ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 300 - (hide annotations) (download)
Mon Jan 14 19:43:19 2008 UTC (7 years, 3 months ago) by ph10
File size: 148714 byte(s)
Fix buffer overrun for class with very many characters whose codepoints are 
above 255.

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


Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

ViewVC Help
Powered by ViewVC 1.1.12