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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 285 - (show annotations) (download)
Wed Dec 12 17:03:50 2007 UTC (6 years, 9 months ago) by ph10
File size: 146155 byte(s)
Fix bad compiled code for things like /\pL{2}+/ in which a possessive 
quantifier with a fixed limit was applied to a character property.

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

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