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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 234 - (show annotations) (download)
Tue Sep 11 11:36:27 2007 UTC (6 years, 10 months ago) by ph10
File size: 140194 byte(s)
Macros for snprintf, strtoll, strtoull when compiling the C++ things under 
Windows.

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

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