (--fixed-strings) option.
1280    
1281     (k) Allow "-" to be used as a file name for -f as well as for a data file.
1282    
1283     (l) Added the --colo(u)r option.
1284    
1285     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1286     is not present by default.
1287    
1288     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1289     items such as (?R) or (?1), when the recursion could match a number of
1290     alternatives. If it matched one of the alternatives, but subsequently,
1291     outside the recursion, there was a failure, the code tried to back up into
1292     the recursion. However, because of the way PCRE is implemented, this is not
1293     possible, and the result was an incorrect result from the match.
1294    
1295     In order to prevent this happening, the specification of recursion has
1296     been changed so that all such subpatterns are automatically treated as
1297     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1298    
1299     4. I had overlooked the fact that, in some locales, there are characters for
1300     which isalpha() is true but neither isupper() nor islower() are true. In
1301     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1302     and ordfeminine) are like this. This affected the treatment of \w and \W
1303     when they appeared in character classes, but not when they appeared outside
1304     a character class. The bit map for "word" characters is now created
1305     separately from the results of isalnum() instead of just taking it from the
1306     upper, lower, and digit maps. (Plus the underscore character, of course.)
1307    
1308     5. The above bug also affected the handling of POSIX character classes such as
1309     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1310     permanent tables. Instead, the bit maps for such a class were previously
1311     created as the appropriate unions of the upper, lower, and digit bitmaps.
1312     Now they are created by subtraction from the [[:word:]] class, which has
1313     its own bitmap.
1314    
1315     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1316     It is created by subtracting the vertical space characters (\x09, \x0a,
1317     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1318     subtraction was done in the overall bitmap for a character class, meaning
1319     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1320     be recognized. This bug has been fixed.
1321    
1322     7. Patches from the folks at Google:
1323    
1324     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1325     real life, but is still worth protecting against".
1326    
1327     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1328     regular expressions".
1329    
1330     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1331     have it.
1332    
1333     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1334     "configure" and the latter not, in order to fix a problem somebody had
1335     with compiling the Arg class on HP-UX.
1336    
1337     (e) Improve the error-handling of the C++ wrapper a little bit.
1338    
1339     (f) New tests for checking recursion limiting.
1340    
1341     8. The pcre_memmove() function, which is used only if the environment does not
1342     have a standard memmove() function (and is therefore rarely compiled),
1343     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1344     returning a result (though PCRE never actually uses the result).
1345    
1346     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1347     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1348     returned instead of calling malloc() with an overflowing number that would
1349     most likely cause subsequent chaos.
1350    
1351     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1352    
1353     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1354     with this option is matched, the nmatch and pmatch options of regexec() are
1355     ignored.
1356    
1357     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1358     provided in case anyone wants to the the POSIX interface with UTF-8
1359     strings.
1360    
1361     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1362     C++ linking (needed for some HP-UX environments).
1363    
1364     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1365     (unused parameter) and in the pcre_printint() function (omitted "default"
1366     switch label when the default is to do nothing).
1367    
1368     15. Added some code to make it possible, when PCRE is compiled as a C++
1369     library, to replace subject pointers for pcre_exec() with a smart pointer
1370     class, thus making it possible to process discontinuous strings.
1371    
1372     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1373     much the same function. They were added by different people who were trying
1374     to make PCRE easy to compile on non-Unix systems. It has been suggested
1375     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1376     for compiling on Windows systems. I have therefore replaced it with
1377     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1378     defaults to "extern" for C or "extern C" for C++, which works fine on
1379     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1380     SCOPE with something explicit in config.h. In addition:
1381    
1382     (a) pcreposix.h still had just "extern" instead of either of these macros;
1383     I have replaced it with PCRE_DATA_SCOPE.
1384    
1385     (b) Functions such as _pcre_xclass(), which are internal to the library,
1386     but external in the C sense, all had PCRE_EXPORT in their definitions.
1387     This is apparently wrong for the Windows case, so I have removed it.
1388     (It makes no difference on Unix-like systems.)
1389    
1390     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1391     of recursive calls to match(). This is different to MATCH_LIMIT because
1392     that limits the total number of calls to match(), not all of which increase
1393     the depth of recursion. Limiting the recursion depth limits the amount of
1394     stack (or heap if NO_RECURSE is set) that is used. The default can be set
1395     when PCRE is compiled, and changed at run time. A patch from Google adds
1396     this functionality to the C++ interface.
1397    
1398     18. Changes to the handling of Unicode character properties:
1399    
1400     (a) Updated the table to Unicode 4.1.0.
1401    
1402     (b) Recognize characters that are not in the table as "Cn" (undefined).
1403    
1404     (c) I revised the way the table is implemented to a much improved format
1405     which includes recognition of ranges. It now supports the ranges that
1406     are defined in UnicodeData.txt, and it also amalgamates other
1407     characters into ranges. This has reduced the number of entries in the
1408     table from around 16,000 to around 3,000, thus reducing its size
1409     considerably. I realized I did not need to use a tree structure after
1410     all - a binary chop search is just as efficient. Having reduced the
1411     number of entries, I extended their size from 6 bytes to 8 bytes to
1412     allow for more data.
1413    
1414     (d) Added support for Unicode script names via properties such as \p{Han}.
1415    
1416     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1417     matching that character.
1418    
1419     20. When matching a repeated Unicode property with a minimum greater than zero,
1420     (for example \pL{2,}), PCRE could look past the end of the subject if it
1421     reached it while seeking the minimum number of characters. This could
1422     happen only if some of the characters were more than one byte long, because
1423     there is a check for at least the minimum number of bytes.
1424    
1425     21. Refactored the implementation of \p and \P so as to be more general, to
1426     allow for more different types of property in future. This has changed the
1427     compiled form incompatibly. Anybody with saved compiled patterns that use
1428     \p or \P will have to recompile them.
1429    
1430     22. Added "Any" and "L&" to the supported property types.
1431    
1432     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1433     but give a compile time error if the value is greater than 0xff.
1434    
1435     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1436     accidentally not being installed or uninstalled.
1437    
1438     25. The pcre.h file was built from pcre.h.in, but the only changes that were
1439     made were to insert the current release number. This seemed silly, because
1440     it made things harder for people building PCRE on systems that don't run
1441     "configure". I have turned pcre.h into a distributed file, no longer built
1442     by "configure", with the version identification directly included. There is
1443     no longer a pcre.h.in file.
1444    
1445     However, this change necessitated a change to the pcre-config script as
1446     well. It is built from pcre-config.in, and one of the substitutions was the
1447     release number. I have updated configure.ac so that ./configure now finds
1448     the release number by grepping pcre.h.
1449    
1450     26. Added the ability to run the tests under valgrind.
1451    
1452    
1453 nigel 85 Version 6.4 05-Sep-05
1454     ---------------------
1455    
1456     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1457     "--" to be printed when multiple files were scanned, even when none of the
1458     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1459     consider it to be a bug, and have restored the previous behaviour.
1460    
1461     2. A couple of code tidies to get rid of compiler warnings.
1462    
1463     3. The pcretest program used to cheat by referring to symbols in the library
1464     whose names begin with _pcre_. These are internal symbols that are not
1465     really supposed to be visible externally, and in some environments it is
1466     possible to suppress them. The cheating is now confined to including
1467     certain files from the library's source, which is a bit cleaner.
1468    
1469     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1470     file's purpose clearer.
1471    
1472     5. Reorganized pcre_ucp_findchar().
1473    
1474    
1475 nigel 83 Version 6.3 15-Aug-05
1476     ---------------------
1477    
1478     1. The file libpcre.pc.in did not have general read permission in the tarball.
1479    
1480     2. There were some problems when building without C++ support:
1481    
1482     (a) If C++ support was not built, "make install" and "make test" still
1483     tried to test it.
1484    
1485     (b) There were problems when the value of CXX was explicitly set. Some
1486     changes have been made to try to fix these, and ...
1487    
1488     (c) --disable-cpp can now be used to explicitly disable C++ support.
1489    
1490     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1491     backslash in a target when C++ was disabled. This confuses some
1492     versions of "make", apparently. Using an intermediate variable solves
1493     this. (Same for CPP_LOBJ.)
1494    
1495     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1496     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1497     necessary on certain architectures.
1498    
1499     4. Added a setting of -export-symbols-regex to the link command to remove
1500     those symbols that are exported in the C sense, but actually are local
1501     within the library, and not documented. Their names all begin with
1502     "_pcre_". This is not a perfect job, because (a) we have to except some
1503     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1504     available (and never for static libraries). I have made a note to try to
1505     find a way round (a) in the future.
1506    
1507    
1508 nigel 81 Version 6.2 01-Aug-05
1509     ---------------------
1510    
1511     1. There was no test for integer overflow of quantifier values. A construction
1512     such as {1111111111111111} would give undefined results. What is worse, if
1513     a minimum quantifier for a parenthesized subpattern overflowed and became
1514     negative, the calculation of the memory size went wrong. This could have
1515     led to memory overwriting.
1516    
1517     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1518    
1519     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1520     operating environments where this matters.
1521    
1522     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1523     PCRE options from within the C++ wrapper.
1524    
1525     5. Named capturing subpatterns were not being correctly counted when a pattern
1526     was compiled. This caused two problems: (a) If there were more than 100
1527     such subpatterns, the calculation of the memory needed for the whole
1528     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1529     back references of the form \12, where the number was greater than 9, were
1530     not recognized as back references, even though there were sufficient
1531     previous subpatterns.
1532    
1533     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1534     versions of gcc, e.g. 2.95.4.
1535    
1536    
1537 nigel 79 Version 6.1 21-Jun-05
1538     ---------------------
1539    
1540     1. There was one reference to the variable "posix" in pcretest.c that was not
1541     surrounded by "#if !defined NOPOSIX".
1542    
1543     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1544     the cross-check on the old pcre_info() function, for the benefit of the
1545     cut-down version of PCRE that is currently imported into Exim.
1546    
1547     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1548     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1549     patterns that start (?i) but I don't think that matters. The use of (?i) is
1550     just an example; this all applies to the other options as well.
1551    
1552     4. Since libtool seems to echo the compile commands it is issuing, the output
1553     from "make" can be reduced a bit by putting "@" in front of each libtool
1554     compile command.
1555    
1556     5. Patch from the folks at Google for configure.in to be a bit more thorough
1557     in checking for a suitable C++ installation before trying to compile the
1558     C++ stuff. This should fix a reported problem when a compiler was present,
1559     but no suitable headers.
1560    
1561     6. The man pages all had just "PCRE" as their title. I have changed them to
1562     be the relevant file name. I have also arranged that these names are
1563     retained in the file doc/pcre.txt, which is a concatenation in text format
1564     of all the man pages except the little individual ones for each function.
1565    
1566     7. The NON-UNIX-USE file had not been updated for the different set of source
1567     files that come with release 6. I also added a few comments about the C++
1568     wrapper.
1569    
1570    
1571 nigel 77 Version 6.0 07-Jun-05
1572     ---------------------
1573    
1574     1. Some minor internal re-organization to help with my DFA experiments.
1575    
1576     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1577     didn't matter for the library itself when fully configured, but did matter
1578     when compiling without UCP support, or within Exim, where the ucp files are
1579     not imported.
1580    
1581     3. Refactoring of the library code to split up the various functions into
1582     different source modules. The addition of the new DFA matching code (see
1583     below) to a single monolithic source would have made it really too
1584     unwieldy, quite apart from causing all the code to be include in a
1585     statically linked application, when only some functions are used. This is
1586     relevant even without the DFA addition now that patterns can be compiled in
1587     one application and matched in another.
1588    
1589     The downside of splitting up is that there have to be some external
1590     functions and data tables that are used internally in different modules of
1591     the library but which are not part of the API. These have all had their
1592     names changed to start with "_pcre_" so that they are unlikely to clash
1593     with other external names.
1594    
1595     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1596     a different (DFA) algorithm. Although it is slower than the original
1597     function, it does have some advantages for certain types of matching
1598     problem.
1599    
1600     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1601     including restarting after a partial match.
1602    
1603     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1604     defined when compiling for Windows was sent to me. I have put it into the
1605     code, though I have no means of testing or verifying it.
1606    
1607     7. Added the pcre_refcount() auxiliary function.
1608    
1609     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1610     match before or at the first newline in the subject string. In pcretest,
1611     the /f option on a pattern can be used to set this.
1612    
1613     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1614     would behave wrongly. This has been present in PCRE since release 4.0.
1615    
1616     10. A number of changes to the pcregrep command:
1617    
1618     (a) Refactored how -x works; insert ^(...)$ instead of setting
1619     PCRE_ANCHORED and checking the length, in preparation for adding
1620     something similar for -w.
1621    
1622     (b) Added the -w (match as a word) option.
1623    
1624     (c) Refactored the way lines are read and buffered so as to have more
1625     than one at a time available.
1626    
1627     (d) Implemented a pcregrep test script.
1628    
1629     (e) Added the -M (multiline match) option. This allows patterns to match
1630     over several lines of the subject. The buffering ensures that at least
1631     8K, or the rest of the document (whichever is the shorter) is available
1632     for matching (and similarly the previous 8K for lookbehind assertions).
1633    
1634     (f) Changed the --help output so that it now says
1635    
1636     -w, --word-regex(p)
1637    
1638     instead of two lines, one with "regex" and the other with "regexp"
1639     because that confused at least one person since the short forms are the
1640     same. (This required a bit of code, as the output is generated
1641     automatically from a table. It wasn't just a text change.)
1642    
1643     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1644     option but starts with a hyphen. Could be a pattern or a path name
1645     starting with a hyphen, for instance.
1646    
1647     (h) "-" can be given as a file name to represent stdin.
1648    
1649     (i) When file names are being printed, "(standard input)" is used for
1650     the standard input, for compatibility with GNU grep. Previously
1651     "<stdin>" was used.
1652    
1653     (j) The option --label=xxx can be used to supply a name to be used for
1654     stdin when file names are being printed. There is no short form.
1655    
1656     (k) Re-factored the options decoding logic because we are going to add
1657     two more options that take data. Such options can now be given in four
1658     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1659    
1660     (l) Added the -A, -B, and -C options for requesting that lines of context
1661     around matches be printed.
1662    
1663     (m) Added the -L option to print the names of files that do not contain
1664     any matching lines, that is, the complement of -l.
1665    
1666     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1667     continue to scan other files.
1668    
1669     (o) The -s option was incorrectly implemented. For compatibility with other
1670     greps, it now suppresses the error message for a non-existent or non-
1671     accessible file (but not the return code). There is a new option called
1672     -q that suppresses the output of matching lines, which was what -s was
1673     previously doing.
1674    
1675     (p) Added --include and --exclude options to specify files for inclusion
1676     and exclusion when recursing.
1677    
1678     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1679     Hopefully, it now does.
1680    
1681     12. Missing cast in pcre_study().
1682    
1683     13. Added an "uninstall" target to the makefile.
1684    
1685     14. Replaced "extern" in the function prototypes in Makefile.in with
1686     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1687     world, but is set differently for Windows.
1688    
1689     15. Added a second compiling function called pcre_compile2(). The only
1690     difference is that it has an extra argument, which is a pointer to an
1691     integer error code. When there is a compile-time failure, this is set
1692     non-zero, in addition to the error test pointer being set to point to an
1693     error message. The new argument may be NULL if no error number is required
1694     (but then you may as well call pcre_compile(), which is now just a
1695     wrapper). This facility is provided because some applications need a
1696     numeric error indication, but it has also enabled me to tidy up the way
1697     compile-time errors are handled in the POSIX wrapper.
1698    
1699     16. Added VPATH=.libs to the makefile; this should help when building with one
1700     prefix path and installing with another. (Or so I'm told by someone who
1701     knows more about this stuff than I do.)
1702    
1703     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1704     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1705     match everything, including newlines. This is not POSIX-compatible, but
1706     somebody wanted the feature. From pcretest it can be activated by using
1707     both the P and the s flags.
1708    
1709     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1710    
1711     19. libpcre.pc was being incorrectly installed as executable.
1712    
1713     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1714     it now also looks for '\r' so that it will work unmodified on Windows.
1715    
1716     21. Added Google's contributed C++ wrapper to the distribution.
1717    
1718     22. Added some untidy missing memory free() calls in pcretest, to keep
1719     Electric Fence happy when testing.
1720    
1721    
1722    
1723 nigel 75 Version 5.0 13-Sep-04
1724     ---------------------
1725    
1726     1. Internal change: literal characters are no longer packed up into items
1727     containing multiple characters in a single byte-string. Each character
1728     is now matched using a separate opcode. However, there may be more than one
1729     byte in the character in UTF-8 mode.
1730    
1731     2. The pcre_callout_block structure has two new fields: pattern_position and
1732     next_item_length. These contain the offset in the pattern to the next match
1733     item, and its length, respectively.
1734    
1735     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1736     insertion of callouts before each pattern item. Added the /C option to
1737     pcretest to make use of this.
1738    
1739     4. On the advice of a Windows user, the lines
1740    
1741     #if defined(_WIN32) || defined(WIN32)
1742     _setmode( _fileno( stdout ), 0x8000 );
1743     #endif /* defined(_WIN32) || defined(WIN32) */
1744    
1745     have been added to the source of pcretest. This apparently does useful
1746     magic in relation to line terminators.
1747    
1748     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1749     for the benefit of those environments where the "b" makes a difference.
1750    
1751     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1752     to know about it. I have put a hack into configure.in that adds in code
1753     to set GCC=yes if CC=icc. This seems to end up at a point in the
1754     generated configure script that is early enough to affect the setting of
1755     compiler options, which is what is needed, but I have no means of testing
1756     whether it really works. (The user who reported this had patched the
1757     generated configure script, which of course I cannot do.)
1758    
1759     LATER: After change 22 below (new libtool files), the configure script
1760     seems to know about icc (and also ecc). Therefore, I have commented out
1761     this hack in configure.in.
1762    
1763     7. Added support for pkg-config (2 patches were sent in).
1764    
1765     8. Negated POSIX character classes that used a combination of internal tables
1766     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1767     [[:^ascii]]. Typically, they would match almost any characters. The other
1768     POSIX classes were not broken in this way.
1769    
1770     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1771     to find the match, as PCRE was deluded into thinking that the match had to
1772     start at the start point or following a newline. The same bug applied to
1773     patterns with negative forward assertions or any backward assertions
1774     preceding ".*" at the start, unless the pattern required a fixed first
1775     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1776    
1777     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1778     starting at the last subject character, bytes beyond the end of the subject
1779     string were read.
1780    
1781     11. Renamed the variable "class" as "classbits" to make life easier for C++
1782     users. (Previously there was a macro definition, but it apparently wasn't
1783     enough.)
1784    
1785     12. Added the new field "tables" to the extra data so that tables can be passed
1786     in at exec time, or the internal tables can be re-selected. This allows
1787     a compiled regex to be saved and re-used at a later time by a different
1788     program that might have everything at different addresses.
1789    
1790     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1791     -R library as well as a -L library.
1792    
1793     14. The debugging options of pcretest (-d on the command line or D on a
1794     pattern) showed incorrect output for anything following an extended class
1795     that contained multibyte characters and which was followed by a quantifier.
1796    
1797     15. Added optional support for general category Unicode character properties
1798     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1799     support. It adds about 90K to the size of the library. The meanings of the
1800     inbuilt class escapes such as \d and \s have NOT been changed.
1801    
1802     16. Updated pcredemo.c to include calls to free() to release the memory for the
1803     compiled pattern.
1804    
1805     17. The generated file chartables.c was being created in the source directory
1806     instead of in the building directory. This caused the build to fail if the
1807     source directory was different from the building directory, and was
1808     read-only.
1809    
1810     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1811     file. No doubt somebody will tell me if they don't make sense... Also added
1812     Dan Mooney's comments about building on OpenVMS.
1813    
1814     19. Added support for partial matching via the PCRE_PARTIAL option for
1815     pcre_exec() and the \P data escape in pcretest.
1816    
1817     20. Extended pcretest with 3 new pattern features:
1818    
1819     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1820     write the compiled pattern to the file whose name is "rest-of-line".
1821     This is a straight binary dump of the data, with the saved pointer to
1822     the character tables forced to be NULL. The study data, if any, is
1823     written too. After writing, pcretest reads a new pattern.
1824    
1825     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1826     compiled pattern from the given file. There must not be any
1827     occurrences of "<" in the file name (pretty unlikely); if there are,
1828     pcretest will instead treat the initial "<" as a pattern delimiter.
1829     After reading in the pattern, pcretest goes on to read data lines as
1830     usual.
1831    
1832     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1833     and 16-bit fields in a compiled pattern, to simulate a pattern that
1834     was compiled on a host of opposite endianness.
1835    
1836     21. The pcre-exec() function can now cope with patterns that were compiled on
1837     hosts of opposite endianness, with this restriction:
1838    
1839     As for any compiled expression that is saved and used later, the tables
1840     pointer field cannot be preserved; the extra_data field in the arguments
1841     to pcre_exec() should be used to pass in a tables address if a value
1842     other than the default internal tables were used at compile time.
1843    
1844     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1845     now diagnosed as an error. Previously, most of the time, a negative number
1846     would have been treated as zero, but if in addition "ovector" was passed as
1847     NULL, a crash could occur.
1848    
1849     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1850     new versions from the libtool 1.5 distribution (the last one is a copy of
1851     a file called libtool.m4). This seems to have fixed the need to patch
1852     "configure" to support Darwin 1.3 (which I used to do). However, I still
1853     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1854     workstation).
1855    
1856     24. Changed the PCRE licence to be the more standard "BSD" licence.
1857    
1858    
1859 nigel 73 Version 4.5 01-Dec-03
1860     ---------------------
1861    
1862     1. There has been some re-arrangement of the code for the match() function so
1863     that it can be compiled in a version that does not call itself recursively.
1864     Instead, it keeps those local variables that need separate instances for
1865     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1866     needs to "recurse". Keeping track of where control must go is done by means
1867     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1868     hide most of the details from the main code, and operates only if
1869     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1870     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1871     operating.
1872    
1873     To make it easier for callers to provide specially tailored get/free
1874     functions for this usage, two new functions, pcre_stack_malloc, and
1875     pcre_stack_free, are used. They are always called in strict stacking order,
1876     and the size of block requested is always the same.
1877    
1878     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1879     PCRE has been compiled to use the stack or the heap for recursion. The
1880     -C option of pcretest uses this to show which version is compiled.
1881    
1882     A new data escape \S, is added to pcretest; it causes the amounts of store
1883     obtained and freed by both kinds of malloc/free at match time to be added
1884     to the output.
1885    
1886     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1887     what's available on my current Linux desktop machine.
1888    
1889     3. When matching a UTF-8 string, the test for a valid string at the start has
1890     been extended. If start_offset is not zero, PCRE now checks that it points
1891     to a byte that is the start of a UTF-8 character. If not, it returns
1892     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1893     this is necessary because there may be backward assertions in the pattern.
1894     When matching the same subject several times, it may save resources to use
1895     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1896    
1897     4. The code for checking the validity of UTF-8 strings has been tightened so
1898     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1899     containing "overlong sequences".
1900    
1901     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1902     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1903     should have been "&", but it just so happened that all the cases this let
1904     through by mistake were picked up later in the function.
1905    
1906     6. I had used a variable called "isblank" - this is a C99 function, causing
1907     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1908    
1909     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1910     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1911     so the version is shown; (c) stop "make test" echoing "./RunTest".
1912    
1913     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1914    
1915     9. The prototype for memmove() for systems that don't have it was using
1916     size_t, but the inclusion of the header that defines size_t was later. I've
1917     moved the #includes for the C headers earlier to avoid this.
1918    
1919     10. Added some adjustments to the code to make it easier to compiler on certain
1920     special systems:
1921    
1922     (a) Some "const" qualifiers were missing.
1923     (b) Added the macro EXPORT before all exported functions; by default this
1924     is defined to be empty.
1925     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1926     that it reads its output file name as an argument instead of writing
1927     to the standard output and assuming this can be redirected.
1928    
1929     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1930     class containing characters with values greater than 255, PCRE compilation
1931     went into a loop.
1932    
1933     12. A recursive reference to a subpattern that was within another subpattern
1934     that had a minimum quantifier of zero caused PCRE to crash. For example,
1935     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1936     recursion. If the recursively-called subpattern itself had a zero repeat,
1937     that was OK.
1938    
1939     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1940     buffer into which it was copied (for escape processing) was still set at
1941     1024, so long lines caused crashes.
1942    
1943     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1944     "internal error: code overflow...". This applied to any character class
1945     that was followed by a possessive quantifier.
1946    
1947     15. Modified the Makefile to add libpcre.la as a prerequisite for
1948     libpcreposix.la because I was told this is needed for a parallel build to
1949     work.
1950    
1951     16. If a pattern that contained .* following optional items at the start was
1952     studied, the wrong optimizing data was generated, leading to matching
1953     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1954     matching string must start with a or b or c. The correct conclusion for
1955     this pattern is that a match can start with any character.
1956    
1957    
1958 nigel 71 Version 4.4 13-Aug-03
1959     ---------------------
1960    
1961     1. In UTF-8 mode, a character class containing characters with values between
1962     127 and 255 was not handled correctly if the compiled pattern was studied.
1963     In fixing this, I have also improved the studying algorithm for such
1964     classes (slightly).
1965    
1966     2. Three internal functions had redundant arguments passed to them. Removal
1967     might give a very teeny performance improvement.
1968    
1969     3. Documentation bug: the value of the capture_top field in a callout is *one
1970     more than* the number of the hightest numbered captured substring.
1971    
1972     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1973     in incorrectly linking with a previously installed version. They now link
1974     explicitly with libpcre.la.
1975    
1976     5. configure.in no longer needs to recognize Cygwin specially.
1977    
1978     6. A problem in pcre.in for Windows platforms is fixed.
1979    
1980     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1981     pcretest, it used to include the size of the study block as part of its
1982     output. Unfortunately, the structure contains a field that has a different
1983     size on different hardware architectures. This meant that the tests that
1984     showed this size failed. As the block is currently always of a fixed size,
1985     this information isn't actually particularly useful in pcretest output, so
1986     I have just removed it.
1987    
1988     8. Three pre-processor statements accidentally did not start in column 1.
1989     Sadly, there are *still* compilers around that complain, even though
1990     standard C has not required this for well over a decade. Sigh.
1991    
1992     9. In pcretest, the code for checking callouts passed small integers in the
1993     callout_data field, which is a void * field. However, some picky compilers
1994     complained about the casts involved for this on 64-bit systems. Now
1995     pcretest passes the address of the small integer instead, which should get
1996     rid of the warnings.
1997    
1998     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1999     both compile and run time, and gives an error if an invalid UTF-8 sequence
2000     is found. There is a option for disabling this check in cases where the
2001     string is known to be correct and/or the maximum performance is wanted.
2002    
2003     11. In response to a bug report, I changed one line in Makefile.in from
2004    
2005     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2006     to
2007     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2008    
2009     to look similar to other lines, but I have no way of telling whether this
2010     is the right thing to do, as I do not use Windows. No doubt I'll get told
2011     if it's wrong...
2012    
2013    
2014 nigel 69 Version 4.3 21-May-03
2015     ---------------------
2016    
2017     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2018     Makefile.
2019    
2020     2. Some refactoring to improve the quality of the code:
2021    
2022     (i) The utf8_table... variables are now declared "const".
2023    
2024     (ii) The code for \cx, which used the "case flipping" table to upper case
2025     lower case letters, now just substracts 32. This is ASCII-specific,
2026     but the whole concept of \cx is ASCII-specific, so it seems
2027     reasonable.
2028    
2029     (iii) PCRE was using its character types table to recognize decimal and
2030     hexadecimal digits in the pattern. This is silly, because it handles
2031     only 0-9, a-f, and A-F, but the character types table is locale-
2032     specific, which means strange things might happen. A private
2033     table is now used for this - though it costs 256 bytes, a table is
2034     much faster than multiple explicit tests. Of course, the standard
2035     character types table is still used for matching digits in subject
2036     strings against \d.
2037    
2038     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2039     ending in _t are). So I've renamed it as ESC_tee.
2040    
2041     3. The first argument for regexec() in the POSIX wrapper should have been
2042     defined as "const".
2043    
2044     4. Changed pcretest to use malloc() for its buffers so that they can be
2045     Electric Fenced for debugging.
2046    
2047     5. There were several places in the code where, in UTF-8 mode, PCRE would try
2048     to read one or more bytes before the start of the subject string. Often this
2049     had no effect on PCRE's behaviour, but in some circumstances it could
2050     provoke a segmentation fault.
2051    
2052     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2053     to try to read one or more bytes before the start of the subject string.
2054    
2055     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2056     UTF-8 support could misbehave in various ways if the subject string
2057     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2058     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2059     back over UTF-8 characters.)
2060    
2061    
2062 nigel 67 Version 4.2 14-Apr-03
2063     ---------------------
2064    
2065     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2066    
2067     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2068     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2069     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2070     [WIN_PREFIX]: new variable, "cyg" for Cygwin
2071     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2072     and BUILD_EXEEXT
2073     Note: automatic setting of the BUILD variables is not yet working
2074     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2075     compile-time but not at link-time
2076     [LINK]: use for linking executables only
2077     make different versions for Windows and non-Windows
2078     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2079     libraries
2080     [LINK_FOR_BUILD]: new variable
2081     [OBJEXT]: use throughout
2082     [EXEEXT]: use throughout
2083     <winshared>: new target
2084     <wininstall>: new target
2085     <dftables.o>: use native compiler
2086     <dftables>: use native linker
2087     <install>: handle Windows platform correctly
2088     <clean>: ditto
2089     <check>: ditto
2090     copy DLL to top builddir before testing
2091    
2092     As part of these changes, -no-undefined was removed again. This was reported
2093     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2094     in any case.
2095    
2096     3. Some tidies to get rid of compiler warnings:
2097    
2098     . In the match_data structure, match_limit was an unsigned long int, whereas
2099     match_call_count was an int. I've made them both unsigned long ints.
2100    
2101     . In pcretest the fact that a const uschar * doesn't automatically cast to
2102     a void * provoked a warning.
2103    
2104     . Turning on some more compiler warnings threw up some "shadow" variables
2105     and a few more missing casts.
2106    
2107     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2108     option, a class that contained a single character with a value between 128
2109     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2110    
2111     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2112     option, a class that contained several characters, but with at least one
2113     whose value was between 128 and 255 caused PCRE to crash.
2114    
2115    
2116 nigel 65 Version 4.1 12-Mar-03
2117     ---------------------
2118 nigel 63
2119 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
2120     needed, and a string in dftables.c that was longer than standard compilers are
2121     required to support.
2122    
2123     2. Compiling with Sun's compiler found a few more places where the code could
2124     be tidied up in order to avoid warnings.
2125    
2126     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2127     first of these names is deprecated in the latest Autoconf in favour of the name
2128     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2129     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2130     analogy I have changed it to CFLAGS_FOR_BUILD.
2131    
2132     4. Added -no-undefined to the linking command in the Makefile, because this is
2133     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2134     linking step for the pcreposix library.
2135    
2136     5. PCRE was failing to diagnose the case of two named groups with the same
2137     name.
2138    
2139     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2140     literal character that is needed in the subject for a match, and scans along to
2141     ensure that it is present before embarking on the full matching process. This
2142     saves time in cases of nested unlimited repeats that are never going to match.
2143     Problem: the scan can take a lot of time if the subject is very long (e.g.
2144     megabytes), thus penalizing straightforward matches. It is now done only if the
2145     amount of subject to be scanned is less than 1000 bytes.
2146    
2147     7. A lesser problem with the same optimization is that it was recording the
2148     first character of an anchored pattern as "needed", thus provoking a search
2149     right along the subject, even when the first match of the pattern was going to
2150     fail. The "needed" character is now not set for anchored patterns, unless it
2151     follows something in the pattern that is of non-fixed length. Thus, it still
2152     fulfils its original purpose of finding quick non-matches in cases of nested
2153     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2154    
2155    
2156     Version 4.0 17-Feb-03
2157     ---------------------
2158    
2159 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
2160     extended to the end of string, PCRE compiled incorrect data. This could lead to
2161     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2162    
2163     2. Moved to autoconf 2.53 and libtool 1.4.2.
2164    
2165     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2166     the special perltest8 script is no longer needed - all the tests can be run
2167     from a single perltest script.
2168    
2169     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2170     by \s. It has now been removed in PCRE. This means it isn't recognized as
2171     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2172     class [:space:] *does* include VT, thereby creating a mess.
2173    
2174     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2175     space and tab.
2176    
2177     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2178     its new features into the main test script, reducing the number of scripts.
2179    
2180     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2181     were backward compatible, and made the (?i) apply to the whole pattern, as if
2182     /i were given. Now it behaves more logically, and applies the option setting
2183     only to what follows. PCRE has been changed to follow suit. However, if it
2184     finds options settings right at the start of the pattern, it extracts them into
2185     the global options, as before. Thus, they show up in the info data.
2186    
2187     8. Added support for the \Q...\E escape sequence. Characters in between are
2188     treated as literals. This is slightly different from Perl in that $ and @ are
2189     also handled as literals inside the quotes. In Perl, they will cause variable
2190     interpolation. Note the following examples:
2191    
2192     Pattern PCRE matches Perl matches
2193    
2194     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2195     \Qabc\$xyz\E abc\$xyz abc\$xyz
2196     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2197    
2198     For compatibility with Perl, \Q...\E sequences are recognized inside character
2199     classes as well as outside them.
2200    
2201     9. Re-organized 3 code statements in pcretest to avoid "overflow in
2202     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2203     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2204     signed/unsigned warnings.
2205    
2206     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2207     option for pcretest, so I've replaced it by a simple function that does just
2208     that job.
2209    
2210     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2211     "pcregrep -".
2212    
2213     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2214     Java package. This provides some syntactic sugar for simple cases of what my
2215     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2216     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2217     item, you can use this simplified notation. Note that only makes sense with
2218     greedy quantifiers. Consequently, the use of the possessive quantifier forces
2219     greediness, whatever the setting of the PCRE_UNGREEDY option.
2220    
2221     13. A change of greediness default within a pattern was not taking effect at
2222     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2223     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2224     was abstracted outside.
2225    
2226     14. PCRE now supports the \G assertion. It is true when the current matching
2227     position is at the start point of the match. This differs from \A when the
2228     starting offset is non-zero. Used with the /g option of pcretest (or similar
2229     code), it works in the same way as it does for Perl's /g option. If all
2230     alternatives of a regex begin with \G, the expression is anchored to the start
2231     match position, and the "anchored" flag is set in the compiled expression.
2232    
2233     15. Some bugs concerning the handling of certain option changes within patterns
2234     have been fixed. These applied to options other than (?ims). For example,
2235     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2236     been the other way round. Some of this was related to change 7 above.
2237    
2238     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2239     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2240     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2241     POSIX classes only within a class (e.g. /[[:alpha:]]/).
2242    
2243     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2244     mode. Unlike ".", it always matches newline, whatever the setting of
2245     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2246     assertions. Perl allows it, but it doesn't (in general) work because it can't
2247     calculate the length of the lookbehind. At least, that's the case for Perl
2248     5.8.0 - I've been told they are going to document that it doesn't work in
2249     future.
2250    
2251     18. Added an error diagnosis for escapes that PCRE does not support: these are
2252     \L, \l, \N, \P, \p, \U, \u, and \X.
2253    
2254     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2255     reading past the end of the pattern in cases such as /[abcd/.
2256    
2257     20. PCRE was getting more memory than necessary for patterns with classes that
2258     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2259    
2260     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2261     compiling PCRE for use with Virtual Pascal.
2262    
2263     22. Small fix to the Makefile to make it work properly if the build is done
2264     outside the source tree.
2265    
2266     23. Added a new extension: a condition to go with recursion. If a conditional
2267     subpattern starts with (?(R) the "true" branch is used if recursion has
2268     happened, whereas the "false" branch is used only at the top level.
2269    
2270     24. When there was a very long string of literal characters (over 255 bytes
2271     without UTF support, over 250 bytes with UTF support), the computation of how
2272     much memory was required could be incorrect, leading to segfaults or other
2273     strange effects.
2274    
2275     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2276     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2277     there was a subsequent back reference to those brackets. This meant that, for
2278     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2279     possible to check for precisely this case. All we can do is abandon the
2280     optimization if .* occurs inside capturing brackets when there are any back
2281     references whatsoever. (See below for a better fix that came later.)
2282    
2283     26. The handling of the optimization for finding the first character of a
2284     non-anchored pattern, and for finding a character that is required later in the
2285     match were failing in some cases. This didn't break the matching; it just
2286     failed to optimize when it could. The way this is done has been re-implemented.
2287    
2288     27. Fixed typo in error message for invalid (?R item (it said "(?p").
2289    
2290     28. Added a new feature that provides some of the functionality that Perl
2291     provides with (?{...}). The facility is termed a "callout". The way it is done
2292     in PCRE is for the caller to provide an optional function, by setting
2293     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2294     global variable. By default it is unset, which disables all calling out. To get
2295     the function called, the regex must include (?C) at appropriate points. This
2296     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2297     This provides a means of identifying different callout points. When PCRE
2298     reaches such a point in the regex, if pcre_callout has been set, the external
2299     function is called. It is provided with data in a structure called
2300     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2301     matching continues; if it returns a non-zero value, the match at the current
2302     point fails. However, backtracking will occur if possible. [This was changed
2303     later and other features added - see item 49 below.]
2304    
2305     29. pcretest is upgraded to test the callout functionality. It provides a
2306     callout function that displays information. By default, it shows the start of
2307     the match and the current position in the text. There are some new data escapes
2308     to vary what happens:
2309    
2310     \C+ in addition, show current contents of captured substrings
2311     \C- do not supply a callout function
2312     \C!n return 1 when callout number n is reached
2313     \C!n!m return 1 when callout number n is reached for the mth time
2314    
2315     30. If pcregrep was called with the -l option and just a single file name, it
2316     output "<stdin>" if a match was found, instead of the file name.
2317    
2318     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2319     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2320     pcre_exec(). This saves a malloc/free per call. The default value of
2321     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2322     when configuring.
2323    
2324     32. The default maximum size of a compiled pattern is 64K. There have been a
2325     few cases of people hitting this limit. The code now uses macros to handle the
2326     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2327     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2328     configuring. Tests 2 and 5 work only with 2-byte links because they output
2329     debugging information about compiled patterns.
2330    
2331     33. Internal code re-arrangements:
2332    
2333     (a) Moved the debugging function for printing out a compiled regex into
2334     its own source file (printint.c) and used #include to pull it into
2335     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2336     separate copies.
2337    
2338     (b) Defined the list of op-code names for debugging as a macro in
2339     internal.h so that it is next to the definition of the opcodes.
2340    
2341     (c) Defined a table of op-code lengths for simpler skipping along compiled
2342     code. This is again a macro in internal.h so that it is next to the
2343     definition of the opcodes.
2344    
2345     34. Added support for recursive calls to individual subpatterns, along the
2346     lines of Robin Houston's patch (but implemented somewhat differently).
2347    
2348     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2349     allow it to read and process whole directories in Win32. This code was
2350     contributed by Lionel Fourquaux; it has not been tested by me.
2351    
2352     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2353     used to name a group. Names consist of alphanumerics and underscores, and must
2354     be unique. Back references use the syntax (?P=name) and recursive calls use
2355     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2356     numbers. The function pcre_fullinfo() can be used after compilation to extract
2357     a name/number map. There are three relevant calls:
2358    
2359     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2360     PCRE_INFO_NAMECOUNT yields the number of entries
2361     PCRE_INFO_NAMETABLE yields a pointer to the map.
2362    
2363     The map is a vector of fixed-size entries. The size of each entry depends on
2364     the length of the longest name used. The first two bytes of each entry are the
2365     group number, most significant byte first. There follows the corresponding
2366     name, zero terminated. The names are in alphabetical order.
2367    
2368     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2369     case instead of 255. Making it the same both with and without UTF-8 support
2370     means that the same test output works with both.
2371    
2372     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2373     calling malloc() with a zero argument.
2374    
2375     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2376     optimization. I've improved things by keeping a bitmap of backreferences with
2377     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2378     fact referenced, the optimization can be applied. It is unlikely that a
2379     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2380     the match to follow \n) will appear inside brackets with a number greater than
2381     31, but if it does, any back reference > 31 suppresses the optimization.
2382    
2383     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2384     of disabling numbered capturing parentheses. Any opening parenthesis that is
2385     not followed by ? behaves as if it were followed by ?: but named parentheses
2386     can still be used for capturing (and they will acquire numbers in the usual
2387     way).
2388    
2389     41. Redesigned the return codes from the match() function into yes/no/error so
2390     that errors can be passed back from deep inside the nested calls. A malloc
2391     failure while inside a recursive subpattern call now causes the
2392     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2393    
2394     42. It is now possible to set a limit on the number of times the match()
2395     function is called in a call to pcre_exec(). This facility makes it possible to
2396     limit the amount of recursion and backtracking, though not in a directly
2397     obvious way, because the match() function is used in a number of different
2398     circumstances. The count starts from zero for each position in the subject
2399     string (for non-anchored patterns). The default limit is, for compatibility, a
2400     large number, namely 10 000 000. You can change this in two ways:
2401    
2402     (a) When configuring PCRE before making, you can use --with-match-limit=n
2403     to set a default value for the compiled library.
2404    
2405     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2406     a different value is set. See 45 below.
2407    
2408     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2409    
2410     43. Added a new function pcre_config(int, void *) to enable run-time extraction
2411     of things that can be changed at compile time. The first argument specifies
2412     what is wanted and the second points to where the information is to be placed.
2413     The current list of available information is:
2414    
2415     PCRE_CONFIG_UTF8
2416    
2417     The output is an integer that is set to one if UTF-8 support is available;
2418     otherwise it is set to zero.
2419    
2420     PCRE_CONFIG_NEWLINE
2421    
2422     The output is an integer that it set to the value of the code that is used for
2423     newline. It is either LF (10) or CR (13).
2424    
2425     PCRE_CONFIG_LINK_SIZE
2426    
2427     The output is an integer that contains the number of bytes used for internal
2428     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2429    
2430     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2431    
2432     The output is an integer that contains the threshold above which the POSIX
2433     interface uses malloc() for output vectors. See item 31 above.
2434    
2435     PCRE_CONFIG_MATCH_LIMIT
2436    
2437     The output is an unsigned integer that contains the default limit of the number
2438     of match() calls in a pcre_exec() execution. See 42 above.
2439    
2440     44. pcretest has been upgraded by the addition of the -C option. This causes it
2441     to extract all the available output from the new pcre_config() function, and to
2442     output it. The program then exits immediately.
2443    
2444     45. A need has arisen to pass over additional data with calls to pcre_exec() in
2445     order to support additional features. One way would have been to define
2446     pcre_exec2() (for example) with extra arguments, but this would not have been
2447     extensible, and would also have required all calls to the original function to
2448     be mapped to the new one. Instead, I have chosen to extend the mechanism that
2449     is used for passing in "extra" data from pcre_study().
2450    
2451     The pcre_extra structure is now exposed and defined in pcre.h. It currently
2452     contains the following fields:
2453    
2454     flags a bitmap indicating which of the following fields are set
2455     study_data opaque data from pcre_study()
2456     match_limit a way of specifying a limit on match() calls for a specific
2457     call to pcre_exec()
2458     callout_data data for callouts (see 49 below)
2459    
2460     The flag bits are also defined in pcre.h, and are
2461    
2462     PCRE_EXTRA_STUDY_DATA
2463     PCRE_EXTRA_MATCH_LIMIT
2464     PCRE_EXTRA_CALLOUT_DATA
2465    
2466     The pcre_study() function now returns one of these new pcre_extra blocks, with
2467     the actual study data pointed to by the study_data field, and the
2468     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2469     before. That is, this change is entirely upwards-compatible and requires no
2470     change to existing code.
2471    
2472     If you want to pass in additional data to pcre_exec(), you can either place it
2473     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2474     block.
2475    
2476     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2477     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2478     times with different match limits, until it finds the minimum value needed for
2479     pcre_exec() to complete. The value is then output. This can be instructive; for
2480     most simple matches the number is quite small, but for pathological cases it
2481     gets very large very quickly.
2482    
2483     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2484     returns the size of the data block pointed to by the study_data field in a
2485     pcre_extra block, that is, the value that was passed as the argument to
2486     pcre_malloc() when PCRE was getting memory in which to place the information
2487     created by pcre_study(). The fourth argument should point to a size_t variable.
2488     pcretest has been extended so that this information is shown after a successful
2489     pcre_study() call when information about the compiled regex is being displayed.
2490    
2491     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2492     because what follows is always an absolute path. (Later: it turns out that this
2493     is more than cosmetic for MinGW, because it doesn't like empty path
2494     components.)
2495    
2496     49. Some changes have been made to the callout feature (see 28 above):
2497    
2498     (i) A callout function now has three choices for what it returns:
2499    
2500     0 => success, carry on matching
2501     > 0 => failure at this point, but backtrack if possible
2502     < 0 => serious error, return this value from pcre_exec()
2503    
2504     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2505     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2506     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2507     use by callout functions. It will never be used by PCRE itself.
2508    
2509     (ii) The pcre_extra structure (see 45 above) has a void * field called
2510     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2511     pcre_callout_block structure has a field of the same name. The contents of
2512     the field passed in the pcre_extra structure are passed to the callout
2513     function in the corresponding field in the callout block. This makes it
2514     easier to use the same callout-containing regex from multiple threads. For
2515     testing, the pcretest program has a new data escape
2516    
2517     \C*n pass the number n (may be negative) as callout_data
2518    
2519     If the callout function in pcretest receives a non-zero value as
2520     callout_data, it returns that value.
2521    
2522     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2523     there were some redundant $(CFLAGS) in commands that are now specified as
2524     $(LINK), which already includes $(CFLAGS).
2525    
2526     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2527     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2528     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2529     one-byte characters throughout. Note that case-insensitive matching applies
2530     only to characters whose values are less than 256. PCRE doesn't support the
2531     notion of cases for higher-valued characters.
2532    
2533     (i) A character class whose characters are all within 0-255 is handled as
2534     a bit map, and the map is inverted for negative classes. Previously, a
2535     character > 255 always failed to match such a class; however it should
2536     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2537    
2538     (ii) A negated character class with a single character < 255 is coded as
2539     "not this character" (OP_NOT). This wasn't working properly when the test
2540     character was multibyte, either singly or repeated.
2541    
2542     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2543     mode, for example: \x{100}{2,3}.
2544    
2545     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2546     singly or repeated) now correctly test multibyte characters. However,
2547     PCRE doesn't recognize any characters with values greater than 255 as
2548     digits, spaces, or word characters. Such characters always match \D, \S,
2549     and \W, and never match \d, \s, or \w.
2550    
2551     (v) Classes may now contain characters and character ranges with values
2552     greater than 255. For example: [ab\x{100}-\x{400}].
2553    
2554     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2555     PCRE in UTF-8 mode.
2556    
2557     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2558     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2559     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2560     value.)
2561    
2562     53. The single man page has become too large. I have therefore split it up into
2563     a number of separate man pages. These also give rise to individual HTML pages;
2564     these are now put in a separate directory, and there is an index.html page that
2565     lists them all. Some hyperlinking between the pages has been installed.
2566    
2567     54. Added convenience functions for handling named capturing parentheses.
2568    
2569     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2570     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2571     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2572     are faulted.
2573    
2574     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2575     calling configure. These values are used when compiling the dftables.c program
2576     which is run to generate the source of the default character tables. They
2577     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2578     you will need to set these values.
2579    
2580     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2581    
2582    
2583     Version 3.9 02-Jan-02
2584 nigel 61 ---------------------
2585    
2586     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2587    
2588     2. If --disable-static was given, the building process failed when trying to
2589     build pcretest and pcregrep. (For some reason it was using libtool to compile
2590     them, which is not right, as they aren't part of the library.)
2591    
2592    
2593 nigel 59 Version 3.8 18-Dec-01
2594     ---------------------
2595    
2596     1. The experimental UTF-8 code was completely screwed up. It was packing the
2597     bytes in the wrong order. How dumb can you get?
2598    
2599    
2600 nigel 57 Version 3.7 29-Oct-01
2601     ---------------------
2602    
2603     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2604     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2605     this didn't happen under Solaris 8, where I normally test things.
2606    
2607 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2608     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2609     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2610     this fix an hour or so after the initial 3.7 release.)
2611 nigel 57
2612 nigel 59
2613 nigel 55 Version 3.6 23-Oct-01
2614     ---------------------
2615    
2616     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2617     offsets passed as NULL with zero offset count.
2618    
2619     2. The config.guess and config.sub files had not been updated when I moved to
2620     the latest autoconf.
2621    
2622    
2623 nigel 53 Version 3.5 15-Aug-01
2624     ---------------------
2625 nigel 5
2626 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2627     had been forgotten.
2628    
2629     2. By using declared but undefined structures, we can avoid using "void"
2630     definitions in pcre.h while keeping the internal definitions of the structures
2631     private.
2632    
2633     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2634     user point of view, this means that both static and shared libraries are built
2635     by default, but this can be individually controlled. More of the work of
2636     handling this static/shared cases is now inside libtool instead of PCRE's make
2637     file.
2638    
2639     4. The pcretest utility is now installed along with pcregrep because it is
2640     useful for users (to test regexs) and by doing this, it automatically gets
2641     relinked by libtool. The documentation has been turned into a man page, so
2642     there are now .1, .txt, and .html versions in /doc.
2643    
2644     5. Upgrades to pcregrep:
2645     (i) Added long-form option names like gnu grep.
2646     (ii) Added --help to list all options with an explanatory phrase.
2647     (iii) Added -r, --recursive to recurse into sub-directories.
2648     (iv) Added -f, --file to read patterns from a file.
2649    
2650     6. pcre_exec() was referring to its "code" argument before testing that
2651     argument for NULL (and giving an error if it was NULL).
2652    
2653     7. Upgraded Makefile.in to allow for compiling in a different directory from
2654     the source directory.
2655    
2656     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2657     options bits, the pointer it was passed was to an int instead of to an unsigned
2658     long int. This mattered only on 64-bit systems.
2659    
2660     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2661     generated) instead of pcre.in, which it its source. Also made the same change
2662     in several of the .c files.
2663    
2664     10. A new release of gcc defines printf() as a macro, which broke pcretest
2665     because it had an ifdef in the middle of a string argument for printf(). Fixed
2666     by using separate calls to printf().
2667    
2668     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2669     script, to force use of CR or LF instead of \n in the source. On non-Unix
2670     systems, the value can be set in config.h.
2671    
2672     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2673     absolute limit. Changed the text of the error message to make this clear, and
2674     likewise updated the man page.
2675    
2676     13. The limit of 99 on the number of capturing subpatterns has been removed.
2677     The new limit is 65535, which I hope will not be a "real" limit.
2678    
2679    
2680 nigel 51 Version 3.4 22-Aug-00
2681     ---------------------
2682    
2683     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2684    
2685     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2686    
2687    
2688 nigel 49 Version 3.3 01-Aug-00
2689     ---------------------
2690    
2691     1. If an octal character was given, but the value was greater than \377, it
2692     was not getting masked to the least significant bits, as documented. This could
2693     lead to crashes in some systems.
2694    
2695     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2696     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2697    
2698     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2699     These just pass their arguments on to (pcre_free)(), but they are provided
2700     because some uses of PCRE bind it to non-C systems that can call its functions,
2701     but cannot call free() or pcre_free() directly.
2702    
2703     4. Add "make test" as a synonym for "make check". Corrected some comments in
2704     the Makefile.
2705    
2706     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2707     Makefile.
2708    
2709     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2710     command called pgrep for grepping around the active processes.
2711    
2712     7. Added the beginnings of support for UTF-8 character strings.
2713    
2714     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2715     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2716     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2717     out for the ar command.)
2718    
2719    
2720 nigel 47 Version 3.2 12-May-00
2721     ---------------------
2722    
2723     This is purely a bug fixing release.
2724    
2725     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2726     of ZA. This was just one example of several cases that could provoke this bug,
2727     which was introduced by change 9 of version 2.00. The code for breaking
2728     infinite loops after an iteration that matches an empty string was't working
2729     correctly.
2730    
2731     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2732     when matched against abbab (for example). After matching an empty string, it
2733     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2734     caused it to match further down the string than it should.
2735    
2736     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2737     was there because it doesn't seem to be needed, and it causes trouble on some
2738     systems, as it is not a Standard C header. It has been removed.
2739    
2740     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2741     were reported on the Macintosh. The changes were from
2742    
2743     while ((c = *(++ptr)) != 0 && c != '\n');
2744     to
2745     while ((c = *(++ptr)) != 0 && c != '\n') ;
2746    
2747     Totally extraordinary, but if that's what it takes...
2748    
2749     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2750     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2751     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2752     assumes the way PCRE uses memmove() (always moving upwards).
2753    
2754     6. PCRE is being used in one environment where strchr() is not available. There
2755     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2756     faster code anyway.
2757    
2758    
2759 nigel 45 Version 3.1 09-Feb-00
2760     ---------------------
2761    
2762     The only change in this release is the fixing of some bugs in Makefile.in for
2763     the "install" target:
2764    
2765     (1) It was failing to install pcreposix.h.
2766    
2767     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2768    
2769    
2770 nigel 43 Version 3.0 01-Feb-00
2771     ---------------------
2772 nigel 41
2773     1. Add support for the /+ modifier to perltest (to output $` like it does in
2774     pcretest).
2775    
2776     2. Add support for the /g modifier to perltest.
2777    
2778     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2779     matches null strings.
2780    
2781     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2782     pattern. Perl treats empty patterns specially - it reuses the most recent
2783     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2784     effect.
2785    
2786     5. The POSIX interface was broken in that it was just handing over the POSIX
2787     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2788     required a bigger vector, with some working space on the end. This means that
2789     the POSIX wrapper now has to get and free some memory, and copy the results.
2790    
2791 nigel 43 6. Added some simple autoconf support, placing the test data and the
2792     documentation in separate directories, re-organizing some of the
2793     information files, and making it build pcre-config (a GNU standard). Also added
2794     libtool support for building PCRE as a shared library, which is now the
2795     default.
2796 nigel 41
2797 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2798     09 are not valid octal constants. Single digits will be used for minor values
2799     less than 10.
2800    
2801     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2802     existing programs that set these in the POSIX interface can use PCRE without
2803     modification.
2804    
2805     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2806     return all that pcre_info() returns, plus additional data. The pcre_info()
2807     function is retained for compatibility, but is considered to be obsolete.
2808    
2809     10. Added experimental recursion feature (?R) to handle one common case that
2810     Perl 5.6 will be able to do with (?p{...}).
2811    
2812     11. Added support for POSIX character classes like [:alpha:], which Perl is
2813     adopting.
2814    
2815    
2816 nigel 39 Version 2.08 31-Aug-99
2817     ----------------------
2818    
2819     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2820     trying to match at the startoffset position, but instead was moving forward to
2821     the next newline as if a previous match had failed.
2822    
2823     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2824     and could get into a loop if a null string was matched other than at the start
2825     of the subject.
2826    
2827     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2828     be distinguished at compile time, and for completeness also added PCRE_DATE.
2829    
2830     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2831     in GnuWin32 environments.
2832    
2833    
2834 nigel 37 Version 2.07 29-Jul-99
2835     ----------------------
2836    
2837     1. The documentation is now supplied in plain text form and HTML as well as in
2838     the form of man page sources.
2839    
2840     2. C++ compilers don't like assigning (void *) values to other pointer types.
2841     In particular this affects malloc(). Although there is no problem in Standard
2842     C, I've put in casts to keep C++ compilers happy.
2843    
2844     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2845     should be (const char *).
2846    
2847     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2848     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2849     However, I haven't made this a standard facility. The documentation doesn't
2850     mention it, and the Makefile doesn't support it.
2851    
2852     5. The Makefile now contains an "install" target, with editable destinations at
2853     the top of the file. The pcretest program is not installed.
2854    
2855     6. pgrep -V now gives the PCRE version number and date.
2856    
2857     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2858     causing the entire string to be ignored, instead of just the last character.
2859    
2860     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2861     non-matching string, it can take a very, very long time, even for strings of
2862     quite modest length, because of the nested recursion. PCRE now does better in
2863     some of these cases. It does this by remembering the last required literal
2864     character in the pattern, and pre-searching the subject to ensure it is present
2865     before running the real match. In other words, it applies a heuristic to detect
2866     some types of certain failure quickly, and in the above example, if presented
2867     with a string that has no trailing " it gives "no match" very quickly.
2868    
2869     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2870     other alternatives are tried instead.
2871    
2872    
2873 nigel 35 Version 2.06 09-Jun-99
2874     ----------------------
2875    
2876     1. Change pcretest's output for amount of store used to show just the code
2877     space, because the remainder (the data block) varies in size between 32-bit and
2878     64-bit systems.
2879    
2880     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2881     start matching at. This allows lookbehinds to work when searching for multiple
2882     occurrences in a string.
2883    
2884     3. Added additional options to pcretest for testing multiple occurrences:
2885    
2886     /+ outputs the rest of the string that follows a match
2887     /g loops for multiple occurrences, using the new startoffset argument
2888     /G loops for multiple occurrences by passing an incremented pointer
2889    
2890     4. PCRE wasn't doing the "first character" optimization for patterns starting
2891     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2892     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2893     the letter 'x'. On long subject strings, this gives a significant speed-up.
2894    
2895    
2896 nigel 33 Version 2.05 21-Apr-99
2897     ----------------------
2898    
2899     1. Changed the type of magic_number from int to long int so that it works
2900     properly on 16-bit systems.
2901    
2902     2. Fixed a bug which caused patterns starting with .* not to work correctly
2903     when the subject string contained newline characters. PCRE was assuming
2904     anchoring for such patterns in all cases, which is not correct because .* will
2905     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2906     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2907     must be retried after every newline in the subject.
2908    
2909    
2910 nigel 31 Version 2.04 18-Feb-99
2911     ----------------------
2912    
2913     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2914     computation of the store needed to hold the pattern was incorrect (too large).
2915     If such patterns were nested a few deep, this could multiply and become a real
2916     problem.
2917    
2918     2. Added /M option to pcretest to show the memory requirement of a specific
2919     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2920    
2921     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2922     compiled in such a way that the backtracking after subsequent failure was
2923     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2924     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2925    
2926    
2927 nigel 29 Version 2.03 02-Feb-99
2928     ----------------------
2929    
2930     1. Fixed typo and small mistake in man page.
2931    
2932 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2933     LICENCE file containing the conditions.
2934 nigel 29
2935     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2936     Perl, that is the internal \ allows the delimiter to be included in the
2937     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2938     the final delimiter, add \ to the end of the pattern (to test the error).
2939    
2940     4. Added the convenience functions for extracting substrings after a successful
2941     match. Updated pcretest to make it able to test these functions.
2942    
2943    
2944 nigel 27 Version 2.02 14-Jan-99
2945     ----------------------
2946    
2947     1. Initialized the working variables associated with each extraction so that
2948     their saving and restoring doesn't refer to uninitialized store.
2949    
2950     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2951     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2952     fix the problem.
2953    
2954     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2955     calls, and wasn't printing the correct value for compiling calls. Increased the
2956     default value of LOOPREPEAT, and the number of significant figures in the
2957     times.
2958    
2959     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2960    
2961     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2962     a building problem on Windows NT with a FAT file system.
2963    
2964    
2965 nigel 25 Version 2.01 21-Oct-98
2966     ----------------------
2967    
2968     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2969     to character tables built by pcre_maketables() in the current locale. If NULL
2970     is passed, the default tables are used.
2971    
2972    
2973 nigel 23 Version 2.00 24-Sep-98
2974 nigel 21 ----------------------
2975    
2976 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2977     it any more.
2978    
2979     2. Allow quantification of (?>) groups, and make it work correctly.
2980    
2981     3. The first character computation wasn't working for (?>) groups.
2982    
2983     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2984     end of the subject) and add 5.005's \z, which really does match only at the
2985     very end of the subject.
2986    
2987     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2988    
2989     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2990     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2991     localized options. All options to pcre_study() were also removed.
2992    
2993     7. Add other new features from 5.005:
2994    
2995     $(?<= positive lookbehind
2996     $(?<! negative lookbehind
2997     (?imsx-imsx) added the unsetting capability
2998     such a setting is global if at outer level; local otherwise
2999     (?imsx-imsx:) non-capturing groups with option setting
3000     (?(cond)re|re) conditional pattern matching
3001    
3002     A backreference to itself in a repeated group matches the previous
3003     captured string.
3004    
3005     8. General tidying up of studying (both automatic and via "study")
3006     consequential on the addition of new assertions.
3007    
3008     9. As in 5.005, unlimited repeated groups that could match an empty substring
3009     are no longer faulted at compile time. Instead, the loop is forcibly broken at
3010     runtime if any iteration does actually match an empty substring.
3011    
3012     10. Include the RunTest script in the distribution.
3013    
3014     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3015     discrepancies, some of which were old and were also with respect to 5.004. They
3016     have now been fixed.
3017    
3018    
3019     Version 1.09 28-Apr-98
3020     ----------------------
3021    
3022 nigel 21 1. A negated single character class followed by a quantifier with a minimum
3023     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3024     program crashes, or just wrong answers. This did not apply to negated classes
3025     containing more than one character, or to minima other than one.
3026    
3027    
3028 nigel 19 Version 1.08 27-Mar-98
3029     ----------------------
3030    
3031     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3032    
3033     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3034     latter must appear before anything that relies on it in the pattern.
3035    
3036    
3037 nigel 17 Version 1.07 16-Feb-98
3038     ----------------------
3039    
3040     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3041     repeat of a potentially empty string).
3042    
3043    
3044 nigel 15 Version 1.06 23-Jan-98
3045     ----------------------
3046    
3047     1. Added Markus Oberhumer's little patches for C++.
3048    
3049     2. Literal strings longer than 255 characters were broken.
3050    
3051    
3052 nigel 13 Version 1.05 23-Dec-97
3053     ----------------------
3054    
3055     1. Negated character classes containing more than one character were failing if
3056     PCRE_CASELESS was set at run time.
3057    
3058    
3059 nigel 11 Version 1.04 19-Dec-97
3060     ----------------------
3061    
3062     1. Corrected the man page, where some "const" qualifiers had been omitted.
3063    
3064     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3065     input syntax.
3066    
3067     3. Fixed memory leak which occurred when a regex with back references was
3068     matched with an offsets vector that wasn't big enough. The temporary memory
3069     that is used in this case wasn't being freed if the match failed.
3070    
3071     4. Tidied pcretest to ensure it frees memory that it gets.
3072    
3073     5. Temporary memory was being obtained in the case where the passed offsets
3074     vector was exactly big enough.
3075    
3076     6. Corrected definition of offsetof() from change 5 below.
3077    
3078     7. I had screwed up change 6 below and broken the rules for the use of
3079     setjmp(). Now fixed.
3080    
3081    
3082 nigel 9 Version 1.03 18-Dec-97
3083     ----------------------
3084    
3085     1. A erroneous regex with a missing opening parenthesis was correctly
3086     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3087     on some systems.
3088    
3089     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3090     it was reported that one broken compiler failed on the former because "code" is
3091     also an independent variable.
3092    
3093     3. The erroneous regex a[]b caused an array overrun reference.
3094    
3095     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3096     fail on data ending with that character. (It was going on too far, and checking
3097     the next character, typically a binary zero.) This was specific to the
3098     optimized code for single-character negative classes.
3099    
3100     5. Added a contributed patch from the TIN world which does the following:
3101    
3102     + Add an undef for memmove, in case the the system defines a macro for it.
3103    
3104     + Add a definition of offsetof(), in case there isn't one. (I don't know
3105     the reason behind this - offsetof() is part of the ANSI standard - but
3106     it does no harm).
3107    
3108     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3109     most of the places where whitespace preceded '#'. I have given up and
3110     allowed the remaining 2 cases to be at the margin.
3111    
3112     + Rename some variables in pcre to eliminate shadowing. This seems very
3113     pedantic, but does no harm, of course.
3114    
3115     6. Moved the call to setjmp() into its own function, to get rid of warnings
3116     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3117    
3118     7. Constructs such as \d{8,} were compiling into the equivalent of
3119 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3120 nigel 9 outcome, but in this particular case used more store than had been allocated,
3121     which caused the bug to be discovered because it threw up an internal error.
3122    
3123     8. The debugging code in both pcre and pcretest for outputting the compiled
3124     form of a regex was going wrong in the case of back references followed by
3125     curly-bracketed repeats.
3126    
3127    
3128 nigel 7 Version 1.02 12-Dec-97
3129     ----------------------
3130    
3131     1. Typos in pcre.3 and comments in the source fixed.
3132    
3133     2. Applied a contributed patch to get rid of places where it used to remove
3134     'const' from variables, and fixed some signed/unsigned and uninitialized
3135     variable warnings.
3136    
3137     3. Added the "runtest" target to Makefile.
3138    
3139     4. Set default compiler flag to -O2 rather than just -O.
3140    
3141    
3142 nigel 5 Version 1.01 19-Nov-97
3143     ----------------------
3144    
3145     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3146     like /([ab]*)*/, that is, for classes with more than one character in them.
3147    
3148     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3149     as /((?>a*))*/ (a PCRE_EXTRA facility).
3150    
3151    
3152     Version 1.00 18-Nov-97
3153     ----------------------
3154    
3155     1. Added compile-time macros to support systems such as SunOS4 which don't have
3156     memmove() or strerror() but have other things that can be used instead.
3157    
3158     2. Arranged that "make clean" removes the executables.
3159    
3160    
3161 nigel 3 Version 0.99 27-Oct-97
3162     ----------------------
3163    
3164     1. Fixed bug in code for optimizing classes with only one character. It was
3165     initializing a 32-byte map regardless, which could cause it to run off the end
3166     of the memory it had got.
3167    
3168     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3169    
3170    
3171     Version 0.98 22-Oct-97
3172     ----------------------
3173    
3174     1. Fixed bug in code for handling temporary memory usage when there are more
3175     back references than supplied space in the ovector. This could cause segfaults.
3176    
3177    
3178     Version 0.97 21-Oct-97
3179     ----------------------
3180    
3181     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3182    
3183     2. Optimized negated single characters not to use a bit map.
3184    
3185     3. Brought error texts together as macro definitions; clarified some of them;
3186     fixed one that was wrong - it said "range out of order" when it meant "invalid
3187     escape sequence".
3188    
3189     4. Changed some char * arguments to const char *.
3190    
3191     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3192    
3193     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3194     pcretest.
3195    
3196    
3197     Version 0.96 16-Oct-97
3198     ----------------------
3199    
3200     1. Added a simple "pgrep" utility to the distribution.
3201    
3202     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3203     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3204     where "ddd" means "one or more decimal digits".
3205    
3206     3. Fixed serious bug. If a pattern had a back reference, but the call to
3207     pcre_exec() didn't supply a large enough ovector to record the related
3208     identifying subpattern, the match always failed. PCRE now remembers the number
3209     of the largest back reference, and gets some temporary memory in which to save
3210     the offsets during matching if necessary, in order to ensure that
3211     backreferences always work.
3212    
3213     4. Increased the compatibility with Perl in a number of ways:
3214    
3215     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3216     to request this handling. The option can be set at compile or exec time.
3217    
3218     (b) $ matches before a terminating newline by default; an option
3219     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3220     mode). The option can be set at compile or exec time.
3221    
3222     (c) The handling of \ followed by a digit other than 0 is now supposed to be
3223     the same as Perl's. If the decimal number it represents is less than 10
3224     or there aren't that many previous left capturing parentheses, an octal
3225     escape is read. Inside a character class, it's always an octal escape,
3226     even if it is a single digit.
3227    
3228     (d) An escaped but undefined alphabetic character is taken as a literal,
3229     unless PCRE_EXTRA is set. Currently this just reserves the remaining
3230     escapes.
3231    
3232     (e) {0} is now permitted. (The previous item is removed from the compiled
3233     pattern).
3234    
3235     5. Changed all the names of code files so that the basic parts are no longer
3236     than 10 characters, and abolished the teeny "globals.c" file.
3237    
3238     6. Changed the handling of character classes; they are now done with a 32-byte
3239     bit map always.
3240    
3241     7. Added the -d and /D options to pcretest to make it possible to look at the
3242     internals of compilation without having to recompile pcre.
3243    
3244    
3245     Version 0.95 23-Sep-97
3246     ----------------------
3247    
3248     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3249     \x20 at the start of a run of normal characters. These were being treated as
3250     real characters, instead of the source characters being re-checked.
3251    
3252    
3253     Version 0.94 18-Sep-97
3254     ----------------------
3255    
3256     1. The functions are now thread-safe, with the caveat that the global variables
3257     containing pointers to malloc() and free() or alternative functions are the
3258     same for all threads.
3259    
3260     2. Get pcre_study() to generate a bitmap of initial characters for non-
3261     anchored patterns when this is possible, and use it if passed to pcre_exec().
3262    
3263    
3264     Version 0.93 15-Sep-97
3265     ----------------------
3266    
3267     1. /(b)|(:+)/ was computing an incorrect first character.
3268    
3269     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3270     but not actually doing anything yet.
3271    
3272     3. Treat "-" characters in classes that cannot be part of ranges as literals,
3273     as Perl does (e.g. [-az] or [az-]).
3274    
3275     4. Set the anchored flag if a branch starts with .* or .*? because that tests
3276     all possible positions.
3277    
3278     5. Split up into different modules to avoid including unneeded functions in a
3279     compiled binary. However, compile and exec are still in one module. The "study"
3280     function is split off.
3281    
3282     6. The character tables are now in a separate module whose source is generated
3283     by an auxiliary program - but can then be edited by hand if required. There are
3284     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3285     toupper() in the code.
3286    
3287     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3288     make them global. Abolish the function for setting them, as the caller can now
3289     set them directly.
3290    
3291    
3292     Version 0.92 11-Sep-97
3293     ----------------------
3294    
3295     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3296     (e.g. /a{1,3}/) was broken (I mis-optimized it).
3297    
3298     2. Caseless matching was not working in character classes if the characters in
3299     the pattern were in upper case.
3300    
3301     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3302    
3303     4. Make PCRE_ANCHORED public and accept as a compile option.
3304    
3305     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3306     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3307     pass them.
3308    
3309     6. Give an error if bad option bits passed at compile or run time.
3310    
3311     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3312     pcretest to cause it to pass that flag.
3313    
3314     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3315     options, and the first character, if set.
3316    
3317     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3318    
3319    
3320     Version 0.91 10-Sep-97
3321     ----------------------
3322    
3323     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3324     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3325    
3326     2. PCRE was looping on encountering an indefinitely repeated back reference to
3327     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3328     Perl does - treats the match as successful.
3329    
3330     ****

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