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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 286 - (show annotations) (download)
Mon Dec 17 14:46:11 2007 UTC (6 years, 10 months ago) by ph10
File size: 146126 byte(s)
Add .gz and .bz2 optional support to pcregrep.

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 17. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
95
96
97 Version 7.4 21-Sep-07
98 ---------------------
99
100 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
101 means that a class such as [\s] counted as "explicit reference to CR or
102 LF". That isn't really right - the whole point of the change was to try to
103 help when there was an actual mention of one of the two characters. So now
104 the change happens only if \r or \n (or a literal CR or LF) character is
105 encountered.
106
107 2. The 32-bit options word was also used for 6 internal flags, but the numbers
108 of both had grown to the point where there were only 3 bits left.
109 Fortunately, there was spare space in the data structure, and so I have
110 moved the internal flags into a new 16-bit field to free up more option
111 bits.
112
113 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
114 but did not set the internal JCHANGED flag - either of these is enough to
115 control the way the "get" function works - but the PCRE_INFO_JCHANGED
116 facility is supposed to tell if (?J) was ever used, so now (?J) at the
117 start sets both bits.
118
119 4. Added options (at build time, compile time, exec time) to change \R from
120 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
121
122 5. doc/pcresyntax.html was missing from the distribution.
123
124 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
125 compatibility, even though it is no longer used.
126
127 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
128 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
129 windows.h file is present (where different names are used). [This was
130 reversed later after testing - see 16 below.]
131
132 8. Changed all #include <config.h> to #include "config.h". There were also
133 some further <pcre.h> cases that I changed to "pcre.h".
134
135 9. When pcregrep was used with the --colour option, it missed the line ending
136 sequence off the lines that it output.
137
138 10. It was pointed out to me that arrays of string pointers cause lots of
139 relocations when a shared library is dynamically loaded. A technique of
140 using a single long string with a table of offsets can drastically reduce
141 these. I have refactored PCRE in four places to do this. The result is
142 dramatic:
143
144 Originally: 290
145 After changing UCP table: 187
146 After changing error message table: 43
147 After changing table of "verbs" 36
148 After changing table of Posix names 22
149
150 Thanks to the folks working on Gregex for glib for this insight.
151
152 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
153 unicode-properties was also set.
154
155 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
156
157 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
158 checked only for CRLF.
159
160 14. Added casts to pcretest.c to avoid compiler warnings.
161
162 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
163
164 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
165 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
166 entirely. This removes changes made in 7 above.
167
168 17. The CMake files have been updated, and there is now more information about
169 building with CMake in the NON-UNIX-USE document.
170
171
172 Version 7.3 28-Aug-07
173 ---------------------
174
175 1. In the rejigging of the build system that eventually resulted in 7.1, the
176 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
177 brackets there is not right, since it causes compilers to look for an
178 installed pcre.h, not the version that is in the source that is being
179 compiled (which of course may be different). I have changed it back to:
180
181 #include "pcre.h"
182
183 I have a vague recollection that the change was concerned with compiling in
184 different directories, but in the new build system, that is taken care of
185 by the VPATH setting the Makefile.
186
187 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
188 when the subject happened to end in the byte 0x85 (e.g. if the last
189 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
190 characters but of course it shouldn't be taken as a newline when it is part
191 of another character. The bug was that, for an unlimited repeat of . in
192 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
193 characters when looking for a newline.
194
195 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
196
197 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
198 in debug output.
199
200 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
201 long printing in the pcrecpp unittest when running under MinGW.
202
203 6. ESC_K was left out of the EBCDIC table.
204
205 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
206 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
207 limit also applies to "virtual nesting" when a pattern is recursive, and in
208 this case 1000 isn't so big. I have been able to remove this limit at the
209 expense of backing off one optimization in certain circumstances. Normally,
210 when pcre_exec() would call its internal match() function recursively and
211 immediately return the result unconditionally, it uses a "tail recursion"
212 feature to save stack. However, when a subpattern that can match an empty
213 string has an unlimited repetition quantifier, it no longer makes this
214 optimization. That gives it a stack frame in which to save the data for
215 checking that an empty string has been matched. Previously this was taken
216 from the 1000-entry workspace that had been reserved. So now there is no
217 explicit limit, but more stack is used.
218
219 8. Applied Daniel's patches to solve problems with the import/export magic
220 syntax that is required for Windows, and which was going wrong for the
221 pcreposix and pcrecpp parts of the library. These were overlooked when this
222 problem was solved for the main library.
223
224 9. There were some crude static tests to avoid integer overflow when computing
225 the size of patterns that contain repeated groups with explicit upper
226 limits. As the maximum quantifier is 65535, the maximum group length was
227 set at 30,000 so that the product of these two numbers did not overflow a
228 32-bit integer. However, it turns out that people want to use groups that
229 are longer than 30,000 bytes (though not repeat them that many times).
230 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
231 made it possible to implement the integer overflow checks in a much more
232 dynamic way, which I have now done. The artificial limitation on group
233 length has been removed - we now have only the limit on the total length of
234 the compiled pattern, which depends on the LINK_SIZE setting.
235
236 10. Fixed a bug in the documentation for get/copy named substring when
237 duplicate names are permitted. If none of the named substrings are set, the
238 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
239 empty string.
240
241 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
242 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
243 because the ] is interpreted as the first data character and the
244 terminating ] is not found. PCRE has been made compatible with Perl in this
245 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
246 cause memory overwriting.
247
248 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
249 string has been matched (to stop an infinite loop). It was not recognizing
250 a conditional subpattern that could match an empty string if that
251 subpattern was within another subpattern. For example, it looped when
252 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
253 condition was not nested. This bug has been fixed.
254
255 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
256 past the start of the subject in the presence of bytes with the top bit
257 set, for example "\x8aBCD".
258
259 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
260 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
261
262 14. Optimized (?!) to (*FAIL).
263
264 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
265 This restricts code points to be within the range 0 to 0x10FFFF, excluding
266 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
267 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
268 does: it's just the validity check that is more restrictive.
269
270 16. Inserted checks for integer overflows during escape sequence (backslash)
271 processing, and also fixed erroneous offset values for syntax errors during
272 backslash processing.
273
274 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
275 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
276
277 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
278 caused an overrun.
279
280 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
281 something other than just ASCII characters) inside a group that had an
282 unlimited repeat caused a loop at compile time (while checking to see
283 whether the group could match an empty string).
284
285 20. Debugging a pattern containing \p or \P could cause a crash. For example,
286 [\P{Any}] did so. (Error in the code for printing property names.)
287
288 21. An orphan \E inside a character class could cause a crash.
289
290 22. A repeated capturing bracket such as (A)? could cause a wild memory
291 reference during compilation.
292
293 23. There are several functions in pcre_compile() that scan along a compiled
294 expression for various reasons (e.g. to see if it's fixed length for look
295 behind). There were bugs in these functions when a repeated \p or \P was
296 present in the pattern. These operators have additional parameters compared
297 with \d, etc, and these were not being taken into account when moving along
298 the compiled data. Specifically:
299
300 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
301 length.
302
303 (b) An item such as \pL+ within a repeated group could cause crashes or
304 loops.
305
306 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
307 "reference to non-existent subpattern" error.
308
309 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
310
311 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
312 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
313
314 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
315
316 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
317 character were causing crashes (broken optimization).
318
319 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
320 \p or \P) caused a compile-time loop.
321
322 28. More problems have arisen in unanchored patterns when CRLF is a valid line
323 break. For example, the unstudied pattern [\r\n]A does not match the string
324 "\r\nA" because change 7.0/46 below moves the current point on by two
325 characters after failing to match at the start. However, the pattern \nA
326 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
327 the same is true. There doesn't seem any very clean way out of this, but
328 what I have chosen to do makes the common cases work: PCRE now takes note
329 of whether there can be an explicit match for \r or \n anywhere in the
330 pattern, and if so, 7.0/46 no longer applies. As part of this change,
331 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
332 pattern has explicit CR or LF references.
333
334 29. Added (*CR) etc for changing newline setting at start of pattern.
335
336
337 Version 7.2 19-Jun-07
338 ---------------------
339
340 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
341 which is apparently normally available under Windows.
342
343 2. Re-jig the pcregrep tests with different newline settings in an attempt
344 to make them independent of the local environment's newline setting.
345
346 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
347
348 4. Some of the "internals" tests were previously cut out when the link size
349 was not 2, because the output contained actual offsets. The recent new
350 "Z" feature of pcretest means that these can be cut out, making the tests
351 usable with all link sizes.
352
353 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
354 stack recursion. This gives a massive performance boost under BSD, but just
355 a small improvement under Linux. However, it saves one field in the frame
356 in all cases.
357
358 6. Added more features from the forthcoming Perl 5.10:
359
360 (a) (?-n) (where n is a string of digits) is a relative subroutine or
361 recursion call. It refers to the nth most recently opened parentheses.
362
363 (b) (?+n) is also a relative subroutine call; it refers to the nth next
364 to be opened parentheses.
365
366 (c) Conditions that refer to capturing parentheses can be specified
367 relatively, for example, (?(-2)... or (?(+3)...
368
369 (d) \K resets the start of the current match so that everything before
370 is not part of it.
371
372 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
373
374 (f) \g{name} is another synonym - part of Perl 5.10's unification of
375 reference syntax.
376
377 (g) (?| introduces a group in which the numbering of parentheses in each
378 alternative starts with the same number.
379
380 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
381
382 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
383 PCRE_INFO_JCHANGED.
384
385 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
386 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
387 for detecting groups that can match an empty string.
388
389 9. A pattern with a very large number of alternatives (more than several
390 hundred) was running out of internal workspace during the pre-compile
391 phase, where pcre_compile() figures out how much memory will be needed. A
392 bit of new cunning has reduced the workspace needed for groups with
393 alternatives. The 1000-alternative test pattern now uses 12 bytes of
394 workspace instead of running out of the 4096 that are available.
395
396 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
397
398 11. Applied patch from Google to remove an optimization that didn't quite work.
399 The report of the bug said:
400
401 pcrecpp::RE("a*").FullMatch("aaa") matches, while
402 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
403 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
404
405 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
406 it matched the wrong number of bytes.
407
408
409 Version 7.1 24-Apr-07
410 ---------------------
411
412 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
413 that is more "standard", making use of automake and other Autotools. There
414 is some re-arrangement of the files and adjustment of comments consequent
415 on this.
416
417 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
418 for recursive directory scanning broke on some systems because the files
419 are not scanned in any specific order and on different systems the order
420 was different. A call to "sort" has been inserted into RunGrepTest for the
421 approprate test as a short-term fix. In the longer term there may be an
422 alternative.
423
424 3. I had an email from Eric Raymond about problems translating some of PCRE's
425 man pages to HTML (despite the fact that I distribute HTML pages, some
426 people do their own conversions for various reasons). The problems
427 concerned the use of low-level troff macros .br and .in. I have therefore
428 removed all such uses from the man pages (some were redundant, some could
429 be replaced by .nf/.fi pairs). The 132html script that I use to generate
430 HTML has been updated to handle .nf/.fi and to complain if it encounters
431 .br or .in.
432
433 4. Updated comments in configure.ac that get placed in config.h.in and also
434 arranged for config.h to be included in the distribution, with the name
435 config.h.generic, for the benefit of those who have to compile without
436 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
437
438 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
439 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
440 makevp.bat for the current PCRE, using the additional files
441 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
442
443 6. A Windows user reported a minor discrepancy with test 2, which turned out
444 to be caused by a trailing space on an input line that had got lost in his
445 copy. The trailing space was an accident, so I've just removed it.
446
447 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
448 that is needed.
449
450 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
451 as "const" (a) because they are and (b) because it helps the PHP
452 maintainers who have recently made a script to detect big data structures
453 in the php code that should be moved to the .rodata section. I remembered
454 to update Builducptable as well, so it won't revert if ucptable.h is ever
455 re-created.
456
457 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
458 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
459 order to be able to cut out the UTF-8 tables in the latter when UTF-8
460 support is not required. This saves 1.5-2K of code, which is important in
461 some applications.
462
463 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
464 so as not to refer to the tables, even though these functions will never be
465 called when UTF-8 support is disabled. Otherwise there are problems with a
466 shared library.
467
468 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
469
470 (a) It was defining its arguments as char * instead of void *.
471
472 (b) It was assuming that all moves were upwards in memory; this was true
473 a long time ago when I wrote it, but is no longer the case.
474
475 The emulated memove() is provided for those environments that have neither
476 memmove() nor bcopy(). I didn't think anyone used it these days, but that
477 is clearly not the case, as these two bugs were recently reported.
478
479 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
480 and Detrail to create the HTML documentation, the .txt form of the man
481 pages, and it removes trailing spaces from listed files. It also creates
482 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
483 case, it wraps all the #defines with #ifndefs. This script should be run
484 before "make dist".
485
486 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
487 with Unicode property support.
488
489 (a) For a maximizing quantifier, if the two different cases of the
490 character were of different lengths in their UTF-8 codings (there are
491 some cases like this - I found 11), and the matching function had to
492 back up over a mixture of the two cases, it incorrectly assumed they
493 were both the same length.
494
495 (b) When PCRE was configured to use the heap rather than the stack for
496 recursion during matching, it was not correctly preserving the data for
497 the other case of a UTF-8 character when checking ahead for a match
498 while processing a minimizing repeat. If the check also involved
499 matching a wide character, but failed, corruption could cause an
500 erroneous result when trying to check for a repeat of the original
501 character.
502
503 13. Some tidying changes to the testing mechanism:
504
505 (a) The RunTest script now detects the internal link size and whether there
506 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
507 values substituted by "configure". (The RunGrepTest script already did
508 this for UTF-8.) The configure.ac script no longer substitutes the
509 relevant variables.
510
511 (b) The debugging options /B and /D in pcretest show the compiled bytecode
512 with length and offset values. This means that the output is different
513 for different internal link sizes. Test 2 is skipped for link sizes
514 other than 2 because of this, bypassing the problem. Unfortunately,
515 there was also a test in test 3 (the locale tests) that used /B and
516 failed for link sizes other than 2. Rather than cut the whole test out,
517 I have added a new /Z option to pcretest that replaces the length and
518 offset values with spaces. This is now used to make test 3 independent
519 of link size. (Test 2 will be tidied up later.)
520
521 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
522 segmentation fault instead of returning the appropriate error message.
523
524 15. In multiline mode when the newline sequence was set to "any", the pattern
525 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
526 This doesn't seem right; it now treats the CRLF combination as the line
527 ending, and so does not match in that case. It's only a pattern such as ^$
528 that would hit this one: something like ^ABC$ would have failed after \r
529 and then tried again after \r\n.
530
531 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
532 in an attempt to make files that differ only in their line terminators
533 compare equal. This works on Linux.
534
535 17. Under certain error circumstances pcregrep might try to free random memory
536 as it exited. This is now fixed, thanks to valgrind.
537
538 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
539 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
540 was because its rules for how to advance for /g after matching an empty
541 string at the end of a line did not allow for this case. They now check for
542 it specially.
543
544 20. pcretest is supposed to handle patterns and data of any length, by
545 extending its buffers when necessary. It was getting this wrong when the
546 buffer for a data line had to be extended.
547
548 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
549 CRLF as a newline sequence.
550
551 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
552 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
553 I have nevertheless tidied it up.
554
555 23. Added some casts to kill warnings from HP-UX ia64 compiler.
556
557 24. Added a man page for pcre-config.
558
559
560 Version 7.0 19-Dec-06
561 ---------------------
562
563 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
564 moving to gcc 4.1.1.
565
566 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
567 sys/time.h, which is documented as needed for this function. It doesn't
568 seem to matter on Linux, but it showed up on some releases of OS X.
569
570 3. It seems that there are systems where bytes whose values are greater than
571 127 match isprint() in the "C" locale. The "C" locale should be the
572 default when a C program starts up. In most systems, only ASCII printing
573 characters match isprint(). This difference caused the output from pcretest
574 to vary, making some of the tests fail. I have changed pcretest so that:
575
576 (a) When it is outputting text in the compiled version of a pattern, bytes
577 other than 32-126 are always shown as hex escapes.
578
579 (b) When it is outputting text that is a matched part of a subject string,
580 it does the same, unless a different locale has been set for the match
581 (using the /L modifier). In this case, it uses isprint() to decide.
582
583 4. Fixed a major bug that caused incorrect computation of the amount of memory
584 required for a compiled pattern when options that changed within the
585 pattern affected the logic of the preliminary scan that determines the
586 length. The relevant options are -x, and -i in UTF-8 mode. The result was
587 that the computed length was too small. The symptoms of this bug were
588 either the PCRE error "internal error: code overflow" from pcre_compile(),
589 or a glibc crash with a message such as "pcretest: free(): invalid next
590 size (fast)". Examples of patterns that provoked this bug (shown in
591 pcretest format) are:
592
593 /(?-x: )/x
594 /(?x)(?-x: \s*#\s*)/
595 /((?i)[\x{c0}])/8
596 /(?i:[\x{c0}])/8
597
598 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
599 is now done differently.
600
601 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
602 wrapper classes; (b) implement a new function in the C++ scanner that is
603 more efficient than the old way of doing things because it avoids levels of
604 recursion in the regex matching; (c) add a paragraph to the documentation
605 for the FullMatch() function.
606
607 6. The escape sequence \n was being treated as whatever was defined as
608 "newline". Not only was this contrary to the documentation, which states
609 that \n is character 10 (hex 0A), but it also went horribly wrong when
610 "newline" was defined as CRLF. This has been fixed.
611
612 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
613 was being set to -1 for the "end of line" case (supposedly a value that no
614 character can have). Though this value is never used (the check for end of
615 line is "zero bytes in current character"), it caused compiler complaints.
616 I've changed it to 0xffffffff.
617
618 8. In pcre_version.c, the version string was being built by a sequence of
619 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
620 string (as it is for production releases) called a macro with an empty
621 argument. The C standard says the result of this is undefined. The gcc
622 compiler treats it as an empty string (which was what was wanted) but it is
623 reported that Visual C gives an error. The source has been hacked around to
624 avoid this problem.
625
626 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
627 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
628 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
629 of them did).
630
631 10. Originally, pcretest opened its input and output without "b"; then I was
632 told that "b" was needed in some environments, so it was added for release
633 5.0 to both the input and output. (It makes no difference on Unix-like
634 systems.) Later I was told that it is wrong for the input on Windows. I've
635 now abstracted the modes into two macros, to make it easier to fiddle with
636 them, and removed "b" from the input mode under Windows.
637
638 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
639
640 12. Added -help and --help to pcretest as an official way of being reminded
641 of the options.
642
643 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
644 and pcrecpp.cc because they annoy compilers at high warning levels.
645
646 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
647
648 15. Fixed an occurrence of == in configure.ac that should have been = (shell
649 scripts are not C programs :-) and which was not noticed because it works
650 on Linux.
651
652 16. pcretest is supposed to handle any length of pattern and data line (as one
653 line or as a continued sequence of lines) by extending its input buffer if
654 necessary. This feature was broken for very long pattern lines, leading to
655 a string of junk being passed to pcre_compile() if the pattern was longer
656 than about 50K.
657
658 17. I have done a major re-factoring of the way pcre_compile() computes the
659 amount of memory needed for a compiled pattern. Previously, there was code
660 that made a preliminary scan of the pattern in order to do this. That was
661 OK when PCRE was new, but as the facilities have expanded, it has become
662 harder and harder to keep it in step with the real compile phase, and there
663 have been a number of bugs (see for example, 4 above). I have now found a
664 cunning way of running the real compile function in a "fake" mode that
665 enables it to compute how much memory it would need, while actually only
666 ever using a few hundred bytes of working memory and without too many
667 tests of the mode. This should make future maintenance and development
668 easier. A side effect of this work is that the limit of 200 on the nesting
669 depth of parentheses has been removed (though this was never a serious
670 limitation, I suspect). However, there is a downside: pcre_compile() now
671 runs more slowly than before (30% or more, depending on the pattern). I
672 hope this isn't a big issue. There is no effect on runtime performance.
673
674 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
675 newline (only possible for the last line of a file) and it was a
676 pattern that set a locale (followed by /Lsomething), pcretest crashed.
677
678 19. Added additional timing features to pcretest. (1) The -tm option now times
679 matching only, not compiling. (2) Both -t and -tm can be followed, as a
680 separate command line item, by a number that specifies the number of
681 repeats to use when timing. The default is 50000; this gives better
682 precision, but takes uncomfortably long for very large patterns.
683
684 20. Extended pcre_study() to be more clever in cases where a branch of a
685 subpattern has no definite first character. For example, (a*|b*)[cd] would
686 previously give no result from pcre_study(). Now it recognizes that the
687 first character must be a, b, c, or d.
688
689 21. There was an incorrect error "recursive call could loop indefinitely" if
690 a subpattern (or the entire pattern) that was being tested for matching an
691 empty string contained only one non-empty item after a nested subpattern.
692 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
693 incorrectly, because the \d was being skipped in the check.
694
695 22. The pcretest program now has a new pattern option /B and a command line
696 option -b, which is equivalent to adding /B to every pattern. This causes
697 it to show the compiled bytecode, without the additional information that
698 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
699 is the same as /B/I).
700
701 23. A new optimization is now able automatically to treat some sequences such
702 as a*b as a*+b. More specifically, if something simple (such as a character
703 or a simple class like \d) has an unlimited quantifier, and is followed by
704 something that cannot possibly match the quantified thing, the quantifier
705 is automatically "possessified".
706
707 24. A recursive reference to a subpattern whose number was greater than 39
708 went wrong under certain circumstances in UTF-8 mode. This bug could also
709 have affected the operation of pcre_study().
710
711 25. Realized that a little bit of performance could be had by replacing
712 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
713
714 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
715
716 27. Possessive quantifiers such as a++ were previously implemented by turning
717 them into atomic groups such as ($>a+). Now they have their own opcodes,
718 which improves performance. This includes the automatically created ones
719 from 23 above.
720
721 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
722 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
723 the first matched character to be a colon. This applied both to named and
724 numbered groups.
725
726 29. The ucpinternal.h header file was missing its idempotency #ifdef.
727
728 30. I was sent a "project" file called libpcre.a.dev which I understand makes
729 building PCRE on Windows easier, so I have included it in the distribution.
730
731 31. There is now a check in pcretest against a ridiculously large number being
732 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
733 loop, the loop is abandoned.
734
735 32. Forward references to subpatterns in conditions such as (?(2)...) where
736 subpattern 2 is defined later cause pcre_compile() to search forwards in
737 the pattern for the relevant set of parentheses. This search went wrong
738 when there were unescaped parentheses in a character class, parentheses
739 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
740
741 33. "Subroutine" calls and backreferences were previously restricted to
742 referencing subpatterns earlier in the regex. This restriction has now
743 been removed.
744
745 34. Added a number of extra features that are going to be in Perl 5.10. On the
746 whole, these are just syntactic alternatives for features that PCRE had
747 previously implemented using the Python syntax or my own invention. The
748 other formats are all retained for compatibility.
749
750 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
751 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
752 also .NET compatible.
753
754 (b) A recursion or subroutine call to a named group can now be defined as
755 (?&name) as well as (?P>name).
756
757 (c) A backreference to a named group can now be defined as \k<name> or
758 \k'name' as well as (?P=name). The new forms, as well as being in Perl
759 5.10, are also .NET compatible.
760
761 (d) A conditional reference to a named group can now use the syntax
762 (?(<name>) or (?('name') as well as (?(name).
763
764 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
765 groups (named and numbered) that are never evaluated inline, but can be
766 called as "subroutines" from elsewhere. In effect, the DEFINE condition
767 is always false. There may be only one alternative in such a group.
768
769 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
770 as the simple (?(R). The condition is true only if the most recent
771 recursion is that of the given number or name. It does not search out
772 through the entire recursion stack.
773
774 (g) The escape \gN or \g{N} has been added, where N is a positive or
775 negative number, specifying an absolute or relative reference.
776
777 35. Tidied to get rid of some further signed/unsigned compiler warnings and
778 some "unreachable code" warnings.
779
780 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
781 things, this adds five new scripts.
782
783 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
784 There were also incompatibilities regarding the handling of \Q..\E inside
785 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
786 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
787
788 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
789 matches an empty string, and forcibly breaks the loop. There were bugs in
790 this code in non-simple cases. For a pattern such as ^(a()*)* matched
791 against aaaa the result was just "a" rather than "aaaa", for example. Two
792 separate and independent bugs (that affected different cases) have been
793 fixed.
794
795 39. Refactored the code to abolish the use of different opcodes for small
796 capturing bracket numbers. This is a tidy that I avoided doing when I
797 removed the limit on the number of capturing brackets for 3.5 back in 2001.
798 The new approach is not only tidier, it makes it possible to reduce the
799 memory needed to fix the previous bug (38).
800
801 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
802 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
803 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
804 mode.
805
806 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
807 report.
808
809 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
810 copy construction and assignment in the C++ wrapper.
811
812 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
813 couple of bugs that could have given wrong results in the "--newline=crlf"
814 case.
815
816 44. Added a number of casts and did some reorganization of signed/unsigned int
817 variables following suggestions from Dair Grant. Also renamed the variable
818 "this" as "item" because it is a C++ keyword.
819
820 45. Arranged for dftables to add
821
822 #include "pcre_internal.h"
823
824 to pcre_chartables.c because without it, gcc 4.x may remove the array
825 definition from the final binary if PCRE is built into a static library and
826 dead code stripping is activated.
827
828 46. For an unanchored pattern, if a match attempt fails at the start of a
829 newline sequence, and the newline setting is CRLF or ANY, and the next two
830 characters are CRLF, advance by two characters instead of one.
831
832
833 Version 6.7 04-Jul-06
834 ---------------------
835
836 1. In order to handle tests when input lines are enormously long, pcretest has
837 been re-factored so that it automatically extends its buffers when
838 necessary. The code is crude, but this _is_ just a test program. The
839 default size has been increased from 32K to 50K.
840
841 2. The code in pcre_study() was using the value of the re argument before
842 testing it for NULL. (Of course, in any sensible call of the function, it
843 won't be NULL.)
844
845 3. The memmove() emulation function in pcre_internal.h, which is used on
846 systems that lack both memmove() and bcopy() - that is, hardly ever -
847 was missing a "static" storage class specifier.
848
849 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
850 containing an extended class (one that cannot be represented by a bitmap
851 because it contains high-valued characters or Unicode property items, e.g.
852 [\pZ]). Almost always one would set UTF-8 mode when processing such a
853 pattern, but PCRE should not loop if you do not (it no longer does).
854 [Detail: two cases were found: (a) a repeated subpattern containing an
855 extended class; (b) a recursive reference to a subpattern that followed a
856 previous extended class. It wasn't skipping over the extended class
857 correctly when UTF-8 mode was not set.]
858
859 5. A negated single-character class was not being recognized as fixed-length
860 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
861 compile error "lookbehind assertion is not fixed length".
862
863 6. The RunPerlTest auxiliary script was showing an unexpected difference
864 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
865 write a Perl script that can interpret lines of an input file either as
866 byte characters or as UTF-8, which is what "perltest" was being required to
867 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
868 can't do is switch easily at run time between having the "use utf8;" pragma
869 or not. In the end, I fudged it by using the RunPerlTest script to insert
870 "use utf8;" explicitly for the UTF-8 tests.
871
872 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
873 the end of the subject string, contrary to the documentation and to what
874 Perl does. This was true of both matching functions. Now it matches only at
875 the start of the subject and immediately after *internal* newlines.
876
877 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
878 a pointer to an int instead of a pointer to an unsigned long int. This
879 caused problems on 64-bit systems.
880
881 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
882 instance of the 'standard' template library not being so standard".
883
884 10. There was no check on the number of named subpatterns nor the maximum
885 length of a subpattern name. The product of these values is used to compute
886 the size of the memory block for a compiled pattern. By supplying a very
887 long subpattern name and a large number of named subpatterns, the size
888 computation could be caused to overflow. This is now prevented by limiting
889 the length of names to 32 characters, and the number of named subpatterns
890 to 10,000.
891
892 11. Subpatterns that are repeated with specific counts have to be replicated in
893 the compiled pattern. The size of memory for this was computed from the
894 length of the subpattern and the repeat count. The latter is limited to
895 65535, but there was no limit on the former, meaning that integer overflow
896 could in principle occur. The compiled length of a repeated subpattern is
897 now limited to 30,000 bytes in order to prevent this.
898
899 12. Added the optional facility to have named substrings with the same name.
900
901 13. Added the ability to use a named substring as a condition, using the
902 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
903 are numbers (not recommended). Forward references are permitted.
904
905 14. Added forward references in named backreferences (if you see what I mean).
906
907 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
908 pattern could run off the end of the subject. For example, the pattern
909 "(?s)(.{1,5})"8 did this with the subject "ab".
910
911 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
912 PCRE_CASELESS was set when matching characters that were quantified with ?
913 or *.
914
915 17. A character class other than a single negated character that had a minimum
916 but no maximum quantifier - for example [ab]{6,} - was not handled
917 correctly by pce_dfa_exec(). It would match only one character.
918
919 18. A valid (though odd) pattern that looked like a POSIX character
920 class but used an invalid character after [ (for example [[,abc,]]) caused
921 pcre_compile() to give the error "Failed: internal error: code overflow" or
922 in some cases to crash with a glibc free() error. This could even happen if
923 the pattern terminated after [[ but there just happened to be a sequence of
924 letters, a binary zero, and a closing ] in the memory that followed.
925
926 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
927 over the years. Originally (before any Unicode support), just the bottom 8
928 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
929 output from "man perlunicode" includes this:
930
931 The regular expression compiler produces polymorphic opcodes. That
932 is, the pattern adapts to the data and automatically switches to
933 the Unicode character scheme when presented with Unicode data--or
934 instead uses a traditional byte scheme when presented with byte
935 data.
936
937 Sadly, a wide octal escape does not cause a switch, and in a string with
938 no other multibyte characters, these octal escapes are treated as before.
939 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
940 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
941 Unicode string.
942
943 I have not perpetrated such confusion in PCRE. Up till now, it took just
944 the bottom 8 bits, as in old Perl. I have now made octal escapes with
945 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
946 translate to the appropriate multibyte character.
947
948 29. Applied some refactoring to reduce the number of warnings from Microsoft
949 and Borland compilers. This has included removing the fudge introduced
950 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
951 a warning about an unused variable.
952
953 21. PCRE has not included VT (character 0x0b) in the set of whitespace
954 characters since release 4.0, because Perl (from release 5.004) does not.
955 [Or at least, is documented not to: some releases seem to be in conflict
956 with the documentation.] However, when a pattern was studied with
957 pcre_study() and all its branches started with \s, PCRE still included VT
958 as a possible starting character. Of course, this did no harm; it just
959 caused an unnecessary match attempt.
960
961 22. Removed a now-redundant internal flag bit that recorded the fact that case
962 dependency changed within the pattern. This was once needed for "required
963 byte" processing, but is no longer used. This recovers a now-scarce options
964 bit. Also moved the least significant internal flag bit to the most-
965 significant bit of the word, which was not previously used (hangover from
966 the days when it was an int rather than a uint) to free up another bit for
967 the future.
968
969 23. Added support for CRLF line endings as well as CR and LF. As well as the
970 default being selectable at build time, it can now be changed at runtime
971 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
972 specify that it is scanning data with non-default line endings.
973
974 24. Changed the definition of CXXLINK to make it agree with the definition of
975 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
976
977 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
978 recursions. This makes a big different to stack usage for some patterns.
979
980 26. If a subpattern containing a named recursion or subroutine reference such
981 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
982 the space required for the compiled pattern went wrong and gave too small a
983 value. Depending on the environment, this could lead to "Failed: internal
984 error: code overflow at offset 49" or "glibc detected double free or
985 corruption" errors.
986
987 27. Applied patches from Google (a) to support the new newline modes and (b) to
988 advance over multibyte UTF-8 characters in GlobalReplace.
989
990 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
991 difference for some implementation of PCRE in some Windows version.
992
993 29. Added some extra testing facilities to pcretest:
994
995 \q<number> in a data line sets the "match limit" value
996 \Q<number> in a data line sets the "match recursion limt" value
997 -S <number> sets the stack size, where <number> is in megabytes
998
999 The -S option isn't available for Windows.
1000
1001
1002 Version 6.6 06-Feb-06
1003 ---------------------
1004
1005 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1006 in pcreposix.h. I have copied the definition from pcre.h.
1007
1008 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1009 because pcre.h is no longer a built file.
1010
1011 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1012 not normally included in the compiled code.
1013
1014
1015 Version 6.5 01-Feb-06
1016 ---------------------
1017
1018 1. When using the partial match feature with pcre_dfa_exec(), it was not
1019 anchoring the second and subsequent partial matches at the new starting
1020 point. This could lead to incorrect results. For example, with the pattern
1021 /1234/, partially matching against "123" and then "a4" gave a match.
1022
1023 2. Changes to pcregrep:
1024
1025 (a) All non-match returns from pcre_exec() were being treated as failures
1026 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1027 error message is output. Some extra information is given for the
1028 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1029 probably the only errors that are likely to be caused by users (by
1030 specifying a regex that has nested indefinite repeats, for instance).
1031 If there are more than 20 of these errors, pcregrep is abandoned.
1032
1033 (b) A binary zero was treated as data while matching, but terminated the
1034 output line if it was written out. This has been fixed: binary zeroes
1035 are now no different to any other data bytes.
1036
1037 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1038 used to set a locale for matching. The --locale=xxxx long option has
1039 been added (no short equivalent) to specify a locale explicitly on the
1040 pcregrep command, overriding the environment variables.
1041
1042 (d) When -B was used with -n, some line numbers in the output were one less
1043 than they should have been.
1044
1045 (e) Added the -o (--only-matching) option.
1046
1047 (f) If -A or -C was used with -c (count only), some lines of context were
1048 accidentally printed for the final match.
1049
1050 (g) Added the -H (--with-filename) option.
1051
1052 (h) The combination of options -rh failed to suppress file names for files
1053 that were found from directory arguments.
1054
1055 (i) Added the -D (--devices) and -d (--directories) options.
1056
1057 (j) Added the -F (--fixed-strings) option.
1058
1059 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1060
1061 (l) Added the --colo(u)r option.
1062
1063 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1064 is not present by default.
1065
1066 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1067 items such as (?R) or (?1), when the recursion could match a number of
1068 alternatives. If it matched one of the alternatives, but subsequently,
1069 outside the recursion, there was a failure, the code tried to back up into
1070 the recursion. However, because of the way PCRE is implemented, this is not
1071 possible, and the result was an incorrect result from the match.
1072
1073 In order to prevent this happening, the specification of recursion has
1074 been changed so that all such subpatterns are automatically treated as
1075 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1076
1077 4. I had overlooked the fact that, in some locales, there are characters for
1078 which isalpha() is true but neither isupper() nor islower() are true. In
1079 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1080 and ordfeminine) are like this. This affected the treatment of \w and \W
1081 when they appeared in character classes, but not when they appeared outside
1082 a character class. The bit map for "word" characters is now created
1083 separately from the results of isalnum() instead of just taking it from the
1084 upper, lower, and digit maps. (Plus the underscore character, of course.)
1085
1086 5. The above bug also affected the handling of POSIX character classes such as
1087 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1088 permanent tables. Instead, the bit maps for such a class were previously
1089 created as the appropriate unions of the upper, lower, and digit bitmaps.
1090 Now they are created by subtraction from the [[:word:]] class, which has
1091 its own bitmap.
1092
1093 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1094 It is created by subtracting the vertical space characters (\x09, \x0a,
1095 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1096 subtraction was done in the overall bitmap for a character class, meaning
1097 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1098 be recognized. This bug has been fixed.
1099
1100 7. Patches from the folks at Google:
1101
1102 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1103 real life, but is still worth protecting against".
1104
1105 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1106 regular expressions".
1107
1108 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1109 have it.
1110
1111 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1112 "configure" and the latter not, in order to fix a problem somebody had
1113 with compiling the Arg class on HP-UX.
1114
1115 (e) Improve the error-handling of the C++ wrapper a little bit.
1116
1117 (f) New tests for checking recursion limiting.
1118
1119 8. The pcre_memmove() function, which is used only if the environment does not
1120 have a standard memmove() function (and is therefore rarely compiled),
1121 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1122 returning a result (though PCRE never actually uses the result).
1123
1124 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1125 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1126 returned instead of calling malloc() with an overflowing number that would
1127 most likely cause subsequent chaos.
1128
1129 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1130
1131 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1132 with this option is matched, the nmatch and pmatch options of regexec() are
1133 ignored.
1134
1135 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1136 provided in case anyone wants to the the POSIX interface with UTF-8
1137 strings.
1138
1139 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1140 C++ linking (needed for some HP-UX environments).
1141
1142 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1143 (unused parameter) and in the pcre_printint() function (omitted "default"
1144 switch label when the default is to do nothing).
1145
1146 15. Added some code to make it possible, when PCRE is compiled as a C++
1147 library, to replace subject pointers for pcre_exec() with a smart pointer
1148 class, thus making it possible to process discontinuous strings.
1149
1150 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1151 much the same function. They were added by different people who were trying
1152 to make PCRE easy to compile on non-Unix systems. It has been suggested
1153 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1154 for compiling on Windows systems. I have therefore replaced it with
1155 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1156 defaults to "extern" for C or "extern C" for C++, which works fine on
1157 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1158 SCOPE with something explicit in config.h. In addition:
1159
1160 (a) pcreposix.h still had just "extern" instead of either of these macros;
1161 I have replaced it with PCRE_DATA_SCOPE.
1162
1163 (b) Functions such as _pcre_xclass(), which are internal to the library,
1164 but external in the C sense, all had PCRE_EXPORT in their definitions.
1165 This is apparently wrong for the Windows case, so I have removed it.
1166 (It makes no difference on Unix-like systems.)
1167
1168 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1169 of recursive calls to match(). This is different to MATCH_LIMIT because
1170 that limits the total number of calls to match(), not all of which increase
1171 the depth of recursion. Limiting the recursion depth limits the amount of
1172 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1173 when PCRE is compiled, and changed at run time. A patch from Google adds
1174 this functionality to the C++ interface.
1175
1176 18. Changes to the handling of Unicode character properties:
1177
1178 (a) Updated the table to Unicode 4.1.0.
1179
1180 (b) Recognize characters that are not in the table as "Cn" (undefined).
1181
1182 (c) I revised the way the table is implemented to a much improved format
1183 which includes recognition of ranges. It now supports the ranges that
1184 are defined in UnicodeData.txt, and it also amalgamates other
1185 characters into ranges. This has reduced the number of entries in the
1186 table from around 16,000 to around 3,000, thus reducing its size
1187 considerably. I realized I did not need to use a tree structure after
1188 all - a binary chop search is just as efficient. Having reduced the
1189 number of entries, I extended their size from 6 bytes to 8 bytes to
1190 allow for more data.
1191
1192 (d) Added support for Unicode script names via properties such as \p{Han}.
1193
1194 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1195 matching that character.
1196
1197 20. When matching a repeated Unicode property with a minimum greater than zero,
1198 (for example \pL{2,}), PCRE could look past the end of the subject if it
1199 reached it while seeking the minimum number of characters. This could
1200 happen only if some of the characters were more than one byte long, because
1201 there is a check for at least the minimum number of bytes.
1202
1203 21. Refactored the implementation of \p and \P so as to be more general, to
1204 allow for more different types of property in future. This has changed the
1205 compiled form incompatibly. Anybody with saved compiled patterns that use
1206 \p or \P will have to recompile them.
1207
1208 22. Added "Any" and "L&" to the supported property types.
1209
1210 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1211 but give a compile time error if the value is greater than 0xff.
1212
1213 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1214 accidentally not being installed or uninstalled.
1215
1216 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1217 made were to insert the current release number. This seemed silly, because
1218 it made things harder for people building PCRE on systems that don't run
1219 "configure". I have turned pcre.h into a distributed file, no longer built
1220 by "configure", with the version identification directly included. There is
1221 no longer a pcre.h.in file.
1222
1223 However, this change necessitated a change to the pcre-config script as
1224 well. It is built from pcre-config.in, and one of the substitutions was the
1225 release number. I have updated configure.ac so that ./configure now finds
1226 the release number by grepping pcre.h.
1227
1228 26. Added the ability to run the tests under valgrind.
1229
1230
1231 Version 6.4 05-Sep-05
1232 ---------------------
1233
1234 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1235 "--" to be printed when multiple files were scanned, even when none of the
1236 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1237 consider it to be a bug, and have restored the previous behaviour.
1238
1239 2. A couple of code tidies to get rid of compiler warnings.
1240
1241 3. The pcretest program used to cheat by referring to symbols in the library
1242 whose names begin with _pcre_. These are internal symbols that are not
1243 really supposed to be visible externally, and in some environments it is
1244 possible to suppress them. The cheating is now confined to including
1245 certain files from the library's source, which is a bit cleaner.
1246
1247 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1248 file's purpose clearer.
1249
1250 5. Reorganized pcre_ucp_findchar().
1251
1252
1253 Version 6.3 15-Aug-05
1254 ---------------------
1255
1256 1. The file libpcre.pc.in did not have general read permission in the tarball.
1257
1258 2. There were some problems when building without C++ support:
1259
1260 (a) If C++ support was not built, "make install" and "make test" still
1261 tried to test it.
1262
1263 (b) There were problems when the value of CXX was explicitly set. Some
1264 changes have been made to try to fix these, and ...
1265
1266 (c) --disable-cpp can now be used to explicitly disable C++ support.
1267
1268 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1269 backslash in a target when C++ was disabled. This confuses some
1270 versions of "make", apparently. Using an intermediate variable solves
1271 this. (Same for CPP_LOBJ.)
1272
1273 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1274 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1275 necessary on certain architectures.
1276
1277 4. Added a setting of -export-symbols-regex to the link command to remove
1278 those symbols that are exported in the C sense, but actually are local
1279 within the library, and not documented. Their names all begin with
1280 "_pcre_". This is not a perfect job, because (a) we have to except some
1281 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1282 available (and never for static libraries). I have made a note to try to
1283 find a way round (a) in the future.
1284
1285
1286 Version 6.2 01-Aug-05
1287 ---------------------
1288
1289 1. There was no test for integer overflow of quantifier values. A construction
1290 such as {1111111111111111} would give undefined results. What is worse, if
1291 a minimum quantifier for a parenthesized subpattern overflowed and became
1292 negative, the calculation of the memory size went wrong. This could have
1293 led to memory overwriting.
1294
1295 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1296
1297 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1298 operating environments where this matters.
1299
1300 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1301 PCRE options from within the C++ wrapper.
1302
1303 5. Named capturing subpatterns were not being correctly counted when a pattern
1304 was compiled. This caused two problems: (a) If there were more than 100
1305 such subpatterns, the calculation of the memory needed for the whole
1306 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1307 back references of the form \12, where the number was greater than 9, were
1308 not recognized as back references, even though there were sufficient
1309 previous subpatterns.
1310
1311 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1312 versions of gcc, e.g. 2.95.4.
1313
1314
1315 Version 6.1 21-Jun-05
1316 ---------------------
1317
1318 1. There was one reference to the variable "posix" in pcretest.c that was not
1319 surrounded by "#if !defined NOPOSIX".
1320
1321 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1322 the cross-check on the old pcre_info() function, for the benefit of the
1323 cut-down version of PCRE that is currently imported into Exim.
1324
1325 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1326 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1327 patterns that start (?i) but I don't think that matters. The use of (?i) is
1328 just an example; this all applies to the other options as well.
1329
1330 4. Since libtool seems to echo the compile commands it is issuing, the output
1331 from "make" can be reduced a bit by putting "@" in front of each libtool
1332 compile command.
1333
1334 5. Patch from the folks at Google for configure.in to be a bit more thorough
1335 in checking for a suitable C++ installation before trying to compile the
1336 C++ stuff. This should fix a reported problem when a compiler was present,
1337 but no suitable headers.
1338
1339 6. The man pages all had just "PCRE" as their title. I have changed them to
1340 be the relevant file name. I have also arranged that these names are
1341 retained in the file doc/pcre.txt, which is a concatenation in text format
1342 of all the man pages except the little individual ones for each function.
1343
1344 7. The NON-UNIX-USE file had not been updated for the different set of source
1345 files that come with release 6. I also added a few comments about the C++
1346 wrapper.
1347
1348
1349 Version 6.0 07-Jun-05
1350 ---------------------
1351
1352 1. Some minor internal re-organization to help with my DFA experiments.
1353
1354 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1355 didn't matter for the library itself when fully configured, but did matter
1356 when compiling without UCP support, or within Exim, where the ucp files are
1357 not imported.
1358
1359 3. Refactoring of the library code to split up the various functions into
1360 different source modules. The addition of the new DFA matching code (see
1361 below) to a single monolithic source would have made it really too
1362 unwieldy, quite apart from causing all the code to be include in a
1363 statically linked application, when only some functions are used. This is
1364 relevant even without the DFA addition now that patterns can be compiled in
1365 one application and matched in another.
1366
1367 The downside of splitting up is that there have to be some external
1368 functions and data tables that are used internally in different modules of
1369 the library but which are not part of the API. These have all had their
1370 names changed to start with "_pcre_" so that they are unlikely to clash
1371 with other external names.
1372
1373 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1374 a different (DFA) algorithm. Although it is slower than the original
1375 function, it does have some advantages for certain types of matching
1376 problem.
1377
1378 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1379 including restarting after a partial match.
1380
1381 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1382 defined when compiling for Windows was sent to me. I have put it into the
1383 code, though I have no means of testing or verifying it.
1384
1385 7. Added the pcre_refcount() auxiliary function.
1386
1387 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1388 match before or at the first newline in the subject string. In pcretest,
1389 the /f option on a pattern can be used to set this.
1390
1391 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1392 would behave wrongly. This has been present in PCRE since release 4.0.
1393
1394 10. A number of changes to the pcregrep command:
1395
1396 (a) Refactored how -x works; insert ^(...)$ instead of setting
1397 PCRE_ANCHORED and checking the length, in preparation for adding
1398 something similar for -w.
1399
1400 (b) Added the -w (match as a word) option.
1401
1402 (c) Refactored the way lines are read and buffered so as to have more
1403 than one at a time available.
1404
1405 (d) Implemented a pcregrep test script.
1406
1407 (e) Added the -M (multiline match) option. This allows patterns to match
1408 over several lines of the subject. The buffering ensures that at least
1409 8K, or the rest of the document (whichever is the shorter) is available
1410 for matching (and similarly the previous 8K for lookbehind assertions).
1411
1412 (f) Changed the --help output so that it now says
1413
1414 -w, --word-regex(p)
1415
1416 instead of two lines, one with "regex" and the other with "regexp"
1417 because that confused at least one person since the short forms are the
1418 same. (This required a bit of code, as the output is generated
1419 automatically from a table. It wasn't just a text change.)
1420
1421 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1422 option but starts with a hyphen. Could be a pattern or a path name
1423 starting with a hyphen, for instance.
1424
1425 (h) "-" can be given as a file name to represent stdin.
1426
1427 (i) When file names are being printed, "(standard input)" is used for
1428 the standard input, for compatibility with GNU grep. Previously
1429 "<stdin>" was used.
1430
1431 (j) The option --label=xxx can be used to supply a name to be used for
1432 stdin when file names are being printed. There is no short form.
1433
1434 (k) Re-factored the options decoding logic because we are going to add
1435 two more options that take data. Such options can now be given in four
1436 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1437
1438 (l) Added the -A, -B, and -C options for requesting that lines of context
1439 around matches be printed.
1440
1441 (m) Added the -L option to print the names of files that do not contain
1442 any matching lines, that is, the complement of -l.
1443
1444 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1445 continue to scan other files.
1446
1447 (o) The -s option was incorrectly implemented. For compatibility with other
1448 greps, it now suppresses the error message for a non-existent or non-
1449 accessible file (but not the return code). There is a new option called
1450 -q that suppresses the output of matching lines, which was what -s was
1451 previously doing.
1452
1453 (p) Added --include and --exclude options to specify files for inclusion
1454 and exclusion when recursing.
1455
1456 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1457 Hopefully, it now does.
1458
1459 12. Missing cast in pcre_study().
1460
1461 13. Added an "uninstall" target to the makefile.
1462
1463 14. Replaced "extern" in the function prototypes in Makefile.in with
1464 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1465 world, but is set differently for Windows.
1466
1467 15. Added a second compiling function called pcre_compile2(). The only
1468 difference is that it has an extra argument, which is a pointer to an
1469 integer error code. When there is a compile-time failure, this is set
1470 non-zero, in addition to the error test pointer being set to point to an
1471 error message. The new argument may be NULL if no error number is required
1472 (but then you may as well call pcre_compile(), which is now just a
1473 wrapper). This facility is provided because some applications need a
1474 numeric error indication, but it has also enabled me to tidy up the way
1475 compile-time errors are handled in the POSIX wrapper.
1476
1477 16. Added VPATH=.libs to the makefile; this should help when building with one
1478 prefix path and installing with another. (Or so I'm told by someone who
1479 knows more about this stuff than I do.)
1480
1481 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1482 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1483 match everything, including newlines. This is not POSIX-compatible, but
1484 somebody wanted the feature. From pcretest it can be activated by using
1485 both the P and the s flags.
1486
1487 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1488
1489 19. libpcre.pc was being incorrectly installed as executable.
1490
1491 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1492 it now also looks for '\r' so that it will work unmodified on Windows.
1493
1494 21. Added Google's contributed C++ wrapper to the distribution.
1495
1496 22. Added some untidy missing memory free() calls in pcretest, to keep
1497 Electric Fence happy when testing.
1498
1499
1500
1501 Version 5.0 13-Sep-04
1502 ---------------------
1503
1504 1. Internal change: literal characters are no longer packed up into items
1505 containing multiple characters in a single byte-string. Each character
1506 is now matched using a separate opcode. However, there may be more than one
1507 byte in the character in UTF-8 mode.
1508
1509 2. The pcre_callout_block structure has two new fields: pattern_position and
1510 next_item_length. These contain the offset in the pattern to the next match
1511 item, and its length, respectively.
1512
1513 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1514 insertion of callouts before each pattern item. Added the /C option to
1515 pcretest to make use of this.
1516
1517 4. On the advice of a Windows user, the lines
1518
1519 #if defined(_WIN32) || defined(WIN32)
1520 _setmode( _fileno( stdout ), 0x8000 );
1521 #endif /* defined(_WIN32) || defined(WIN32) */
1522
1523 have been added to the source of pcretest. This apparently does useful
1524 magic in relation to line terminators.
1525
1526 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1527 for the benefit of those environments where the "b" makes a difference.
1528
1529 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1530 to know about it. I have put a hack into configure.in that adds in code
1531 to set GCC=yes if CC=icc. This seems to end up at a point in the
1532 generated configure script that is early enough to affect the setting of
1533 compiler options, which is what is needed, but I have no means of testing
1534 whether it really works. (The user who reported this had patched the
1535 generated configure script, which of course I cannot do.)
1536
1537 LATER: After change 22 below (new libtool files), the configure script
1538 seems to know about icc (and also ecc). Therefore, I have commented out
1539 this hack in configure.in.
1540
1541 7. Added support for pkg-config (2 patches were sent in).
1542
1543 8. Negated POSIX character classes that used a combination of internal tables
1544 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1545 [[:^ascii]]. Typically, they would match almost any characters. The other
1546 POSIX classes were not broken in this way.
1547
1548 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1549 to find the match, as PCRE was deluded into thinking that the match had to
1550 start at the start point or following a newline. The same bug applied to
1551 patterns with negative forward assertions or any backward assertions
1552 preceding ".*" at the start, unless the pattern required a fixed first
1553 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1554
1555 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1556 starting at the last subject character, bytes beyond the end of the subject
1557 string were read.
1558
1559 11. Renamed the variable "class" as "classbits" to make life easier for C++
1560 users. (Previously there was a macro definition, but it apparently wasn't
1561 enough.)
1562
1563 12. Added the new field "tables" to the extra data so that tables can be passed
1564 in at exec time, or the internal tables can be re-selected. This allows
1565 a compiled regex to be saved and re-used at a later time by a different
1566 program that might have everything at different addresses.
1567
1568 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1569 -R library as well as a -L library.
1570
1571 14. The debugging options of pcretest (-d on the command line or D on a
1572 pattern) showed incorrect output for anything following an extended class
1573 that contained multibyte characters and which was followed by a quantifier.
1574
1575 15. Added optional support for general category Unicode character properties
1576 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1577 support. It adds about 90K to the size of the library. The meanings of the
1578 inbuilt class escapes such as \d and \s have NOT been changed.
1579
1580 16. Updated pcredemo.c to include calls to free() to release the memory for the
1581 compiled pattern.
1582
1583 17. The generated file chartables.c was being created in the source directory
1584 instead of in the building directory. This caused the build to fail if the
1585 source directory was different from the building directory, and was
1586 read-only.
1587
1588 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1589 file. No doubt somebody will tell me if they don't make sense... Also added
1590 Dan Mooney's comments about building on OpenVMS.
1591
1592 19. Added support for partial matching via the PCRE_PARTIAL option for
1593 pcre_exec() and the \P data escape in pcretest.
1594
1595 20. Extended pcretest with 3 new pattern features:
1596
1597 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1598 write the compiled pattern to the file whose name is "rest-of-line".
1599 This is a straight binary dump of the data, with the saved pointer to
1600 the character tables forced to be NULL. The study data, if any, is
1601 written too. After writing, pcretest reads a new pattern.
1602
1603 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1604 compiled pattern from the given file. There must not be any
1605 occurrences of "<" in the file name (pretty unlikely); if there are,
1606 pcretest will instead treat the initial "<" as a pattern delimiter.
1607 After reading in the pattern, pcretest goes on to read data lines as
1608 usual.
1609
1610 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1611 and 16-bit fields in a compiled pattern, to simulate a pattern that
1612 was compiled on a host of opposite endianness.
1613
1614 21. The pcre-exec() function can now cope with patterns that were compiled on
1615 hosts of opposite endianness, with this restriction:
1616
1617 As for any compiled expression that is saved and used later, the tables
1618 pointer field cannot be preserved; the extra_data field in the arguments
1619 to pcre_exec() should be used to pass in a tables address if a value
1620 other than the default internal tables were used at compile time.
1621
1622 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1623 now diagnosed as an error. Previously, most of the time, a negative number
1624 would have been treated as zero, but if in addition "ovector" was passed as
1625 NULL, a crash could occur.
1626
1627 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1628 new versions from the libtool 1.5 distribution (the last one is a copy of
1629 a file called libtool.m4). This seems to have fixed the need to patch
1630 "configure" to support Darwin 1.3 (which I used to do). However, I still
1631 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1632 workstation).
1633
1634 24. Changed the PCRE licence to be the more standard "BSD" licence.
1635
1636
1637 Version 4.5 01-Dec-03
1638 ---------------------
1639
1640 1. There has been some re-arrangement of the code for the match() function so
1641 that it can be compiled in a version that does not call itself recursively.
1642 Instead, it keeps those local variables that need separate instances for
1643 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1644 needs to "recurse". Keeping track of where control must go is done by means
1645 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1646 hide most of the details from the main code, and operates only if
1647 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1648 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1649 operating.
1650
1651 To make it easier for callers to provide specially tailored get/free
1652 functions for this usage, two new functions, pcre_stack_malloc, and
1653 pcre_stack_free, are used. They are always called in strict stacking order,
1654 and the size of block requested is always the same.
1655
1656 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1657 PCRE has been compiled to use the stack or the heap for recursion. The
1658 -C option of pcretest uses this to show which version is compiled.
1659
1660 A new data escape \S, is added to pcretest; it causes the amounts of store
1661 obtained and freed by both kinds of malloc/free at match time to be added
1662 to the output.
1663
1664 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1665 what's available on my current Linux desktop machine.
1666
1667 3. When matching a UTF-8 string, the test for a valid string at the start has
1668 been extended. If start_offset is not zero, PCRE now checks that it points
1669 to a byte that is the start of a UTF-8 character. If not, it returns
1670 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1671 this is necessary because there may be backward assertions in the pattern.
1672 When matching the same subject several times, it may save resources to use
1673 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1674
1675 4. The code for checking the validity of UTF-8 strings has been tightened so
1676 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1677 containing "overlong sequences".
1678
1679 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1680 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1681 should have been "&", but it just so happened that all the cases this let
1682 through by mistake were picked up later in the function.
1683
1684 6. I had used a variable called "isblank" - this is a C99 function, causing
1685 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1686
1687 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1688 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1689 so the version is shown; (c) stop "make test" echoing "./RunTest".
1690
1691 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1692
1693 9. The prototype for memmove() for systems that don't have it was using
1694 size_t, but the inclusion of the header that defines size_t was later. I've
1695 moved the #includes for the C headers earlier to avoid this.
1696
1697 10. Added some adjustments to the code to make it easier to compiler on certain
1698 special systems:
1699
1700 (a) Some "const" qualifiers were missing.
1701 (b) Added the macro EXPORT before all exported functions; by default this
1702 is defined to be empty.
1703 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1704 that it reads its output file name as an argument instead of writing
1705 to the standard output and assuming this can be redirected.
1706
1707 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1708 class containing characters with values greater than 255, PCRE compilation
1709 went into a loop.
1710
1711 12. A recursive reference to a subpattern that was within another subpattern
1712 that had a minimum quantifier of zero caused PCRE to crash. For example,
1713 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1714 recursion. If the recursively-called subpattern itself had a zero repeat,
1715 that was OK.
1716
1717 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1718 buffer into which it was copied (for escape processing) was still set at
1719 1024, so long lines caused crashes.
1720
1721 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1722 "internal error: code overflow...". This applied to any character class
1723 that was followed by a possessive quantifier.
1724
1725 15. Modified the Makefile to add libpcre.la as a prerequisite for
1726 libpcreposix.la because I was told this is needed for a parallel build to
1727 work.
1728
1729 16. If a pattern that contained .* following optional items at the start was
1730 studied, the wrong optimizing data was generated, leading to matching
1731 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1732 matching string must start with a or b or c. The correct conclusion for
1733 this pattern is that a match can start with any character.
1734
1735
1736 Version 4.4 13-Aug-03
1737 ---------------------
1738
1739 1. In UTF-8 mode, a character class containing characters with values between
1740 127 and 255 was not handled correctly if the compiled pattern was studied.
1741 In fixing this, I have also improved the studying algorithm for such
1742 classes (slightly).
1743
1744 2. Three internal functions had redundant arguments passed to them. Removal
1745 might give a very teeny performance improvement.
1746
1747 3. Documentation bug: the value of the capture_top field in a callout is *one
1748 more than* the number of the hightest numbered captured substring.
1749
1750 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1751 in incorrectly linking with a previously installed version. They now link
1752 explicitly with libpcre.la.
1753
1754 5. configure.in no longer needs to recognize Cygwin specially.
1755
1756 6. A problem in pcre.in for Windows platforms is fixed.
1757
1758 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1759 pcretest, it used to include the size of the study block as part of its
1760 output. Unfortunately, the structure contains a field that has a different
1761 size on different hardware architectures. This meant that the tests that
1762 showed this size failed. As the block is currently always of a fixed size,
1763 this information isn't actually particularly useful in pcretest output, so
1764 I have just removed it.
1765
1766 8. Three pre-processor statements accidentally did not start in column 1.
1767 Sadly, there are *still* compilers around that complain, even though
1768 standard C has not required this for well over a decade. Sigh.
1769
1770 9. In pcretest, the code for checking callouts passed small integers in the
1771 callout_data field, which is a void * field. However, some picky compilers
1772 complained about the casts involved for this on 64-bit systems. Now
1773 pcretest passes the address of the small integer instead, which should get
1774 rid of the warnings.
1775
1776 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1777 both compile and run time, and gives an error if an invalid UTF-8 sequence
1778 is found. There is a option for disabling this check in cases where the
1779 string is known to be correct and/or the maximum performance is wanted.
1780
1781 11. In response to a bug report, I changed one line in Makefile.in from
1782
1783 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1784 to
1785 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1786
1787 to look similar to other lines, but I have no way of telling whether this
1788 is the right thing to do, as I do not use Windows. No doubt I'll get told
1789 if it's wrong...
1790
1791
1792 Version 4.3 21-May-03
1793 ---------------------
1794
1795 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1796 Makefile.
1797
1798 2. Some refactoring to improve the quality of the code:
1799
1800 (i) The utf8_table... variables are now declared "const".
1801
1802 (ii) The code for \cx, which used the "case flipping" table to upper case
1803 lower case letters, now just substracts 32. This is ASCII-specific,
1804 but the whole concept of \cx is ASCII-specific, so it seems
1805 reasonable.
1806
1807 (iii) PCRE was using its character types table to recognize decimal and
1808 hexadecimal digits in the pattern. This is silly, because it handles
1809 only 0-9, a-f, and A-F, but the character types table is locale-
1810 specific, which means strange things might happen. A private
1811 table is now used for this - though it costs 256 bytes, a table is
1812 much faster than multiple explicit tests. Of course, the standard
1813 character types table is still used for matching digits in subject
1814 strings against \d.
1815
1816 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1817 ending in _t are). So I've renamed it as ESC_tee.
1818
1819 3. The first argument for regexec() in the POSIX wrapper should have been
1820 defined as "const".
1821
1822 4. Changed pcretest to use malloc() for its buffers so that they can be
1823 Electric Fenced for debugging.
1824
1825 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1826 to read one or more bytes before the start of the subject string. Often this
1827 had no effect on PCRE's behaviour, but in some circumstances it could
1828 provoke a segmentation fault.
1829
1830 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1831 to try to read one or more bytes before the start of the subject string.
1832
1833 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1834 UTF-8 support could misbehave in various ways if the subject string
1835 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1836 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1837 back over UTF-8 characters.)
1838
1839
1840 Version 4.2 14-Apr-03
1841 ---------------------
1842
1843 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1844
1845 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1846 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1847 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1848 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1849 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1850 and BUILD_EXEEXT
1851 Note: automatic setting of the BUILD variables is not yet working
1852 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1853 compile-time but not at link-time
1854 [LINK]: use for linking executables only
1855 make different versions for Windows and non-Windows
1856 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1857 libraries
1858 [LINK_FOR_BUILD]: new variable
1859 [OBJEXT]: use throughout
1860 [EXEEXT]: use throughout
1861 <winshared>: new target
1862 <wininstall>: new target
1863 <dftables.o>: use native compiler
1864 <dftables>: use native linker
1865 <install>: handle Windows platform correctly
1866 <clean>: ditto
1867 <check>: ditto
1868 copy DLL to top builddir before testing
1869
1870 As part of these changes, -no-undefined was removed again. This was reported
1871 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1872 in any case.
1873
1874 3. Some tidies to get rid of compiler warnings:
1875
1876 . In the match_data structure, match_limit was an unsigned long int, whereas
1877 match_call_count was an int. I've made them both unsigned long ints.
1878
1879 . In pcretest the fact that a const uschar * doesn't automatically cast to
1880 a void * provoked a warning.
1881
1882 . Turning on some more compiler warnings threw up some "shadow" variables
1883 and a few more missing casts.
1884
1885 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1886 option, a class that contained a single character with a value between 128
1887 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1888
1889 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1890 option, a class that contained several characters, but with at least one
1891 whose value was between 128 and 255 caused PCRE to crash.
1892
1893
1894 Version 4.1 12-Mar-03
1895 ---------------------
1896
1897 1. Compiling with gcc -pedantic found a couple of places where casts were
1898 needed, and a string in dftables.c that was longer than standard compilers are
1899 required to support.
1900
1901 2. Compiling with Sun's compiler found a few more places where the code could
1902 be tidied up in order to avoid warnings.
1903
1904 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1905 first of these names is deprecated in the latest Autoconf in favour of the name
1906 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1907 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1908 analogy I have changed it to CFLAGS_FOR_BUILD.
1909
1910 4. Added -no-undefined to the linking command in the Makefile, because this is
1911 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1912 linking step for the pcreposix library.
1913
1914 5. PCRE was failing to diagnose the case of two named groups with the same
1915 name.
1916
1917 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1918 literal character that is needed in the subject for a match, and scans along to
1919 ensure that it is present before embarking on the full matching process. This
1920 saves time in cases of nested unlimited repeats that are never going to match.
1921 Problem: the scan can take a lot of time if the subject is very long (e.g.
1922 megabytes), thus penalizing straightforward matches. It is now done only if the
1923 amount of subject to be scanned is less than 1000 bytes.
1924
1925 7. A lesser problem with the same optimization is that it was recording the
1926 first character of an anchored pattern as "needed", thus provoking a search
1927 right along the subject, even when the first match of the pattern was going to
1928 fail. The "needed" character is now not set for anchored patterns, unless it
1929 follows something in the pattern that is of non-fixed length. Thus, it still
1930 fulfils its original purpose of finding quick non-matches in cases of nested
1931 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1932
1933
1934 Version 4.0 17-Feb-03
1935 ---------------------
1936
1937 1. If a comment in an extended regex that started immediately after a meta-item
1938 extended to the end of string, PCRE compiled incorrect data. This could lead to
1939 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1940
1941 2. Moved to autoconf 2.53 and libtool 1.4.2.
1942
1943 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1944 the special perltest8 script is no longer needed - all the tests can be run
1945 from a single perltest script.
1946
1947 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1948 by \s. It has now been removed in PCRE. This means it isn't recognized as
1949 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1950 class [:space:] *does* include VT, thereby creating a mess.
1951
1952 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1953 space and tab.
1954
1955 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1956 its new features into the main test script, reducing the number of scripts.
1957
1958 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1959 were backward compatible, and made the (?i) apply to the whole pattern, as if
1960 /i were given. Now it behaves more logically, and applies the option setting
1961 only to what follows. PCRE has been changed to follow suit. However, if it
1962 finds options settings right at the start of the pattern, it extracts them into
1963 the global options, as before. Thus, they show up in the info data.
1964
1965 8. Added support for the \Q...\E escape sequence. Characters in between are
1966 treated as literals. This is slightly different from Perl in that $ and @ are
1967 also handled as literals inside the quotes. In Perl, they will cause variable
1968 interpolation. Note the following examples:
1969
1970 Pattern PCRE matches Perl matches
1971
1972 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1973 \Qabc\$xyz\E abc\$xyz abc\$xyz
1974 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1975
1976 For compatibility with Perl, \Q...\E sequences are recognized inside character
1977 classes as well as outside them.
1978
1979 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1980 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1981 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1982 signed/unsigned warnings.
1983
1984 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1985 option for pcretest, so I've replaced it by a simple function that does just
1986 that job.
1987
1988 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1989 "pcregrep -".
1990
1991 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1992 Java package. This provides some syntactic sugar for simple cases of what my
1993 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1994 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1995 item, you can use this simplified notation. Note that only makes sense with
1996 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1997 greediness, whatever the setting of the PCRE_UNGREEDY option.
1998
1999 13. A change of greediness default within a pattern was not taking effect at
2000 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2001 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2002 was abstracted outside.
2003
2004 14. PCRE now supports the \G assertion. It is true when the current matching
2005 position is at the start point of the match. This differs from \A when the
2006 starting offset is non-zero. Used with the /g option of pcretest (or similar
2007 code), it works in the same way as it does for Perl's /g option. If all
2008 alternatives of a regex begin with \G, the expression is anchored to the start
2009 match position, and the "anchored" flag is set in the compiled expression.
2010
2011 15. Some bugs concerning the handling of certain option changes within patterns
2012 have been fixed. These applied to options other than (?ims). For example,
2013 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2014 been the other way round. Some of this was related to change 7 above.
2015
2016 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2017 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2018 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2019 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2020
2021 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2022 mode. Unlike ".", it always matches newline, whatever the setting of
2023 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2024 assertions. Perl allows it, but it doesn't (in general) work because it can't
2025 calculate the length of the lookbehind. At least, that's the case for Perl
2026 5.8.0 - I've been told they are going to document that it doesn't work in
2027 future.
2028
2029 18. Added an error diagnosis for escapes that PCRE does not support: these are
2030 \L, \l, \N, \P, \p, \U, \u, and \X.
2031
2032 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2033 reading past the end of the pattern in cases such as /[abcd/.
2034
2035 20. PCRE was getting more memory than necessary for patterns with classes that
2036 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2037
2038 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2039 compiling PCRE for use with Virtual Pascal.
2040
2041 22. Small fix to the Makefile to make it work properly if the build is done
2042 outside the source tree.
2043
2044 23. Added a new extension: a condition to go with recursion. If a conditional
2045 subpattern starts with (?(R) the "true" branch is used if recursion has
2046 happened, whereas the "false" branch is used only at the top level.
2047
2048 24. When there was a very long string of literal characters (over 255 bytes
2049 without UTF support, over 250 bytes with UTF support), the computation of how
2050 much memory was required could be incorrect, leading to segfaults or other
2051 strange effects.
2052
2053 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2054 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2055 there was a subsequent back reference to those brackets. This meant that, for
2056 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2057 possible to check for precisely this case. All we can do is abandon the
2058 optimization if .* occurs inside capturing brackets when there are any back
2059 references whatsoever. (See below for a better fix that came later.)
2060
2061 26. The handling of the optimization for finding the first character of a
2062 non-anchored pattern, and for finding a character that is required later in the
2063 match were failing in some cases. This didn't break the matching; it just
2064 failed to optimize when it could. The way this is done has been re-implemented.
2065
2066 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2067
2068 28. Added a new feature that provides some of the functionality that Perl
2069 provides with (?{...}). The facility is termed a "callout". The way it is done
2070 in PCRE is for the caller to provide an optional function, by setting
2071 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2072 global variable. By default it is unset, which disables all calling out. To get
2073 the function called, the regex must include (?C) at appropriate points. This
2074 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2075 This provides a means of identifying different callout points. When PCRE
2076 reaches such a point in the regex, if pcre_callout has been set, the external
2077 function is called. It is provided with data in a structure called
2078 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2079 matching continues; if it returns a non-zero value, the match at the current
2080 point fails. However, backtracking will occur if possible. [This was changed
2081 later and other features added - see item 49 below.]
2082
2083 29. pcretest is upgraded to test the callout functionality. It provides a
2084 callout function that displays information. By default, it shows the start of
2085 the match and the current position in the text. There are some new data escapes
2086 to vary what happens:
2087
2088 \C+ in addition, show current contents of captured substrings
2089 \C- do not supply a callout function
2090 \C!n return 1 when callout number n is reached
2091 \C!n!m return 1 when callout number n is reached for the mth time
2092
2093 30. If pcregrep was called with the -l option and just a single file name, it
2094 output "<stdin>" if a match was found, instead of the file name.
2095
2096 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2097 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2098 pcre_exec(). This saves a malloc/free per call. The default value of
2099 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2100 when configuring.
2101
2102 32. The default maximum size of a compiled pattern is 64K. There have been a
2103 few cases of people hitting this limit. The code now uses macros to handle the
2104 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2105 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2106 configuring. Tests 2 and 5 work only with 2-byte links because they output
2107 debugging information about compiled patterns.
2108
2109 33. Internal code re-arrangements:
2110
2111 (a) Moved the debugging function for printing out a compiled regex into
2112 its own source file (printint.c) and used #include to pull it into
2113 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2114 separate copies.
2115
2116 (b) Defined the list of op-code names for debugging as a macro in
2117 internal.h so that it is next to the definition of the opcodes.
2118
2119 (c) Defined a table of op-code lengths for simpler skipping along compiled
2120 code. This is again a macro in internal.h so that it is next to the
2121 definition of the opcodes.
2122
2123 34. Added support for recursive calls to individual subpatterns, along the
2124 lines of Robin Houston's patch (but implemented somewhat differently).
2125
2126 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2127 allow it to read and process whole directories in Win32. This code was
2128 contributed by Lionel Fourquaux; it has not been tested by me.
2129
2130 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2131 used to name a group. Names consist of alphanumerics and underscores, and must
2132 be unique. Back references use the syntax (?P=name) and recursive calls use
2133 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2134 numbers. The function pcre_fullinfo() can be used after compilation to extract
2135 a name/number map. There are three relevant calls:
2136
2137 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2138 PCRE_INFO_NAMECOUNT yields the number of entries
2139 PCRE_INFO_NAMETABLE yields a pointer to the map.
2140
2141 The map is a vector of fixed-size entries. The size of each entry depends on
2142 the length of the longest name used. The first two bytes of each entry are the
2143 group number, most significant byte first. There follows the corresponding
2144 name, zero terminated. The names are in alphabetical order.
2145
2146 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2147 case instead of 255. Making it the same both with and without UTF-8 support
2148 means that the same test output works with both.
2149
2150 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2151 calling malloc() with a zero argument.
2152
2153 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2154 optimization. I've improved things by keeping a bitmap of backreferences with
2155 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2156 fact referenced, the optimization can be applied. It is unlikely that a
2157 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2158 the match to follow \n) will appear inside brackets with a number greater than
2159 31, but if it does, any back reference > 31 suppresses the optimization.
2160
2161 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2162 of disabling numbered capturing parentheses. Any opening parenthesis that is
2163 not followed by ? behaves as if it were followed by ?: but named parentheses
2164 can still be used for capturing (and they will acquire numbers in the usual
2165 way).
2166
2167 41. Redesigned the return codes from the match() function into yes/no/error so
2168 that errors can be passed back from deep inside the nested calls. A malloc
2169 failure while inside a recursive subpattern call now causes the
2170 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2171
2172 42. It is now possible to set a limit on the number of times the match()
2173 function is called in a call to pcre_exec(). This facility makes it possible to
2174 limit the amount of recursion and backtracking, though not in a directly
2175 obvious way, because the match() function is used in a number of different
2176 circumstances. The count starts from zero for each position in the subject
2177 string (for non-anchored patterns). The default limit is, for compatibility, a
2178 large number, namely 10 000 000. You can change this in two ways:
2179
2180 (a) When configuring PCRE before making, you can use --with-match-limit=n
2181 to set a default value for the compiled library.
2182
2183 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2184 a different value is set. See 45 below.
2185
2186 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2187
2188 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2189 of things that can be changed at compile time. The first argument specifies
2190 what is wanted and the second points to where the information is to be placed.
2191 The current list of available information is:
2192
2193 PCRE_CONFIG_UTF8
2194
2195 The output is an integer that is set to one if UTF-8 support is available;
2196 otherwise it is set to zero.
2197
2198 PCRE_CONFIG_NEWLINE
2199
2200 The output is an integer that it set to the value of the code that is used for
2201 newline. It is either LF (10) or CR (13).
2202
2203 PCRE_CONFIG_LINK_SIZE
2204
2205 The output is an integer that contains the number of bytes used for internal
2206 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2207
2208 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2209
2210 The output is an integer that contains the threshold above which the POSIX
2211 interface uses malloc() for output vectors. See item 31 above.
2212
2213 PCRE_CONFIG_MATCH_LIMIT
2214
2215 The output is an unsigned integer that contains the default limit of the number
2216 of match() calls in a pcre_exec() execution. See 42 above.
2217
2218 44. pcretest has been upgraded by the addition of the -C option. This causes it
2219 to extract all the available output from the new pcre_config() function, and to
2220 output it. The program then exits immediately.
2221
2222 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2223 order to support additional features. One way would have been to define
2224 pcre_exec2() (for example) with extra arguments, but this would not have been
2225 extensible, and would also have required all calls to the original function to
2226 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2227 is used for passing in "extra" data from pcre_study().
2228
2229 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2230 contains the following fields:
2231
2232 flags a bitmap indicating which of the following fields are set
2233 study_data opaque data from pcre_study()
2234 match_limit a way of specifying a limit on match() calls for a specific
2235 call to pcre_exec()
2236 callout_data data for callouts (see 49 below)
2237
2238 The flag bits are also defined in pcre.h, and are
2239
2240 PCRE_EXTRA_STUDY_DATA
2241 PCRE_EXTRA_MATCH_LIMIT
2242 PCRE_EXTRA_CALLOUT_DATA
2243
2244 The pcre_study() function now returns one of these new pcre_extra blocks, with
2245 the actual study data pointed to by the study_data field, and the
2246 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2247 before. That is, this change is entirely upwards-compatible and requires no
2248 change to existing code.
2249
2250 If you want to pass in additional data to pcre_exec(), you can either place it
2251 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2252 block.
2253
2254 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2255 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2256 times with different match limits, until it finds the minimum value needed for
2257 pcre_exec() to complete. The value is then output. This can be instructive; for
2258 most simple matches the number is quite small, but for pathological cases it
2259 gets very large very quickly.
2260
2261 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2262 returns the size of the data block pointed to by the study_data field in a
2263 pcre_extra block, that is, the value that was passed as the argument to
2264 pcre_malloc() when PCRE was getting memory in which to place the information
2265 created by pcre_study(). The fourth argument should point to a size_t variable.
2266 pcretest has been extended so that this information is shown after a successful
2267 pcre_study() call when information about the compiled regex is being displayed.
2268
2269 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2270 because what follows is always an absolute path. (Later: it turns out that this
2271 is more than cosmetic for MinGW, because it doesn't like empty path
2272 components.)
2273
2274 49. Some changes have been made to the callout feature (see 28 above):
2275
2276 (i) A callout function now has three choices for what it returns:
2277
2278 0 => success, carry on matching
2279 > 0 => failure at this point, but backtrack if possible
2280 < 0 => serious error, return this value from pcre_exec()
2281
2282 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2283 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2284 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2285 use by callout functions. It will never be used by PCRE itself.
2286
2287 (ii) The pcre_extra structure (see 45 above) has a void * field called
2288 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2289 pcre_callout_block structure has a field of the same name. The contents of
2290 the field passed in the pcre_extra structure are passed to the callout
2291 function in the corresponding field in the callout block. This makes it
2292 easier to use the same callout-containing regex from multiple threads. For
2293 testing, the pcretest program has a new data escape
2294
2295 \C*n pass the number n (may be negative) as callout_data
2296
2297 If the callout function in pcretest receives a non-zero value as
2298 callout_data, it returns that value.
2299
2300 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2301 there were some redundant $(CFLAGS) in commands that are now specified as
2302 $(LINK), which already includes $(CFLAGS).
2303
2304 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2305 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2306 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2307 one-byte characters throughout. Note that case-insensitive matching applies
2308 only to characters whose values are less than 256. PCRE doesn't support the
2309 notion of cases for higher-valued characters.
2310
2311 (i) A character class whose characters are all within 0-255 is handled as
2312 a bit map, and the map is inverted for negative classes. Previously, a
2313 character > 255 always failed to match such a class; however it should
2314 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2315
2316 (ii) A negated character class with a single character < 255 is coded as
2317 "not this character" (OP_NOT). This wasn't working properly when the test
2318 character was multibyte, either singly or repeated.
2319
2320 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2321 mode, for example: \x{100}{2,3}.
2322
2323 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2324 singly or repeated) now correctly test multibyte characters. However,
2325 PCRE doesn't recognize any characters with values greater than 255 as
2326 digits, spaces, or word characters. Such characters always match \D, \S,
2327 and \W, and never match \d, \s, or \w.
2328
2329 (v) Classes may now contain characters and character ranges with values
2330 greater than 255. For example: [ab\x{100}-\x{400}].
2331
2332 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2333 PCRE in UTF-8 mode.
2334
2335 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2336 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2337 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2338 value.)
2339
2340 53. The single man page has become too large. I have therefore split it up into
2341 a number of separate man pages. These also give rise to individual HTML pages;
2342 these are now put in a separate directory, and there is an index.html page that
2343 lists them all. Some hyperlinking between the pages has been installed.
2344
2345 54. Added convenience functions for handling named capturing parentheses.
2346
2347 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2348 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2349 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2350 are faulted.
2351
2352 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2353 calling configure. These values are used when compiling the dftables.c program
2354 which is run to generate the source of the default character tables. They
2355 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2356 you will need to set these values.
2357
2358 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2359
2360
2361 Version 3.9 02-Jan-02
2362 ---------------------
2363
2364 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2365
2366 2. If --disable-static was given, the building process failed when trying to
2367 build pcretest and pcregrep. (For some reason it was using libtool to compile
2368 them, which is not right, as they aren't part of the library.)
2369
2370
2371 Version 3.8 18-Dec-01
2372 ---------------------
2373
2374 1. The experimental UTF-8 code was completely screwed up. It was packing the
2375 bytes in the wrong order. How dumb can you get?
2376
2377
2378 Version 3.7 29-Oct-01
2379 ---------------------
2380
2381 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2382 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2383 this didn't happen under Solaris 8, where I normally test things.
2384
2385 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2386 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2387 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2388 this fix an hour or so after the initial 3.7 release.)
2389
2390
2391 Version 3.6 23-Oct-01
2392 ---------------------
2393
2394 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2395 offsets passed as NULL with zero offset count.
2396
2397 2. The config.guess and config.sub files had not been updated when I moved to
2398 the latest autoconf.
2399
2400
2401 Version 3.5 15-Aug-01
2402 ---------------------
2403
2404 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2405 had been forgotten.
2406
2407 2. By using declared but undefined structures, we can avoid using "void"
2408 definitions in pcre.h while keeping the internal definitions of the structures
2409 private.
2410
2411 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2412 user point of view, this means that both static and shared libraries are built
2413 by default, but this can be individually controlled. More of the work of
2414 handling this static/shared cases is now inside libtool instead of PCRE's make
2415 file.
2416
2417 4. The pcretest utility is now installed along with pcregrep because it is
2418 useful for users (to test regexs) and by doing this, it automatically gets
2419 relinked by libtool. The documentation has been turned into a man page, so
2420 there are now .1, .txt, and .html versions in /doc.
2421
2422 5. Upgrades to pcregrep:
2423 (i) Added long-form option names like gnu grep.
2424 (ii) Added --help to list all options with an explanatory phrase.
2425 (iii) Added -r, --recursive to recurse into sub-directories.
2426 (iv) Added -f, --file to read patterns from a file.
2427
2428 6. pcre_exec() was referring to its "code" argument before testing that
2429 argument for NULL (and giving an error if it was NULL).
2430
2431 7. Upgraded Makefile.in to allow for compiling in a different directory from
2432 the source directory.
2433
2434 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2435 options bits, the pointer it was passed was to an int instead of to an unsigned
2436 long int. This mattered only on 64-bit systems.
2437
2438 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2439 generated) instead of pcre.in, which it its source. Also made the same change
2440 in several of the .c files.
2441
2442 10. A new release of gcc defines printf() as a macro, which broke pcretest
2443 because it had an ifdef in the middle of a string argument for printf(). Fixed
2444 by using separate calls to printf().
2445
2446 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2447 script, to force use of CR or LF instead of \n in the source. On non-Unix
2448 systems, the value can be set in config.h.
2449
2450 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2451 absolute limit. Changed the text of the error message to make this clear, and
2452 likewise updated the man page.
2453
2454 13. The limit of 99 on the number of capturing subpatterns has been removed.
2455 The new limit is 65535, which I hope will not be a "real" limit.
2456
2457
2458 Version 3.4 22-Aug-00
2459 ---------------------
2460
2461 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2462
2463 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2464
2465
2466 Version 3.3 01-Aug-00
2467 ---------------------
2468
2469 1. If an octal character was given, but the value was greater than \377, it
2470 was not getting masked to the least significant bits, as documented. This could
2471 lead to crashes in some systems.
2472
2473 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2474 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2475
2476 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2477 These just pass their arguments on to (pcre_free)(), but they are provided
2478 because some uses of PCRE bind it to non-C systems that can call its functions,
2479 but cannot call free() or pcre_free() directly.
2480
2481 4. Add "make test" as a synonym for "make check". Corrected some comments in
2482 the Makefile.
2483
2484 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2485 Makefile.
2486
2487 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2488 command called pgrep for grepping around the active processes.
2489
2490 7. Added the beginnings of support for UTF-8 character strings.
2491
2492 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2493 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2494 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2495 out for the ar command.)
2496
2497
2498 Version 3.2 12-May-00
2499 ---------------------
2500
2501 This is purely a bug fixing release.
2502
2503 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2504 of ZA. This was just one example of several cases that could provoke this bug,
2505 which was introduced by change 9 of version 2.00. The code for breaking
2506 infinite loops after an iteration that matches an empty string was't working
2507 correctly.
2508
2509 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2510 when matched against abbab (for example). After matching an empty string, it
2511 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2512 caused it to match further down the string than it should.
2513
2514 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2515 was there because it doesn't seem to be needed, and it causes trouble on some
2516 systems, as it is not a Standard C header. It has been removed.
2517
2518 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2519 were reported on the Macintosh. The changes were from
2520
2521 while ((c = *(++ptr)) != 0 && c != '\n');
2522 to
2523 while ((c = *(++ptr)) != 0 && c != '\n') ;
2524
2525 Totally extraordinary, but if that's what it takes...
2526
2527 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2528 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2529 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2530 assumes the way PCRE uses memmove() (always moving upwards).
2531
2532 6. PCRE is being used in one environment where strchr() is not available. There
2533 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2534 faster code anyway.
2535
2536
2537 Version 3.1 09-Feb-00
2538 ---------------------
2539
2540 The only change in this release is the fixing of some bugs in Makefile.in for
2541 the "install" target:
2542
2543 (1) It was failing to install pcreposix.h.
2544
2545 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2546
2547
2548 Version 3.0 01-Feb-00
2549 ---------------------
2550
2551 1. Add support for the /+ modifier to perltest (to output $` like it does in
2552 pcretest).
2553
2554 2. Add support for the /g modifier to perltest.
2555
2556 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2557 matches null strings.
2558
2559 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2560 pattern. Perl treats empty patterns specially - it reuses the most recent
2561 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2562 effect.
2563
2564 5. The POSIX interface was broken in that it was just handing over the POSIX
2565 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2566 required a bigger vector, with some working space on the end. This means that
2567 the POSIX wrapper now has to get and free some memory, and copy the results.
2568
2569 6. Added some simple autoconf support, placing the test data and the
2570 documentation in separate directories, re-organizing some of the
2571 information files, and making it build pcre-config (a GNU standard). Also added
2572 libtool support for building PCRE as a shared library, which is now the
2573 default.
2574
2575 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2576 09 are not valid octal constants. Single digits will be used for minor values
2577 less than 10.
2578
2579 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2580 existing programs that set these in the POSIX interface can use PCRE without
2581 modification.
2582
2583 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2584 return all that pcre_info() returns, plus additional data. The pcre_info()
2585 function is retained for compatibility, but is considered to be obsolete.
2586
2587 10. Added experimental recursion feature (?R) to handle one common case that
2588 Perl 5.6 will be able to do with (?p{...}).
2589
2590 11. Added support for POSIX character classes like [:alpha:], which Perl is
2591 adopting.
2592
2593
2594 Version 2.08 31-Aug-99
2595 ----------------------
2596
2597 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2598 trying to match at the startoffset position, but instead was moving forward to
2599 the next newline as if a previous match had failed.
2600
2601 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2602 and could get into a loop if a null string was matched other than at the start
2603 of the subject.
2604
2605 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2606 be distinguished at compile time, and for completeness also added PCRE_DATE.
2607
2608 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2609 in GnuWin32 environments.
2610
2611
2612 Version 2.07 29-Jul-99
2613 ----------------------
2614
2615 1. The documentation is now supplied in plain text form and HTML as well as in
2616 the form of man page sources.
2617
2618 2. C++ compilers don't like assigning (void *) values to other pointer types.
2619 In particular this affects malloc(). Although there is no problem in Standard
2620 C, I've put in casts to keep C++ compilers happy.
2621
2622 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2623 should be (const char *).
2624
2625 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2626 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2627 However, I haven't made this a standard facility. The documentation doesn't
2628 mention it, and the Makefile doesn't support it.
2629
2630 5. The Makefile now contains an "install" target, with editable destinations at
2631 the top of the file. The pcretest program is not installed.
2632
2633 6. pgrep -V now gives the PCRE version number and date.
2634
2635 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2636 causing the entire string to be ignored, instead of just the last character.
2637
2638 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2639 non-matching string, it can take a very, very long time, even for strings of
2640 quite modest length, because of the nested recursion. PCRE now does better in
2641 some of these cases. It does this by remembering the last required literal
2642 character in the pattern, and pre-searching the subject to ensure it is present
2643 before running the real match. In other words, it applies a heuristic to detect
2644 some types of certain failure quickly, and in the above example, if presented
2645 with a string that has no trailing " it gives "no match" very quickly.
2646
2647 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2648 other alternatives are tried instead.
2649
2650
2651 Version 2.06 09-Jun-99
2652 ----------------------
2653
2654 1. Change pcretest's output for amount of store used to show just the code
2655 space, because the remainder (the data block) varies in size between 32-bit and
2656 64-bit systems.
2657
2658 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2659 start matching at. This allows lookbehinds to work when searching for multiple
2660 occurrences in a string.
2661
2662 3. Added additional options to pcretest for testing multiple occurrences:
2663
2664 /+ outputs the rest of the string that follows a match
2665 /g loops for multiple occurrences, using the new startoffset argument
2666 /G loops for multiple occurrences by passing an incremented pointer
2667
2668 4. PCRE wasn't doing the "first character" optimization for patterns starting
2669 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2670 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2671 the letter 'x'. On long subject strings, this gives a significant speed-up.
2672
2673
2674 Version 2.05 21-Apr-99
2675 ----------------------
2676
2677 1. Changed the type of magic_number from int to long int so that it works
2678 properly on 16-bit systems.
2679
2680 2. Fixed a bug which caused patterns starting with .* not to work correctly
2681 when the subject string contained newline characters. PCRE was assuming
2682 anchoring for such patterns in all cases, which is not correct because .* will
2683 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2684 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2685 must be retried after every newline in the subject.
2686
2687
2688 Version 2.04 18-Feb-99
2689 ----------------------
2690
2691 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2692 computation of the store needed to hold the pattern was incorrect (too large).
2693 If such patterns were nested a few deep, this could multiply and become a real
2694 problem.
2695
2696 2. Added /M option to pcretest to show the memory requirement of a specific
2697 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2698
2699 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2700 compiled in such a way that the backtracking after subsequent failure was
2701 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2702 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2703
2704
2705 Version 2.03 02-Feb-99
2706 ----------------------
2707
2708 1. Fixed typo and small mistake in man page.
2709
2710 2. Added 4th condition (GPL supersedes if conflict) and created separate
2711 LICENCE file containing the conditions.
2712
2713 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2714 Perl, that is the internal \ allows the delimiter to be included in the
2715 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2716 the final delimiter, add \ to the end of the pattern (to test the error).
2717
2718 4. Added the convenience functions for extracting substrings after a successful
2719 match. Updated pcretest to make it able to test these functions.
2720
2721
2722 Version 2.02 14-Jan-99
2723 ----------------------
2724
2725 1. Initialized the working variables associated with each extraction so that
2726 their saving and restoring doesn't refer to uninitialized store.
2727
2728 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2729 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2730 fix the problem.
2731
2732 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2733 calls, and wasn't printing the correct value for compiling calls. Increased the
2734 default value of LOOPREPEAT, and the number of significant figures in the
2735 times.
2736
2737 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2738
2739 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2740 a building problem on Windows NT with a FAT file system.
2741
2742
2743 Version 2.01 21-Oct-98
2744 ----------------------
2745
2746 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2747 to character tables built by pcre_maketables() in the current locale. If NULL
2748 is passed, the default tables are used.
2749
2750
2751 Version 2.00 24-Sep-98
2752 ----------------------
2753
2754 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2755 it any more.
2756
2757 2. Allow quantification of (?>) groups, and make it work correctly.
2758
2759 3. The first character computation wasn't working for (?>) groups.
2760
2761 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2762 end of the subject) and add 5.005's \z, which really does match only at the
2763 very end of the subject.
2764
2765 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2766
2767 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2768 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2769 localized options. All options to pcre_study() were also removed.
2770
2771 7. Add other new features from 5.005:
2772
2773 $(?<= positive lookbehind
2774 $(?<! negative lookbehind
2775 (?imsx-imsx) added the unsetting capability
2776 such a setting is global if at outer level; local otherwise
2777 (?imsx-imsx:) non-capturing groups with option setting
2778 (?(cond)re|re) conditional pattern matching
2779
2780 A backreference to itself in a repeated group matches the previous
2781 captured string.
2782
2783 8. General tidying up of studying (both automatic and via "study")
2784 consequential on the addition of new assertions.
2785
2786 9. As in 5.005, unlimited repeated groups that could match an empty substring
2787 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2788 runtime if any iteration does actually match an empty substring.
2789
2790 10. Include the RunTest script in the distribution.
2791
2792 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2793 discrepancies, some of which were old and were also with respect to 5.004. They
2794 have now been fixed.
2795
2796
2797 Version 1.09 28-Apr-98
2798 ----------------------
2799
2800 1. A negated single character class followed by a quantifier with a minimum
2801 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2802 program crashes, or just wrong answers. This did not apply to negated classes
2803 containing more than one character, or to minima other than one.
2804
2805
2806 Version 1.08 27-Mar-98
2807 ----------------------
2808
2809 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2810
2811 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2812 latter must appear before anything that relies on it in the pattern.
2813
2814
2815 Version 1.07 16-Feb-98
2816 ----------------------
2817
2818 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2819 repeat of a potentially empty string).
2820
2821
2822 Version 1.06 23-Jan-98
2823 ----------------------
2824
2825 1. Added Markus Oberhumer's little patches for C++.
2826
2827 2. Literal strings longer than 255 characters were broken.
2828
2829
2830 Version 1.05 23-Dec-97
2831 ----------------------
2832
2833 1. Negated character classes containing more than one character were failing if
2834 PCRE_CASELESS was set at run time.
2835
2836
2837 Version 1.04 19-Dec-97
2838 ----------------------
2839
2840 1. Corrected the man page, where some "const" qualifiers had been omitted.
2841
2842 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2843 input syntax.
2844
2845 3. Fixed memory leak which occurred when a regex with back references was
2846 matched with an offsets vector that wasn't big enough. The temporary memory
2847 that is used in this case wasn't being freed if the match failed.
2848
2849 4. Tidied pcretest to ensure it frees memory that it gets.
2850
2851 5. Temporary memory was being obtained in the case where the passed offsets
2852 vector was exactly big enough.
2853
2854 6. Corrected definition of offsetof() from change 5 below.
2855
2856 7. I had screwed up change 6 below and broken the rules for the use of
2857 setjmp(). Now fixed.
2858
2859
2860 Version 1.03 18-Dec-97
2861 ----------------------
2862
2863 1. A erroneous regex with a missing opening parenthesis was correctly
2864 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2865 on some systems.
2866
2867 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2868 it was reported that one broken compiler failed on the former because "code" is
2869 also an independent variable.
2870
2871 3. The erroneous regex a[]b caused an array overrun reference.
2872
2873 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2874 fail on data ending with that character. (It was going on too far, and checking
2875 the next character, typically a binary zero.) This was specific to the
2876 optimized code for single-character negative classes.
2877
2878 5. Added a contributed patch from the TIN world which does the following:
2879
2880 + Add an undef for memmove, in case the the system defines a macro for it.
2881
2882 + Add a definition of offsetof(), in case there isn't one. (I don't know
2883 the reason behind this - offsetof() is part of the ANSI standard - but
2884 it does no harm).
2885
2886 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2887 most of the places where whitespace preceded '#'. I have given up and
2888 allowed the remaining 2 cases to be at the margin.
2889
2890 + Rename some variables in pcre to eliminate shadowing. This seems very
2891 pedantic, but does no harm, of course.
2892
2893 6. Moved the call to setjmp() into its own function, to get rid of warnings
2894 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2895
2896 7. Constructs such as \d{8,} were compiling into the equivalent of
2897 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2898 outcome, but in this particular case used more store than had been allocated,
2899 which caused the bug to be discovered because it threw up an internal error.
2900
2901 8. The debugging code in both pcre and pcretest for outputting the compiled
2902 form of a regex was going wrong in the case of back references followed by
2903 curly-bracketed repeats.
2904
2905
2906 Version 1.02 12-Dec-97
2907 ----------------------
2908
2909 1. Typos in pcre.3 and comments in the source fixed.
2910
2911 2. Applied a contributed patch to get rid of places where it used to remove
2912 'const' from variables, and fixed some signed/unsigned and uninitialized
2913 variable warnings.
2914
2915 3. Added the "runtest" target to Makefile.
2916
2917 4. Set default compiler flag to -O2 rather than just -O.
2918
2919
2920 Version 1.01 19-Nov-97
2921 ----------------------
2922
2923 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2924 like /([ab]*)*/, that is, for classes with more than one character in them.
2925
2926 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2927 as /((?>a*))*/ (a PCRE_EXTRA facility).
2928
2929
2930 Version 1.00 18-Nov-97
2931 ----------------------
2932
2933 1. Added compile-time macros to support systems such as SunOS4 which don't have
2934 memmove() or strerror() but have other things that can be used instead.
2935
2936 2. Arranged that "make clean" removes the executables.
2937
2938
2939 Version 0.99 27-Oct-97
2940 ----------------------
2941
2942 1. Fixed bug in code for optimizing classes with only one character. It was
2943 initializing a 32-byte map regardless, which could cause it to run off the end
2944 of the memory it had got.
2945
2946 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2947
2948
2949 Version 0.98 22-Oct-97
2950 ----------------------
2951
2952 1. Fixed bug in code for handling temporary memory usage when there are more
2953 back references than supplied space in the ovector. This could cause segfaults.
2954
2955
2956 Version 0.97 21-Oct-97
2957 ----------------------
2958
2959 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2960
2961 2. Optimized negated single characters not to use a bit map.
2962
2963 3. Brought error texts together as macro definitions; clarified some of them;
2964 fixed one that was wrong - it said "range out of order" when it meant "invalid
2965 escape sequence".
2966
2967 4. Changed some char * arguments to const char *.
2968
2969 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2970
2971 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2972 pcretest.
2973
2974
2975 Version 0.96 16-Oct-97
2976 ----------------------
2977
2978 1. Added a simple "pgrep" utility to the distribution.
2979
2980 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2981 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2982 where "ddd" means "one or more decimal digits".
2983
2984 3. Fixed serious bug. If a pattern had a back reference, but the call to
2985 pcre_exec() didn't supply a large enough ovector to record the related
2986 identifying subpattern, the match always failed. PCRE now remembers the number
2987 of the largest back reference, and gets some temporary memory in which to save
2988 the offsets during matching if necessary, in order to ensure that
2989 backreferences always work.
2990
2991 4. Increased the compatibility with Perl in a number of ways:
2992
2993 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2994 to request this handling. The option can be set at compile or exec time.
2995
2996 (b) $ matches before a terminating newline by default; an option
2997 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2998 mode). The option can be set at compile or exec time.
2999
3000 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3001 the same as Perl's. If the decimal number it represents is less than 10
3002 or there aren't that many previous left capturing parentheses, an octal
3003 escape is read. Inside a character class, it's always an octal escape,
3004 even if it is a single digit.
3005
3006 (d) An escaped but undefined alphabetic character is taken as a literal,
3007 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3008 escapes.
3009
3010 (e) {0} is now permitted. (The previous item is removed from the compiled
3011 pattern).
3012
3013 5. Changed all the names of code files so that the basic parts are no longer
3014 than 10 characters, and abolished the teeny "globals.c" file.
3015
3016 6. Changed the handling of character classes; they are now done with a 32-byte
3017 bit map always.
3018
3019 7. Added the -d and /D options to pcretest to make it possible to look at the
3020 internals of compilation without having to recompile pcre.
3021
3022
3023 Version 0.95 23-Sep-97
3024 ----------------------
3025
3026 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3027 \x20 at the start of a run of normal characters. These were being treated as
3028 real characters, instead of the source characters being re-checked.
3029
3030
3031 Version 0.94 18-Sep-97
3032 ----------------------
3033
3034 1. The functions are now thread-safe, with the caveat that the global variables
3035 containing pointers to malloc() and free() or alternative functions are the
3036 same for all threads.
3037
3038 2. Get pcre_study() to generate a bitmap of initial characters for non-
3039 anchored patterns when this is possible, and use it if passed to pcre_exec().
3040
3041
3042 Version 0.93 15-Sep-97
3043 ----------------------
3044
3045 1. /(b)|(:+)/ was computing an incorrect first character.
3046
3047 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3048 but not actually doing anything yet.
3049
3050 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3051 as Perl does (e.g. [-az] or [az-]).
3052
3053 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3054 all possible positions.
3055
3056 5. Split up into different modules to avoid including unneeded functions in a
3057 compiled binary. However, compile and exec are still in one module. The "study"
3058 function is split off.
3059
3060 6. The character tables are now in a separate module whose source is generated
3061 by an auxiliary program - but can then be edited by hand if required. There are
3062 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3063 toupper() in the code.
3064
3065 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3066 make them global. Abolish the function for setting them, as the caller can now
3067 set them directly.
3068
3069
3070 Version 0.92 11-Sep-97
3071 ----------------------
3072
3073 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3074 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3075
3076 2. Caseless matching was not working in character classes if the characters in
3077 the pattern were in upper case.
3078
3079 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3080
3081 4. Make PCRE_ANCHORED public and accept as a compile option.
3082
3083 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3084 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3085 pass them.
3086
3087 6. Give an error if bad option bits passed at compile or run time.
3088
3089 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3090 pcretest to cause it to pass that flag.
3091
3092 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3093 options, and the first character, if set.
3094
3095 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3096
3097
3098 Version 0.91 10-Sep-97
3099 ----------------------
3100
3101 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3102 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3103
3104 2. PCRE was looping on encountering an indefinitely repeated back reference to
3105 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3106 Perl does - treats the match as successful.
3107
3108 ****

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