/[pcre]/code/tags/pcre-8.01/ChangeLog
ViewVC logotype

Contents of /code/tags/pcre-8.01/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 248 - (show annotations) (download)
Mon Sep 17 10:09:22 2007 UTC (6 years, 10 months ago) by ph10
Original Path: code/trunk/ChangeLog
File size: 141289 byte(s)
Change condition for the macros for snprintf and strtoll from _WIN32 to 
HAVE_WINDOWS_H because they are needed only when windows.h is present.

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

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12