ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 295 - (hide annotations) (download)
Mon Dec 31 17:00:24 2007 UTC (7 years, 3 months ago) by ph10
File size: 148260 byte(s)
Make POSIX character class parsing more like Perl.

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


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

ViewVC Help
Powered by ViewVC 1.1.12