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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 287 - (show annotations) (download)
Tue Dec 18 20:11:28 2007 UTC (6 years, 9 months ago) by ph10
File size: 146168 byte(s)
Added --enable-pcretest-libreadline.

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 18. Added --enable-pcretest-libreadline.
97
98
99 Version 7.4 21-Sep-07
100 ---------------------
101
102 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
103 means that a class such as [\s] counted as "explicit reference to CR or
104 LF". That isn't really right - the whole point of the change was to try to
105 help when there was an actual mention of one of the two characters. So now
106 the change happens only if \r or \n (or a literal CR or LF) character is
107 encountered.
108
109 2. The 32-bit options word was also used for 6 internal flags, but the numbers
110 of both had grown to the point where there were only 3 bits left.
111 Fortunately, there was spare space in the data structure, and so I have
112 moved the internal flags into a new 16-bit field to free up more option
113 bits.
114
115 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
116 but did not set the internal JCHANGED flag - either of these is enough to
117 control the way the "get" function works - but the PCRE_INFO_JCHANGED
118 facility is supposed to tell if (?J) was ever used, so now (?J) at the
119 start sets both bits.
120
121 4. Added options (at build time, compile time, exec time) to change \R from
122 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
123
124 5. doc/pcresyntax.html was missing from the distribution.
125
126 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
127 compatibility, even though it is no longer used.
128
129 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
130 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
131 windows.h file is present (where different names are used). [This was
132 reversed later after testing - see 16 below.]
133
134 8. Changed all #include <config.h> to #include "config.h". There were also
135 some further <pcre.h> cases that I changed to "pcre.h".
136
137 9. When pcregrep was used with the --colour option, it missed the line ending
138 sequence off the lines that it output.
139
140 10. It was pointed out to me that arrays of string pointers cause lots of
141 relocations when a shared library is dynamically loaded. A technique of
142 using a single long string with a table of offsets can drastically reduce
143 these. I have refactored PCRE in four places to do this. The result is
144 dramatic:
145
146 Originally: 290
147 After changing UCP table: 187
148 After changing error message table: 43
149 After changing table of "verbs" 36
150 After changing table of Posix names 22
151
152 Thanks to the folks working on Gregex for glib for this insight.
153
154 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
155 unicode-properties was also set.
156
157 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
158
159 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
160 checked only for CRLF.
161
162 14. Added casts to pcretest.c to avoid compiler warnings.
163
164 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
165
166 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
167 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
168 entirely. This removes changes made in 7 above.
169
170 17. The CMake files have been updated, and there is now more information about
171 building with CMake in the NON-UNIX-USE document.
172
173
174 Version 7.3 28-Aug-07
175 ---------------------
176
177 1. In the rejigging of the build system that eventually resulted in 7.1, the
178 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
179 brackets there is not right, since it causes compilers to look for an
180 installed pcre.h, not the version that is in the source that is being
181 compiled (which of course may be different). I have changed it back to:
182
183 #include "pcre.h"
184
185 I have a vague recollection that the change was concerned with compiling in
186 different directories, but in the new build system, that is taken care of
187 by the VPATH setting the Makefile.
188
189 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
190 when the subject happened to end in the byte 0x85 (e.g. if the last
191 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
192 characters but of course it shouldn't be taken as a newline when it is part
193 of another character. The bug was that, for an unlimited repeat of . in
194 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
195 characters when looking for a newline.
196
197 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
198
199 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
200 in debug output.
201
202 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
203 long printing in the pcrecpp unittest when running under MinGW.
204
205 6. ESC_K was left out of the EBCDIC table.
206
207 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
208 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
209 limit also applies to "virtual nesting" when a pattern is recursive, and in
210 this case 1000 isn't so big. I have been able to remove this limit at the
211 expense of backing off one optimization in certain circumstances. Normally,
212 when pcre_exec() would call its internal match() function recursively and
213 immediately return the result unconditionally, it uses a "tail recursion"
214 feature to save stack. However, when a subpattern that can match an empty
215 string has an unlimited repetition quantifier, it no longer makes this
216 optimization. That gives it a stack frame in which to save the data for
217 checking that an empty string has been matched. Previously this was taken
218 from the 1000-entry workspace that had been reserved. So now there is no
219 explicit limit, but more stack is used.
220
221 8. Applied Daniel's patches to solve problems with the import/export magic
222 syntax that is required for Windows, and which was going wrong for the
223 pcreposix and pcrecpp parts of the library. These were overlooked when this
224 problem was solved for the main library.
225
226 9. There were some crude static tests to avoid integer overflow when computing
227 the size of patterns that contain repeated groups with explicit upper
228 limits. As the maximum quantifier is 65535, the maximum group length was
229 set at 30,000 so that the product of these two numbers did not overflow a
230 32-bit integer. However, it turns out that people want to use groups that
231 are longer than 30,000 bytes (though not repeat them that many times).
232 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
233 made it possible to implement the integer overflow checks in a much more
234 dynamic way, which I have now done. The artificial limitation on group
235 length has been removed - we now have only the limit on the total length of
236 the compiled pattern, which depends on the LINK_SIZE setting.
237
238 10. Fixed a bug in the documentation for get/copy named substring when
239 duplicate names are permitted. If none of the named substrings are set, the
240 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
241 empty string.
242
243 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
244 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
245 because the ] is interpreted as the first data character and the
246 terminating ] is not found. PCRE has been made compatible with Perl in this
247 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
248 cause memory overwriting.
249
250 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
251 string has been matched (to stop an infinite loop). It was not recognizing
252 a conditional subpattern that could match an empty string if that
253 subpattern was within another subpattern. For example, it looped when
254 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
255 condition was not nested. This bug has been fixed.
256
257 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
258 past the start of the subject in the presence of bytes with the top bit
259 set, for example "\x8aBCD".
260
261 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
262 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
263
264 14. Optimized (?!) to (*FAIL).
265
266 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
267 This restricts code points to be within the range 0 to 0x10FFFF, excluding
268 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
269 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
270 does: it's just the validity check that is more restrictive.
271
272 16. Inserted checks for integer overflows during escape sequence (backslash)
273 processing, and also fixed erroneous offset values for syntax errors during
274 backslash processing.
275
276 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
277 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
278
279 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
280 caused an overrun.
281
282 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
283 something other than just ASCII characters) inside a group that had an
284 unlimited repeat caused a loop at compile time (while checking to see
285 whether the group could match an empty string).
286
287 20. Debugging a pattern containing \p or \P could cause a crash. For example,
288 [\P{Any}] did so. (Error in the code for printing property names.)
289
290 21. An orphan \E inside a character class could cause a crash.
291
292 22. A repeated capturing bracket such as (A)? could cause a wild memory
293 reference during compilation.
294
295 23. There are several functions in pcre_compile() that scan along a compiled
296 expression for various reasons (e.g. to see if it's fixed length for look
297 behind). There were bugs in these functions when a repeated \p or \P was
298 present in the pattern. These operators have additional parameters compared
299 with \d, etc, and these were not being taken into account when moving along
300 the compiled data. Specifically:
301
302 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
303 length.
304
305 (b) An item such as \pL+ within a repeated group could cause crashes or
306 loops.
307
308 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
309 "reference to non-existent subpattern" error.
310
311 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
312
313 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
314 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
315
316 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
317
318 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
319 character were causing crashes (broken optimization).
320
321 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
322 \p or \P) caused a compile-time loop.
323
324 28. More problems have arisen in unanchored patterns when CRLF is a valid line
325 break. For example, the unstudied pattern [\r\n]A does not match the string
326 "\r\nA" because change 7.0/46 below moves the current point on by two
327 characters after failing to match at the start. However, the pattern \nA
328 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
329 the same is true. There doesn't seem any very clean way out of this, but
330 what I have chosen to do makes the common cases work: PCRE now takes note
331 of whether there can be an explicit match for \r or \n anywhere in the
332 pattern, and if so, 7.0/46 no longer applies. As part of this change,
333 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
334 pattern has explicit CR or LF references.
335
336 29. Added (*CR) etc for changing newline setting at start of pattern.
337
338
339 Version 7.2 19-Jun-07
340 ---------------------
341
342 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
343 which is apparently normally available under Windows.
344
345 2. Re-jig the pcregrep tests with different newline settings in an attempt
346 to make them independent of the local environment's newline setting.
347
348 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
349
350 4. Some of the "internals" tests were previously cut out when the link size
351 was not 2, because the output contained actual offsets. The recent new
352 "Z" feature of pcretest means that these can be cut out, making the tests
353 usable with all link sizes.
354
355 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
356 stack recursion. This gives a massive performance boost under BSD, but just
357 a small improvement under Linux. However, it saves one field in the frame
358 in all cases.
359
360 6. Added more features from the forthcoming Perl 5.10:
361
362 (a) (?-n) (where n is a string of digits) is a relative subroutine or
363 recursion call. It refers to the nth most recently opened parentheses.
364
365 (b) (?+n) is also a relative subroutine call; it refers to the nth next
366 to be opened parentheses.
367
368 (c) Conditions that refer to capturing parentheses can be specified
369 relatively, for example, (?(-2)... or (?(+3)...
370
371 (d) \K resets the start of the current match so that everything before
372 is not part of it.
373
374 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
375
376 (f) \g{name} is another synonym - part of Perl 5.10's unification of
377 reference syntax.
378
379 (g) (?| introduces a group in which the numbering of parentheses in each
380 alternative starts with the same number.
381
382 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
383
384 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
385 PCRE_INFO_JCHANGED.
386
387 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
388 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
389 for detecting groups that can match an empty string.
390
391 9. A pattern with a very large number of alternatives (more than several
392 hundred) was running out of internal workspace during the pre-compile
393 phase, where pcre_compile() figures out how much memory will be needed. A
394 bit of new cunning has reduced the workspace needed for groups with
395 alternatives. The 1000-alternative test pattern now uses 12 bytes of
396 workspace instead of running out of the 4096 that are available.
397
398 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
399
400 11. Applied patch from Google to remove an optimization that didn't quite work.
401 The report of the bug said:
402
403 pcrecpp::RE("a*").FullMatch("aaa") matches, while
404 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
405 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
406
407 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
408 it matched the wrong number of bytes.
409
410
411 Version 7.1 24-Apr-07
412 ---------------------
413
414 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
415 that is more "standard", making use of automake and other Autotools. There
416 is some re-arrangement of the files and adjustment of comments consequent
417 on this.
418
419 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
420 for recursive directory scanning broke on some systems because the files
421 are not scanned in any specific order and on different systems the order
422 was different. A call to "sort" has been inserted into RunGrepTest for the
423 approprate test as a short-term fix. In the longer term there may be an
424 alternative.
425
426 3. I had an email from Eric Raymond about problems translating some of PCRE's
427 man pages to HTML (despite the fact that I distribute HTML pages, some
428 people do their own conversions for various reasons). The problems
429 concerned the use of low-level troff macros .br and .in. I have therefore
430 removed all such uses from the man pages (some were redundant, some could
431 be replaced by .nf/.fi pairs). The 132html script that I use to generate
432 HTML has been updated to handle .nf/.fi and to complain if it encounters
433 .br or .in.
434
435 4. Updated comments in configure.ac that get placed in config.h.in and also
436 arranged for config.h to be included in the distribution, with the name
437 config.h.generic, for the benefit of those who have to compile without
438 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
439
440 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
441 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
442 makevp.bat for the current PCRE, using the additional files
443 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
444
445 6. A Windows user reported a minor discrepancy with test 2, which turned out
446 to be caused by a trailing space on an input line that had got lost in his
447 copy. The trailing space was an accident, so I've just removed it.
448
449 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
450 that is needed.
451
452 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
453 as "const" (a) because they are and (b) because it helps the PHP
454 maintainers who have recently made a script to detect big data structures
455 in the php code that should be moved to the .rodata section. I remembered
456 to update Builducptable as well, so it won't revert if ucptable.h is ever
457 re-created.
458
459 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
460 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
461 order to be able to cut out the UTF-8 tables in the latter when UTF-8
462 support is not required. This saves 1.5-2K of code, which is important in
463 some applications.
464
465 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
466 so as not to refer to the tables, even though these functions will never be
467 called when UTF-8 support is disabled. Otherwise there are problems with a
468 shared library.
469
470 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
471
472 (a) It was defining its arguments as char * instead of void *.
473
474 (b) It was assuming that all moves were upwards in memory; this was true
475 a long time ago when I wrote it, but is no longer the case.
476
477 The emulated memove() is provided for those environments that have neither
478 memmove() nor bcopy(). I didn't think anyone used it these days, but that
479 is clearly not the case, as these two bugs were recently reported.
480
481 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
482 and Detrail to create the HTML documentation, the .txt form of the man
483 pages, and it removes trailing spaces from listed files. It also creates
484 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
485 case, it wraps all the #defines with #ifndefs. This script should be run
486 before "make dist".
487
488 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
489 with Unicode property support.
490
491 (a) For a maximizing quantifier, if the two different cases of the
492 character were of different lengths in their UTF-8 codings (there are
493 some cases like this - I found 11), and the matching function had to
494 back up over a mixture of the two cases, it incorrectly assumed they
495 were both the same length.
496
497 (b) When PCRE was configured to use the heap rather than the stack for
498 recursion during matching, it was not correctly preserving the data for
499 the other case of a UTF-8 character when checking ahead for a match
500 while processing a minimizing repeat. If the check also involved
501 matching a wide character, but failed, corruption could cause an
502 erroneous result when trying to check for a repeat of the original
503 character.
504
505 13. Some tidying changes to the testing mechanism:
506
507 (a) The RunTest script now detects the internal link size and whether there
508 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
509 values substituted by "configure". (The RunGrepTest script already did
510 this for UTF-8.) The configure.ac script no longer substitutes the
511 relevant variables.
512
513 (b) The debugging options /B and /D in pcretest show the compiled bytecode
514 with length and offset values. This means that the output is different
515 for different internal link sizes. Test 2 is skipped for link sizes
516 other than 2 because of this, bypassing the problem. Unfortunately,
517 there was also a test in test 3 (the locale tests) that used /B and
518 failed for link sizes other than 2. Rather than cut the whole test out,
519 I have added a new /Z option to pcretest that replaces the length and
520 offset values with spaces. This is now used to make test 3 independent
521 of link size. (Test 2 will be tidied up later.)
522
523 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
524 segmentation fault instead of returning the appropriate error message.
525
526 15. In multiline mode when the newline sequence was set to "any", the pattern
527 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
528 This doesn't seem right; it now treats the CRLF combination as the line
529 ending, and so does not match in that case. It's only a pattern such as ^$
530 that would hit this one: something like ^ABC$ would have failed after \r
531 and then tried again after \r\n.
532
533 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
534 in an attempt to make files that differ only in their line terminators
535 compare equal. This works on Linux.
536
537 17. Under certain error circumstances pcregrep might try to free random memory
538 as it exited. This is now fixed, thanks to valgrind.
539
540 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
541 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
542 was because its rules for how to advance for /g after matching an empty
543 string at the end of a line did not allow for this case. They now check for
544 it specially.
545
546 20. pcretest is supposed to handle patterns and data of any length, by
547 extending its buffers when necessary. It was getting this wrong when the
548 buffer for a data line had to be extended.
549
550 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
551 CRLF as a newline sequence.
552
553 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
554 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
555 I have nevertheless tidied it up.
556
557 23. Added some casts to kill warnings from HP-UX ia64 compiler.
558
559 24. Added a man page for pcre-config.
560
561
562 Version 7.0 19-Dec-06
563 ---------------------
564
565 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
566 moving to gcc 4.1.1.
567
568 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
569 sys/time.h, which is documented as needed for this function. It doesn't
570 seem to matter on Linux, but it showed up on some releases of OS X.
571
572 3. It seems that there are systems where bytes whose values are greater than
573 127 match isprint() in the "C" locale. The "C" locale should be the
574 default when a C program starts up. In most systems, only ASCII printing
575 characters match isprint(). This difference caused the output from pcretest
576 to vary, making some of the tests fail. I have changed pcretest so that:
577
578 (a) When it is outputting text in the compiled version of a pattern, bytes
579 other than 32-126 are always shown as hex escapes.
580
581 (b) When it is outputting text that is a matched part of a subject string,
582 it does the same, unless a different locale has been set for the match
583 (using the /L modifier). In this case, it uses isprint() to decide.
584
585 4. Fixed a major bug that caused incorrect computation of the amount of memory
586 required for a compiled pattern when options that changed within the
587 pattern affected the logic of the preliminary scan that determines the
588 length. The relevant options are -x, and -i in UTF-8 mode. The result was
589 that the computed length was too small. The symptoms of this bug were
590 either the PCRE error "internal error: code overflow" from pcre_compile(),
591 or a glibc crash with a message such as "pcretest: free(): invalid next
592 size (fast)". Examples of patterns that provoked this bug (shown in
593 pcretest format) are:
594
595 /(?-x: )/x
596 /(?x)(?-x: \s*#\s*)/
597 /((?i)[\x{c0}])/8
598 /(?i:[\x{c0}])/8
599
600 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
601 is now done differently.
602
603 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
604 wrapper classes; (b) implement a new function in the C++ scanner that is
605 more efficient than the old way of doing things because it avoids levels of
606 recursion in the regex matching; (c) add a paragraph to the documentation
607 for the FullMatch() function.
608
609 6. The escape sequence \n was being treated as whatever was defined as
610 "newline". Not only was this contrary to the documentation, which states
611 that \n is character 10 (hex 0A), but it also went horribly wrong when
612 "newline" was defined as CRLF. This has been fixed.
613
614 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
615 was being set to -1 for the "end of line" case (supposedly a value that no
616 character can have). Though this value is never used (the check for end of
617 line is "zero bytes in current character"), it caused compiler complaints.
618 I've changed it to 0xffffffff.
619
620 8. In pcre_version.c, the version string was being built by a sequence of
621 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
622 string (as it is for production releases) called a macro with an empty
623 argument. The C standard says the result of this is undefined. The gcc
624 compiler treats it as an empty string (which was what was wanted) but it is
625 reported that Visual C gives an error. The source has been hacked around to
626 avoid this problem.
627
628 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
629 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
630 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
631 of them did).
632
633 10. Originally, pcretest opened its input and output without "b"; then I was
634 told that "b" was needed in some environments, so it was added for release
635 5.0 to both the input and output. (It makes no difference on Unix-like
636 systems.) Later I was told that it is wrong for the input on Windows. I've
637 now abstracted the modes into two macros, to make it easier to fiddle with
638 them, and removed "b" from the input mode under Windows.
639
640 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
641
642 12. Added -help and --help to pcretest as an official way of being reminded
643 of the options.
644
645 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
646 and pcrecpp.cc because they annoy compilers at high warning levels.
647
648 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
649
650 15. Fixed an occurrence of == in configure.ac that should have been = (shell
651 scripts are not C programs :-) and which was not noticed because it works
652 on Linux.
653
654 16. pcretest is supposed to handle any length of pattern and data line (as one
655 line or as a continued sequence of lines) by extending its input buffer if
656 necessary. This feature was broken for very long pattern lines, leading to
657 a string of junk being passed to pcre_compile() if the pattern was longer
658 than about 50K.
659
660 17. I have done a major re-factoring of the way pcre_compile() computes the
661 amount of memory needed for a compiled pattern. Previously, there was code
662 that made a preliminary scan of the pattern in order to do this. That was
663 OK when PCRE was new, but as the facilities have expanded, it has become
664 harder and harder to keep it in step with the real compile phase, and there
665 have been a number of bugs (see for example, 4 above). I have now found a
666 cunning way of running the real compile function in a "fake" mode that
667 enables it to compute how much memory it would need, while actually only
668 ever using a few hundred bytes of working memory and without too many
669 tests of the mode. This should make future maintenance and development
670 easier. A side effect of this work is that the limit of 200 on the nesting
671 depth of parentheses has been removed (though this was never a serious
672 limitation, I suspect). However, there is a downside: pcre_compile() now
673 runs more slowly than before (30% or more, depending on the pattern). I
674 hope this isn't a big issue. There is no effect on runtime performance.
675
676 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
677 newline (only possible for the last line of a file) and it was a
678 pattern that set a locale (followed by /Lsomething), pcretest crashed.
679
680 19. Added additional timing features to pcretest. (1) The -tm option now times
681 matching only, not compiling. (2) Both -t and -tm can be followed, as a
682 separate command line item, by a number that specifies the number of
683 repeats to use when timing. The default is 50000; this gives better
684 precision, but takes uncomfortably long for very large patterns.
685
686 20. Extended pcre_study() to be more clever in cases where a branch of a
687 subpattern has no definite first character. For example, (a*|b*)[cd] would
688 previously give no result from pcre_study(). Now it recognizes that the
689 first character must be a, b, c, or d.
690
691 21. There was an incorrect error "recursive call could loop indefinitely" if
692 a subpattern (or the entire pattern) that was being tested for matching an
693 empty string contained only one non-empty item after a nested subpattern.
694 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
695 incorrectly, because the \d was being skipped in the check.
696
697 22. The pcretest program now has a new pattern option /B and a command line
698 option -b, which is equivalent to adding /B to every pattern. This causes
699 it to show the compiled bytecode, without the additional information that
700 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
701 is the same as /B/I).
702
703 23. A new optimization is now able automatically to treat some sequences such
704 as a*b as a*+b. More specifically, if something simple (such as a character
705 or a simple class like \d) has an unlimited quantifier, and is followed by
706 something that cannot possibly match the quantified thing, the quantifier
707 is automatically "possessified".
708
709 24. A recursive reference to a subpattern whose number was greater than 39
710 went wrong under certain circumstances in UTF-8 mode. This bug could also
711 have affected the operation of pcre_study().
712
713 25. Realized that a little bit of performance could be had by replacing
714 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
715
716 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
717
718 27. Possessive quantifiers such as a++ were previously implemented by turning
719 them into atomic groups such as ($>a+). Now they have their own opcodes,
720 which improves performance. This includes the automatically created ones
721 from 23 above.
722
723 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
724 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
725 the first matched character to be a colon. This applied both to named and
726 numbered groups.
727
728 29. The ucpinternal.h header file was missing its idempotency #ifdef.
729
730 30. I was sent a "project" file called libpcre.a.dev which I understand makes
731 building PCRE on Windows easier, so I have included it in the distribution.
732
733 31. There is now a check in pcretest against a ridiculously large number being
734 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
735 loop, the loop is abandoned.
736
737 32. Forward references to subpatterns in conditions such as (?(2)...) where
738 subpattern 2 is defined later cause pcre_compile() to search forwards in
739 the pattern for the relevant set of parentheses. This search went wrong
740 when there were unescaped parentheses in a character class, parentheses
741 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
742
743 33. "Subroutine" calls and backreferences were previously restricted to
744 referencing subpatterns earlier in the regex. This restriction has now
745 been removed.
746
747 34. Added a number of extra features that are going to be in Perl 5.10. On the
748 whole, these are just syntactic alternatives for features that PCRE had
749 previously implemented using the Python syntax or my own invention. The
750 other formats are all retained for compatibility.
751
752 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
753 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
754 also .NET compatible.
755
756 (b) A recursion or subroutine call to a named group can now be defined as
757 (?&name) as well as (?P>name).
758
759 (c) A backreference to a named group can now be defined as \k<name> or
760 \k'name' as well as (?P=name). The new forms, as well as being in Perl
761 5.10, are also .NET compatible.
762
763 (d) A conditional reference to a named group can now use the syntax
764 (?(<name>) or (?('name') as well as (?(name).
765
766 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
767 groups (named and numbered) that are never evaluated inline, but can be
768 called as "subroutines" from elsewhere. In effect, the DEFINE condition
769 is always false. There may be only one alternative in such a group.
770
771 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
772 as the simple (?(R). The condition is true only if the most recent
773 recursion is that of the given number or name. It does not search out
774 through the entire recursion stack.
775
776 (g) The escape \gN or \g{N} has been added, where N is a positive or
777 negative number, specifying an absolute or relative reference.
778
779 35. Tidied to get rid of some further signed/unsigned compiler warnings and
780 some "unreachable code" warnings.
781
782 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
783 things, this adds five new scripts.
784
785 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
786 There were also incompatibilities regarding the handling of \Q..\E inside
787 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
788 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
789
790 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
791 matches an empty string, and forcibly breaks the loop. There were bugs in
792 this code in non-simple cases. For a pattern such as ^(a()*)* matched
793 against aaaa the result was just "a" rather than "aaaa", for example. Two
794 separate and independent bugs (that affected different cases) have been
795 fixed.
796
797 39. Refactored the code to abolish the use of different opcodes for small
798 capturing bracket numbers. This is a tidy that I avoided doing when I
799 removed the limit on the number of capturing brackets for 3.5 back in 2001.
800 The new approach is not only tidier, it makes it possible to reduce the
801 memory needed to fix the previous bug (38).
802
803 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
804 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
805 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
806 mode.
807
808 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
809 report.
810
811 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
812 copy construction and assignment in the C++ wrapper.
813
814 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
815 couple of bugs that could have given wrong results in the "--newline=crlf"
816 case.
817
818 44. Added a number of casts and did some reorganization of signed/unsigned int
819 variables following suggestions from Dair Grant. Also renamed the variable
820 "this" as "item" because it is a C++ keyword.
821
822 45. Arranged for dftables to add
823
824 #include "pcre_internal.h"
825
826 to pcre_chartables.c because without it, gcc 4.x may remove the array
827 definition from the final binary if PCRE is built into a static library and
828 dead code stripping is activated.
829
830 46. For an unanchored pattern, if a match attempt fails at the start of a
831 newline sequence, and the newline setting is CRLF or ANY, and the next two
832 characters are CRLF, advance by two characters instead of one.
833
834
835 Version 6.7 04-Jul-06
836 ---------------------
837
838 1. In order to handle tests when input lines are enormously long, pcretest has
839 been re-factored so that it automatically extends its buffers when
840 necessary. The code is crude, but this _is_ just a test program. The
841 default size has been increased from 32K to 50K.
842
843 2. The code in pcre_study() was using the value of the re argument before
844 testing it for NULL. (Of course, in any sensible call of the function, it
845 won't be NULL.)
846
847 3. The memmove() emulation function in pcre_internal.h, which is used on
848 systems that lack both memmove() and bcopy() - that is, hardly ever -
849 was missing a "static" storage class specifier.
850
851 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
852 containing an extended class (one that cannot be represented by a bitmap
853 because it contains high-valued characters or Unicode property items, e.g.
854 [\pZ]). Almost always one would set UTF-8 mode when processing such a
855 pattern, but PCRE should not loop if you do not (it no longer does).
856 [Detail: two cases were found: (a) a repeated subpattern containing an
857 extended class; (b) a recursive reference to a subpattern that followed a
858 previous extended class. It wasn't skipping over the extended class
859 correctly when UTF-8 mode was not set.]
860
861 5. A negated single-character class was not being recognized as fixed-length
862 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
863 compile error "lookbehind assertion is not fixed length".
864
865 6. The RunPerlTest auxiliary script was showing an unexpected difference
866 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
867 write a Perl script that can interpret lines of an input file either as
868 byte characters or as UTF-8, which is what "perltest" was being required to
869 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
870 can't do is switch easily at run time between having the "use utf8;" pragma
871 or not. In the end, I fudged it by using the RunPerlTest script to insert
872 "use utf8;" explicitly for the UTF-8 tests.
873
874 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
875 the end of the subject string, contrary to the documentation and to what
876 Perl does. This was true of both matching functions. Now it matches only at
877 the start of the subject and immediately after *internal* newlines.
878
879 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
880 a pointer to an int instead of a pointer to an unsigned long int. This
881 caused problems on 64-bit systems.
882
883 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
884 instance of the 'standard' template library not being so standard".
885
886 10. There was no check on the number of named subpatterns nor the maximum
887 length of a subpattern name. The product of these values is used to compute
888 the size of the memory block for a compiled pattern. By supplying a very
889 long subpattern name and a large number of named subpatterns, the size
890 computation could be caused to overflow. This is now prevented by limiting
891 the length of names to 32 characters, and the number of named subpatterns
892 to 10,000.
893
894 11. Subpatterns that are repeated with specific counts have to be replicated in
895 the compiled pattern. The size of memory for this was computed from the
896 length of the subpattern and the repeat count. The latter is limited to
897 65535, but there was no limit on the former, meaning that integer overflow
898 could in principle occur. The compiled length of a repeated subpattern is
899 now limited to 30,000 bytes in order to prevent this.
900
901 12. Added the optional facility to have named substrings with the same name.
902
903 13. Added the ability to use a named substring as a condition, using the
904 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
905 are numbers (not recommended). Forward references are permitted.
906
907 14. Added forward references in named backreferences (if you see what I mean).
908
909 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
910 pattern could run off the end of the subject. For example, the pattern
911 "(?s)(.{1,5})"8 did this with the subject "ab".
912
913 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
914 PCRE_CASELESS was set when matching characters that were quantified with ?
915 or *.
916
917 17. A character class other than a single negated character that had a minimum
918 but no maximum quantifier - for example [ab]{6,} - was not handled
919 correctly by pce_dfa_exec(). It would match only one character.
920
921 18. A valid (though odd) pattern that looked like a POSIX character
922 class but used an invalid character after [ (for example [[,abc,]]) caused
923 pcre_compile() to give the error "Failed: internal error: code overflow" or
924 in some cases to crash with a glibc free() error. This could even happen if
925 the pattern terminated after [[ but there just happened to be a sequence of
926 letters, a binary zero, and a closing ] in the memory that followed.
927
928 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
929 over the years. Originally (before any Unicode support), just the bottom 8
930 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
931 output from "man perlunicode" includes this:
932
933 The regular expression compiler produces polymorphic opcodes. That
934 is, the pattern adapts to the data and automatically switches to
935 the Unicode character scheme when presented with Unicode data--or
936 instead uses a traditional byte scheme when presented with byte
937 data.
938
939 Sadly, a wide octal escape does not cause a switch, and in a string with
940 no other multibyte characters, these octal escapes are treated as before.
941 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
942 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
943 Unicode string.
944
945 I have not perpetrated such confusion in PCRE. Up till now, it took just
946 the bottom 8 bits, as in old Perl. I have now made octal escapes with
947 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
948 translate to the appropriate multibyte character.
949
950 29. Applied some refactoring to reduce the number of warnings from Microsoft
951 and Borland compilers. This has included removing the fudge introduced
952 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
953 a warning about an unused variable.
954
955 21. PCRE has not included VT (character 0x0b) in the set of whitespace
956 characters since release 4.0, because Perl (from release 5.004) does not.
957 [Or at least, is documented not to: some releases seem to be in conflict
958 with the documentation.] However, when a pattern was studied with
959 pcre_study() and all its branches started with \s, PCRE still included VT
960 as a possible starting character. Of course, this did no harm; it just
961 caused an unnecessary match attempt.
962
963 22. Removed a now-redundant internal flag bit that recorded the fact that case
964 dependency changed within the pattern. This was once needed for "required
965 byte" processing, but is no longer used. This recovers a now-scarce options
966 bit. Also moved the least significant internal flag bit to the most-
967 significant bit of the word, which was not previously used (hangover from
968 the days when it was an int rather than a uint) to free up another bit for
969 the future.
970
971 23. Added support for CRLF line endings as well as CR and LF. As well as the
972 default being selectable at build time, it can now be changed at runtime
973 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
974 specify that it is scanning data with non-default line endings.
975
976 24. Changed the definition of CXXLINK to make it agree with the definition of
977 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
978
979 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
980 recursions. This makes a big different to stack usage for some patterns.
981
982 26. If a subpattern containing a named recursion or subroutine reference such
983 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
984 the space required for the compiled pattern went wrong and gave too small a
985 value. Depending on the environment, this could lead to "Failed: internal
986 error: code overflow at offset 49" or "glibc detected double free or
987 corruption" errors.
988
989 27. Applied patches from Google (a) to support the new newline modes and (b) to
990 advance over multibyte UTF-8 characters in GlobalReplace.
991
992 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
993 difference for some implementation of PCRE in some Windows version.
994
995 29. Added some extra testing facilities to pcretest:
996
997 \q<number> in a data line sets the "match limit" value
998 \Q<number> in a data line sets the "match recursion limt" value
999 -S <number> sets the stack size, where <number> is in megabytes
1000
1001 The -S option isn't available for Windows.
1002
1003
1004 Version 6.6 06-Feb-06
1005 ---------------------
1006
1007 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1008 in pcreposix.h. I have copied the definition from pcre.h.
1009
1010 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1011 because pcre.h is no longer a built file.
1012
1013 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1014 not normally included in the compiled code.
1015
1016
1017 Version 6.5 01-Feb-06
1018 ---------------------
1019
1020 1. When using the partial match feature with pcre_dfa_exec(), it was not
1021 anchoring the second and subsequent partial matches at the new starting
1022 point. This could lead to incorrect results. For example, with the pattern
1023 /1234/, partially matching against "123" and then "a4" gave a match.
1024
1025 2. Changes to pcregrep:
1026
1027 (a) All non-match returns from pcre_exec() were being treated as failures
1028 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1029 error message is output. Some extra information is given for the
1030 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1031 probably the only errors that are likely to be caused by users (by
1032 specifying a regex that has nested indefinite repeats, for instance).
1033 If there are more than 20 of these errors, pcregrep is abandoned.
1034
1035 (b) A binary zero was treated as data while matching, but terminated the
1036 output line if it was written out. This has been fixed: binary zeroes
1037 are now no different to any other data bytes.
1038
1039 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1040 used to set a locale for matching. The --locale=xxxx long option has
1041 been added (no short equivalent) to specify a locale explicitly on the
1042 pcregrep command, overriding the environment variables.
1043
1044 (d) When -B was used with -n, some line numbers in the output were one less
1045 than they should have been.
1046
1047 (e) Added the -o (--only-matching) option.
1048
1049 (f) If -A or -C was used with -c (count only), some lines of context were
1050 accidentally printed for the final match.
1051
1052 (g) Added the -H (--with-filename) option.
1053
1054 (h) The combination of options -rh failed to suppress file names for files
1055 that were found from directory arguments.
1056
1057 (i) Added the -D (--devices) and -d (--directories) options.
1058
1059 (j) Added the -F (--fixed-strings) option.
1060
1061 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1062
1063 (l) Added the --colo(u)r option.
1064
1065 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1066 is not present by default.
1067
1068 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1069 items such as (?R) or (?1), when the recursion could match a number of
1070 alternatives. If it matched one of the alternatives, but subsequently,
1071 outside the recursion, there was a failure, the code tried to back up into
1072 the recursion. However, because of the way PCRE is implemented, this is not
1073 possible, and the result was an incorrect result from the match.
1074
1075 In order to prevent this happening, the specification of recursion has
1076 been changed so that all such subpatterns are automatically treated as
1077 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1078
1079 4. I had overlooked the fact that, in some locales, there are characters for
1080 which isalpha() is true but neither isupper() nor islower() are true. In
1081 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1082 and ordfeminine) are like this. This affected the treatment of \w and \W
1083 when they appeared in character classes, but not when they appeared outside
1084 a character class. The bit map for "word" characters is now created
1085 separately from the results of isalnum() instead of just taking it from the
1086 upper, lower, and digit maps. (Plus the underscore character, of course.)
1087
1088 5. The above bug also affected the handling of POSIX character classes such as
1089 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1090 permanent tables. Instead, the bit maps for such a class were previously
1091 created as the appropriate unions of the upper, lower, and digit bitmaps.
1092 Now they are created by subtraction from the [[:word:]] class, which has
1093 its own bitmap.
1094
1095 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1096 It is created by subtracting the vertical space characters (\x09, \x0a,
1097 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1098 subtraction was done in the overall bitmap for a character class, meaning
1099 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1100 be recognized. This bug has been fixed.
1101
1102 7. Patches from the folks at Google:
1103
1104 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1105 real life, but is still worth protecting against".
1106
1107 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1108 regular expressions".
1109
1110 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1111 have it.
1112
1113 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1114 "configure" and the latter not, in order to fix a problem somebody had
1115 with compiling the Arg class on HP-UX.
1116
1117 (e) Improve the error-handling of the C++ wrapper a little bit.
1118
1119 (f) New tests for checking recursion limiting.
1120
1121 8. The pcre_memmove() function, which is used only if the environment does not
1122 have a standard memmove() function (and is therefore rarely compiled),
1123 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1124 returning a result (though PCRE never actually uses the result).
1125
1126 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1127 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1128 returned instead of calling malloc() with an overflowing number that would
1129 most likely cause subsequent chaos.
1130
1131 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1132
1133 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1134 with this option is matched, the nmatch and pmatch options of regexec() are
1135 ignored.
1136
1137 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1138 provided in case anyone wants to the the POSIX interface with UTF-8
1139 strings.
1140
1141 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1142 C++ linking (needed for some HP-UX environments).
1143
1144 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1145 (unused parameter) and in the pcre_printint() function (omitted "default"
1146 switch label when the default is to do nothing).
1147
1148 15. Added some code to make it possible, when PCRE is compiled as a C++
1149 library, to replace subject pointers for pcre_exec() with a smart pointer
1150 class, thus making it possible to process discontinuous strings.
1151
1152 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1153 much the same function. They were added by different people who were trying
1154 to make PCRE easy to compile on non-Unix systems. It has been suggested
1155 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1156 for compiling on Windows systems. I have therefore replaced it with
1157 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1158 defaults to "extern" for C or "extern C" for C++, which works fine on
1159 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1160 SCOPE with something explicit in config.h. In addition:
1161
1162 (a) pcreposix.h still had just "extern" instead of either of these macros;
1163 I have replaced it with PCRE_DATA_SCOPE.
1164
1165 (b) Functions such as _pcre_xclass(), which are internal to the library,
1166 but external in the C sense, all had PCRE_EXPORT in their definitions.
1167 This is apparently wrong for the Windows case, so I have removed it.
1168 (It makes no difference on Unix-like systems.)
1169
1170 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1171 of recursive calls to match(). This is different to MATCH_LIMIT because
1172 that limits the total number of calls to match(), not all of which increase
1173 the depth of recursion. Limiting the recursion depth limits the amount of
1174 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1175 when PCRE is compiled, and changed at run time. A patch from Google adds
1176 this functionality to the C++ interface.
1177
1178 18. Changes to the handling of Unicode character properties:
1179
1180 (a) Updated the table to Unicode 4.1.0.
1181
1182 (b) Recognize characters that are not in the table as "Cn" (undefined).
1183
1184 (c) I revised the way the table is implemented to a much improved format
1185 which includes recognition of ranges. It now supports the ranges that
1186 are defined in UnicodeData.txt, and it also amalgamates other
1187 characters into ranges. This has reduced the number of entries in the
1188 table from around 16,000 to around 3,000, thus reducing its size
1189 considerably. I realized I did not need to use a tree structure after
1190 all - a binary chop search is just as efficient. Having reduced the
1191 number of entries, I extended their size from 6 bytes to 8 bytes to
1192 allow for more data.
1193
1194 (d) Added support for Unicode script names via properties such as \p{Han}.
1195
1196 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1197 matching that character.
1198
1199 20. When matching a repeated Unicode property with a minimum greater than zero,
1200 (for example \pL{2,}), PCRE could look past the end of the subject if it
1201 reached it while seeking the minimum number of characters. This could
1202 happen only if some of the characters were more than one byte long, because
1203 there is a check for at least the minimum number of bytes.
1204
1205 21. Refactored the implementation of \p and \P so as to be more general, to
1206 allow for more different types of property in future. This has changed the
1207 compiled form incompatibly. Anybody with saved compiled patterns that use
1208 \p or \P will have to recompile them.
1209
1210 22. Added "Any" and "L&" to the supported property types.
1211
1212 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1213 but give a compile time error if the value is greater than 0xff.
1214
1215 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1216 accidentally not being installed or uninstalled.
1217
1218 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1219 made were to insert the current release number. This seemed silly, because
1220 it made things harder for people building PCRE on systems that don't run
1221 "configure". I have turned pcre.h into a distributed file, no longer built
1222 by "configure", with the version identification directly included. There is
1223 no longer a pcre.h.in file.
1224
1225 However, this change necessitated a change to the pcre-config script as
1226 well. It is built from pcre-config.in, and one of the substitutions was the
1227 release number. I have updated configure.ac so that ./configure now finds
1228 the release number by grepping pcre.h.
1229
1230 26. Added the ability to run the tests under valgrind.
1231
1232
1233 Version 6.4 05-Sep-05
1234 ---------------------
1235
1236 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1237 "--" to be printed when multiple files were scanned, even when none of the
1238 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1239 consider it to be a bug, and have restored the previous behaviour.
1240
1241 2. A couple of code tidies to get rid of compiler warnings.
1242
1243 3. The pcretest program used to cheat by referring to symbols in the library
1244 whose names begin with _pcre_. These are internal symbols that are not
1245 really supposed to be visible externally, and in some environments it is
1246 possible to suppress them. The cheating is now confined to including
1247 certain files from the library's source, which is a bit cleaner.
1248
1249 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1250 file's purpose clearer.
1251
1252 5. Reorganized pcre_ucp_findchar().
1253
1254
1255 Version 6.3 15-Aug-05
1256 ---------------------
1257
1258 1. The file libpcre.pc.in did not have general read permission in the tarball.
1259
1260 2. There were some problems when building without C++ support:
1261
1262 (a) If C++ support was not built, "make install" and "make test" still
1263 tried to test it.
1264
1265 (b) There were problems when the value of CXX was explicitly set. Some
1266 changes have been made to try to fix these, and ...
1267
1268 (c) --disable-cpp can now be used to explicitly disable C++ support.
1269
1270 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1271 backslash in a target when C++ was disabled. This confuses some
1272 versions of "make", apparently. Using an intermediate variable solves
1273 this. (Same for CPP_LOBJ.)
1274
1275 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1276 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1277 necessary on certain architectures.
1278
1279 4. Added a setting of -export-symbols-regex to the link command to remove
1280 those symbols that are exported in the C sense, but actually are local
1281 within the library, and not documented. Their names all begin with
1282 "_pcre_". This is not a perfect job, because (a) we have to except some
1283 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1284 available (and never for static libraries). I have made a note to try to
1285 find a way round (a) in the future.
1286
1287
1288 Version 6.2 01-Aug-05
1289 ---------------------
1290
1291 1. There was no test for integer overflow of quantifier values. A construction
1292 such as {1111111111111111} would give undefined results. What is worse, if
1293 a minimum quantifier for a parenthesized subpattern overflowed and became
1294 negative, the calculation of the memory size went wrong. This could have
1295 led to memory overwriting.
1296
1297 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1298
1299 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1300 operating environments where this matters.
1301
1302 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1303 PCRE options from within the C++ wrapper.
1304
1305 5. Named capturing subpatterns were not being correctly counted when a pattern
1306 was compiled. This caused two problems: (a) If there were more than 100
1307 such subpatterns, the calculation of the memory needed for the whole
1308 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1309 back references of the form \12, where the number was greater than 9, were
1310 not recognized as back references, even though there were sufficient
1311 previous subpatterns.
1312
1313 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1314 versions of gcc, e.g. 2.95.4.
1315
1316
1317 Version 6.1 21-Jun-05
1318 ---------------------
1319
1320 1. There was one reference to the variable "posix" in pcretest.c that was not
1321 surrounded by "#if !defined NOPOSIX".
1322
1323 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1324 the cross-check on the old pcre_info() function, for the benefit of the
1325 cut-down version of PCRE that is currently imported into Exim.
1326
1327 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1328 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1329 patterns that start (?i) but I don't think that matters. The use of (?i) is
1330 just an example; this all applies to the other options as well.
1331
1332 4. Since libtool seems to echo the compile commands it is issuing, the output
1333 from "make" can be reduced a bit by putting "@" in front of each libtool
1334 compile command.
1335
1336 5. Patch from the folks at Google for configure.in to be a bit more thorough
1337 in checking for a suitable C++ installation before trying to compile the
1338 C++ stuff. This should fix a reported problem when a compiler was present,
1339 but no suitable headers.
1340
1341 6. The man pages all had just "PCRE" as their title. I have changed them to
1342 be the relevant file name. I have also arranged that these names are
1343 retained in the file doc/pcre.txt, which is a concatenation in text format
1344 of all the man pages except the little individual ones for each function.
1345
1346 7. The NON-UNIX-USE file had not been updated for the different set of source
1347 files that come with release 6. I also added a few comments about the C++
1348 wrapper.
1349
1350
1351 Version 6.0 07-Jun-05
1352 ---------------------
1353
1354 1. Some minor internal re-organization to help with my DFA experiments.
1355
1356 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1357 didn't matter for the library itself when fully configured, but did matter
1358 when compiling without UCP support, or within Exim, where the ucp files are
1359 not imported.
1360
1361 3. Refactoring of the library code to split up the various functions into
1362 different source modules. The addition of the new DFA matching code (see
1363 below) to a single monolithic source would have made it really too
1364 unwieldy, quite apart from causing all the code to be include in a
1365 statically linked application, when only some functions are used. This is
1366 relevant even without the DFA addition now that patterns can be compiled in
1367 one application and matched in another.
1368
1369 The downside of splitting up is that there have to be some external
1370 functions and data tables that are used internally in different modules of
1371 the library but which are not part of the API. These have all had their
1372 names changed to start with "_pcre_" so that they are unlikely to clash
1373 with other external names.
1374
1375 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1376 a different (DFA) algorithm. Although it is slower than the original
1377 function, it does have some advantages for certain types of matching
1378 problem.
1379
1380 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1381 including restarting after a partial match.
1382
1383 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1384 defined when compiling for Windows was sent to me. I have put it into the
1385 code, though I have no means of testing or verifying it.
1386
1387 7. Added the pcre_refcount() auxiliary function.
1388
1389 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1390 match before or at the first newline in the subject string. In pcretest,
1391 the /f option on a pattern can be used to set this.
1392
1393 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1394 would behave wrongly. This has been present in PCRE since release 4.0.
1395
1396 10. A number of changes to the pcregrep command:
1397
1398 (a) Refactored how -x works; insert ^(...)$ instead of setting
1399 PCRE_ANCHORED and checking the length, in preparation for adding
1400 something similar for -w.
1401
1402 (b) Added the -w (match as a word) option.
1403
1404 (c) Refactored the way lines are read and buffered so as to have more
1405 than one at a time available.
1406
1407 (d) Implemented a pcregrep test script.
1408
1409 (e) Added the -M (multiline match) option. This allows patterns to match
1410 over several lines of the subject. The buffering ensures that at least
1411 8K, or the rest of the document (whichever is the shorter) is available
1412 for matching (and similarly the previous 8K for lookbehind assertions).
1413
1414 (f) Changed the --help output so that it now says
1415
1416 -w, --word-regex(p)
1417
1418 instead of two lines, one with "regex" and the other with "regexp"
1419 because that confused at least one person since the short forms are the
1420 same. (This required a bit of code, as the output is generated
1421 automatically from a table. It wasn't just a text change.)
1422
1423 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1424 option but starts with a hyphen. Could be a pattern or a path name
1425 starting with a hyphen, for instance.
1426
1427 (h) "-" can be given as a file name to represent stdin.
1428
1429 (i) When file names are being printed, "(standard input)" is used for
1430 the standard input, for compatibility with GNU grep. Previously
1431 "<stdin>" was used.
1432
1433 (j) The option --label=xxx can be used to supply a name to be used for
1434 stdin when file names are being printed. There is no short form.
1435
1436 (k) Re-factored the options decoding logic because we are going to add
1437 two more options that take data. Such options can now be given in four
1438 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1439
1440 (l) Added the -A, -B, and -C options for requesting that lines of context
1441 around matches be printed.
1442
1443 (m) Added the -L option to print the names of files that do not contain
1444 any matching lines, that is, the complement of -l.
1445
1446 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1447 continue to scan other files.
1448
1449 (o) The -s option was incorrectly implemented. For compatibility with other
1450 greps, it now suppresses the error message for a non-existent or non-
1451 accessible file (but not the return code). There is a new option called
1452 -q that suppresses the output of matching lines, which was what -s was
1453 previously doing.
1454
1455 (p) Added --include and --exclude options to specify files for inclusion
1456 and exclusion when recursing.
1457
1458 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1459 Hopefully, it now does.
1460
1461 12. Missing cast in pcre_study().
1462
1463 13. Added an "uninstall" target to the makefile.
1464
1465 14. Replaced "extern" in the function prototypes in Makefile.in with
1466 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1467 world, but is set differently for Windows.
1468
1469 15. Added a second compiling function called pcre_compile2(). The only
1470 difference is that it has an extra argument, which is a pointer to an
1471 integer error code. When there is a compile-time failure, this is set
1472 non-zero, in addition to the error test pointer being set to point to an
1473 error message. The new argument may be NULL if no error number is required
1474 (but then you may as well call pcre_compile(), which is now just a
1475 wrapper). This facility is provided because some applications need a
1476 numeric error indication, but it has also enabled me to tidy up the way
1477 compile-time errors are handled in the POSIX wrapper.
1478
1479 16. Added VPATH=.libs to the makefile; this should help when building with one
1480 prefix path and installing with another. (Or so I'm told by someone who
1481 knows more about this stuff than I do.)
1482
1483 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1484 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1485 match everything, including newlines. This is not POSIX-compatible, but
1486 somebody wanted the feature. From pcretest it can be activated by using
1487 both the P and the s flags.
1488
1489 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1490
1491 19. libpcre.pc was being incorrectly installed as executable.
1492
1493 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1494 it now also looks for '\r' so that it will work unmodified on Windows.
1495
1496 21. Added Google's contributed C++ wrapper to the distribution.
1497
1498 22. Added some untidy missing memory free() calls in pcretest, to keep
1499 Electric Fence happy when testing.
1500
1501
1502
1503 Version 5.0 13-Sep-04
1504 ---------------------
1505
1506 1. Internal change: literal characters are no longer packed up into items
1507 containing multiple characters in a single byte-string. Each character
1508 is now matched using a separate opcode. However, there may be more than one
1509 byte in the character in UTF-8 mode.
1510
1511 2. The pcre_callout_block structure has two new fields: pattern_position and
1512 next_item_length. These contain the offset in the pattern to the next match
1513 item, and its length, respectively.
1514
1515 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1516 insertion of callouts before each pattern item. Added the /C option to
1517 pcretest to make use of this.
1518
1519 4. On the advice of a Windows user, the lines
1520
1521 #if defined(_WIN32) || defined(WIN32)
1522 _setmode( _fileno( stdout ), 0x8000 );
1523 #endif /* defined(_WIN32) || defined(WIN32) */
1524
1525 have been added to the source of pcretest. This apparently does useful
1526 magic in relation to line terminators.
1527
1528 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1529 for the benefit of those environments where the "b" makes a difference.
1530
1531 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1532 to know about it. I have put a hack into configure.in that adds in code
1533 to set GCC=yes if CC=icc. This seems to end up at a point in the
1534 generated configure script that is early enough to affect the setting of
1535 compiler options, which is what is needed, but I have no means of testing
1536 whether it really works. (The user who reported this had patched the
1537 generated configure script, which of course I cannot do.)
1538
1539 LATER: After change 22 below (new libtool files), the configure script
1540 seems to know about icc (and also ecc). Therefore, I have commented out
1541 this hack in configure.in.
1542
1543 7. Added support for pkg-config (2 patches were sent in).
1544
1545 8. Negated POSIX character classes that used a combination of internal tables
1546 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1547 [[:^ascii]]. Typically, they would match almost any characters. The other
1548 POSIX classes were not broken in this way.
1549
1550 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1551 to find the match, as PCRE was deluded into thinking that the match had to
1552 start at the start point or following a newline. The same bug applied to
1553 patterns with negative forward assertions or any backward assertions
1554 preceding ".*" at the start, unless the pattern required a fixed first
1555 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1556
1557 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1558 starting at the last subject character, bytes beyond the end of the subject
1559 string were read.
1560
1561 11. Renamed the variable "class" as "classbits" to make life easier for C++
1562 users. (Previously there was a macro definition, but it apparently wasn't
1563 enough.)
1564
1565 12. Added the new field "tables" to the extra data so that tables can be passed
1566 in at exec time, or the internal tables can be re-selected. This allows
1567 a compiled regex to be saved and re-used at a later time by a different
1568 program that might have everything at different addresses.
1569
1570 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1571 -R library as well as a -L library.
1572
1573 14. The debugging options of pcretest (-d on the command line or D on a
1574 pattern) showed incorrect output for anything following an extended class
1575 that contained multibyte characters and which was followed by a quantifier.
1576
1577 15. Added optional support for general category Unicode character properties
1578 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1579 support. It adds about 90K to the size of the library. The meanings of the
1580 inbuilt class escapes such as \d and \s have NOT been changed.
1581
1582 16. Updated pcredemo.c to include calls to free() to release the memory for the
1583 compiled pattern.
1584
1585 17. The generated file chartables.c was being created in the source directory
1586 instead of in the building directory. This caused the build to fail if the
1587 source directory was different from the building directory, and was
1588 read-only.
1589
1590 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1591 file. No doubt somebody will tell me if they don't make sense... Also added
1592 Dan Mooney's comments about building on OpenVMS.
1593
1594 19. Added support for partial matching via the PCRE_PARTIAL option for
1595 pcre_exec() and the \P data escape in pcretest.
1596
1597 20. Extended pcretest with 3 new pattern features:
1598
1599 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1600 write the compiled pattern to the file whose name is "rest-of-line".
1601 This is a straight binary dump of the data, with the saved pointer to
1602 the character tables forced to be NULL. The study data, if any, is
1603 written too. After writing, pcretest reads a new pattern.
1604
1605 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1606 compiled pattern from the given file. There must not be any
1607 occurrences of "<" in the file name (pretty unlikely); if there are,
1608 pcretest will instead treat the initial "<" as a pattern delimiter.
1609 After reading in the pattern, pcretest goes on to read data lines as
1610 usual.
1611
1612 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1613 and 16-bit fields in a compiled pattern, to simulate a pattern that
1614 was compiled on a host of opposite endianness.
1615
1616 21. The pcre-exec() function can now cope with patterns that were compiled on
1617 hosts of opposite endianness, with this restriction:
1618
1619 As for any compiled expression that is saved and used later, the tables
1620 pointer field cannot be preserved; the extra_data field in the arguments
1621 to pcre_exec() should be used to pass in a tables address if a value
1622 other than the default internal tables were used at compile time.
1623
1624 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1625 now diagnosed as an error. Previously, most of the time, a negative number
1626 would have been treated as zero, but if in addition "ovector" was passed as
1627 NULL, a crash could occur.
1628
1629 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1630 new versions from the libtool 1.5 distribution (the last one is a copy of
1631 a file called libtool.m4). This seems to have fixed the need to patch
1632 "configure" to support Darwin 1.3 (which I used to do). However, I still
1633 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1634 workstation).
1635
1636 24. Changed the PCRE licence to be the more standard "BSD" licence.
1637
1638
1639 Version 4.5 01-Dec-03
1640 ---------------------
1641
1642 1. There has been some re-arrangement of the code for the match() function so
1643 that it can be compiled in a version that does not call itself recursively.
1644 Instead, it keeps those local variables that need separate instances for
1645 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1646 needs to "recurse". Keeping track of where control must go is done by means
1647 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1648 hide most of the details from the main code, and operates only if
1649 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1650 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1651 operating.
1652
1653 To make it easier for callers to provide specially tailored get/free
1654 functions for this usage, two new functions, pcre_stack_malloc, and
1655 pcre_stack_free, are used. They are always called in strict stacking order,
1656 and the size of block requested is always the same.
1657
1658 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1659 PCRE has been compiled to use the stack or the heap for recursion. The
1660 -C option of pcretest uses this to show which version is compiled.
1661
1662 A new data escape \S, is added to pcretest; it causes the amounts of store
1663 obtained and freed by both kinds of malloc/free at match time to be added
1664 to the output.
1665
1666 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1667 what's available on my current Linux desktop machine.
1668
1669 3. When matching a UTF-8 string, the test for a valid string at the start has
1670 been extended. If start_offset is not zero, PCRE now checks that it points
1671 to a byte that is the start of a UTF-8 character. If not, it returns
1672 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1673 this is necessary because there may be backward assertions in the pattern.
1674 When matching the same subject several times, it may save resources to use
1675 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1676
1677 4. The code for checking the validity of UTF-8 strings has been tightened so
1678 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1679 containing "overlong sequences".
1680
1681 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1682 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1683 should have been "&", but it just so happened that all the cases this let
1684 through by mistake were picked up later in the function.
1685
1686 6. I had used a variable called "isblank" - this is a C99 function, causing
1687 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1688
1689 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1690 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1691 so the version is shown; (c) stop "make test" echoing "./RunTest".
1692
1693 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1694
1695 9. The prototype for memmove() for systems that don't have it was using
1696 size_t, but the inclusion of the header that defines size_t was later. I've
1697 moved the #includes for the C headers earlier to avoid this.
1698
1699 10. Added some adjustments to the code to make it easier to compiler on certain
1700 special systems:
1701
1702 (a) Some "const" qualifiers were missing.
1703 (b) Added the macro EXPORT before all exported functions; by default this
1704 is defined to be empty.
1705 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1706 that it reads its output file name as an argument instead of writing
1707 to the standard output and assuming this can be redirected.
1708
1709 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1710 class containing characters with values greater than 255, PCRE compilation
1711 went into a loop.
1712
1713 12. A recursive reference to a subpattern that was within another subpattern
1714 that had a minimum quantifier of zero caused PCRE to crash. For example,
1715 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1716 recursion. If the recursively-called subpattern itself had a zero repeat,
1717 that was OK.
1718
1719 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1720 buffer into which it was copied (for escape processing) was still set at
1721 1024, so long lines caused crashes.
1722
1723 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1724 "internal error: code overflow...". This applied to any character class
1725 that was followed by a possessive quantifier.
1726
1727 15. Modified the Makefile to add libpcre.la as a prerequisite for
1728 libpcreposix.la because I was told this is needed for a parallel build to
1729 work.
1730
1731 16. If a pattern that contained .* following optional items at the start was
1732 studied, the wrong optimizing data was generated, leading to matching
1733 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1734 matching string must start with a or b or c. The correct conclusion for
1735 this pattern is that a match can start with any character.
1736
1737
1738 Version 4.4 13-Aug-03
1739 ---------------------
1740
1741 1. In UTF-8 mode, a character class containing characters with values between
1742 127 and 255 was not handled correctly if the compiled pattern was studied.
1743 In fixing this, I have also improved the studying algorithm for such
1744 classes (slightly).
1745
1746 2. Three internal functions had redundant arguments passed to them. Removal
1747 might give a very teeny performance improvement.
1748
1749 3. Documentation bug: the value of the capture_top field in a callout is *one
1750 more than* the number of the hightest numbered captured substring.
1751
1752 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1753 in incorrectly linking with a previously installed version. They now link
1754 explicitly with libpcre.la.
1755
1756 5. configure.in no longer needs to recognize Cygwin specially.
1757
1758 6. A problem in pcre.in for Windows platforms is fixed.
1759
1760 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1761 pcretest, it used to include the size of the study block as part of its
1762 output. Unfortunately, the structure contains a field that has a different
1763 size on different hardware architectures. This meant that the tests that
1764 showed this size failed. As the block is currently always of a fixed size,
1765 this information isn't actually particularly useful in pcretest output, so
1766 I have just removed it.
1767
1768 8. Three pre-processor statements accidentally did not start in column 1.
1769 Sadly, there are *still* compilers around that complain, even though
1770 standard C has not required this for well over a decade. Sigh.
1771
1772 9. In pcretest, the code for checking callouts passed small integers in the
1773 callout_data field, which is a void * field. However, some picky compilers
1774 complained about the casts involved for this on 64-bit systems. Now
1775 pcretest passes the address of the small integer instead, which should get
1776 rid of the warnings.
1777
1778 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1779 both compile and run time, and gives an error if an invalid UTF-8 sequence
1780 is found. There is a option for disabling this check in cases where the
1781 string is known to be correct and/or the maximum performance is wanted.
1782
1783 11. In response to a bug report, I changed one line in Makefile.in from
1784
1785 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1786 to
1787 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1788
1789 to look similar to other lines, but I have no way of telling whether this
1790 is the right thing to do, as I do not use Windows. No doubt I'll get told
1791 if it's wrong...
1792
1793
1794 Version 4.3 21-May-03
1795 ---------------------
1796
1797 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1798 Makefile.
1799
1800 2. Some refactoring to improve the quality of the code:
1801
1802 (i) The utf8_table... variables are now declared "const".
1803
1804 (ii) The code for \cx, which used the "case flipping" table to upper case
1805 lower case letters, now just substracts 32. This is ASCII-specific,
1806 but the whole concept of \cx is ASCII-specific, so it seems
1807 reasonable.
1808
1809 (iii) PCRE was using its character types table to recognize decimal and
1810 hexadecimal digits in the pattern. This is silly, because it handles
1811 only 0-9, a-f, and A-F, but the character types table is locale-
1812 specific, which means strange things might happen. A private
1813 table is now used for this - though it costs 256 bytes, a table is
1814 much faster than multiple explicit tests. Of course, the standard
1815 character types table is still used for matching digits in subject
1816 strings against \d.
1817
1818 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1819 ending in _t are). So I've renamed it as ESC_tee.
1820
1821 3. The first argument for regexec() in the POSIX wrapper should have been
1822 defined as "const".
1823
1824 4. Changed pcretest to use malloc() for its buffers so that they can be
1825 Electric Fenced for debugging.
1826
1827 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1828 to read one or more bytes before the start of the subject string. Often this
1829 had no effect on PCRE's behaviour, but in some circumstances it could
1830 provoke a segmentation fault.
1831
1832 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1833 to try to read one or more bytes before the start of the subject string.
1834
1835 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1836 UTF-8 support could misbehave in various ways if the subject string
1837 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1838 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1839 back over UTF-8 characters.)
1840
1841
1842 Version 4.2 14-Apr-03
1843 ---------------------
1844
1845 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1846
1847 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1848 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1849 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1850 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1851 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1852 and BUILD_EXEEXT
1853 Note: automatic setting of the BUILD variables is not yet working
1854 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1855 compile-time but not at link-time
1856 [LINK]: use for linking executables only
1857 make different versions for Windows and non-Windows
1858 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1859 libraries
1860 [LINK_FOR_BUILD]: new variable
1861 [OBJEXT]: use throughout
1862 [EXEEXT]: use throughout
1863 <winshared>: new target
1864 <wininstall>: new target
1865 <dftables.o>: use native compiler
1866 <dftables>: use native linker
1867 <install>: handle Windows platform correctly
1868 <clean>: ditto
1869 <check>: ditto
1870 copy DLL to top builddir before testing
1871
1872 As part of these changes, -no-undefined was removed again. This was reported
1873 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1874 in any case.
1875
1876 3. Some tidies to get rid of compiler warnings:
1877
1878 . In the match_data structure, match_limit was an unsigned long int, whereas
1879 match_call_count was an int. I've made them both unsigned long ints.
1880
1881 . In pcretest the fact that a const uschar * doesn't automatically cast to
1882 a void * provoked a warning.
1883
1884 . Turning on some more compiler warnings threw up some "shadow" variables
1885 and a few more missing casts.
1886
1887 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1888 option, a class that contained a single character with a value between 128
1889 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1890
1891 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1892 option, a class that contained several characters, but with at least one
1893 whose value was between 128 and 255 caused PCRE to crash.
1894
1895
1896 Version 4.1 12-Mar-03
1897 ---------------------
1898
1899 1. Compiling with gcc -pedantic found a couple of places where casts were
1900 needed, and a string in dftables.c that was longer than standard compilers are
1901 required to support.
1902
1903 2. Compiling with Sun's compiler found a few more places where the code could
1904 be tidied up in order to avoid warnings.
1905
1906 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1907 first of these names is deprecated in the latest Autoconf in favour of the name
1908 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1909 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1910 analogy I have changed it to CFLAGS_FOR_BUILD.
1911
1912 4. Added -no-undefined to the linking command in the Makefile, because this is
1913 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1914 linking step for the pcreposix library.
1915
1916 5. PCRE was failing to diagnose the case of two named groups with the same
1917 name.
1918
1919 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1920 literal character that is needed in the subject for a match, and scans along to
1921 ensure that it is present before embarking on the full matching process. This
1922 saves time in cases of nested unlimited repeats that are never going to match.
1923 Problem: the scan can take a lot of time if the subject is very long (e.g.
1924 megabytes), thus penalizing straightforward matches. It is now done only if the
1925 amount of subject to be scanned is less than 1000 bytes.
1926
1927 7. A lesser problem with the same optimization is that it was recording the
1928 first character of an anchored pattern as "needed", thus provoking a search
1929 right along the subject, even when the first match of the pattern was going to
1930 fail. The "needed" character is now not set for anchored patterns, unless it
1931 follows something in the pattern that is of non-fixed length. Thus, it still
1932 fulfils its original purpose of finding quick non-matches in cases of nested
1933 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1934
1935
1936 Version 4.0 17-Feb-03
1937 ---------------------
1938
1939 1. If a comment in an extended regex that started immediately after a meta-item
1940 extended to the end of string, PCRE compiled incorrect data. This could lead to
1941 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1942
1943 2. Moved to autoconf 2.53 and libtool 1.4.2.
1944
1945 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1946 the special perltest8 script is no longer needed - all the tests can be run
1947 from a single perltest script.
1948
1949 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1950 by \s. It has now been removed in PCRE. This means it isn't recognized as
1951 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1952 class [:space:] *does* include VT, thereby creating a mess.
1953
1954 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1955 space and tab.
1956
1957 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1958 its new features into the main test script, reducing the number of scripts.
1959
1960 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1961 were backward compatible, and made the (?i) apply to the whole pattern, as if
1962 /i were given. Now it behaves more logically, and applies the option setting
1963 only to what follows. PCRE has been changed to follow suit. However, if it
1964 finds options settings right at the start of the pattern, it extracts them into
1965 the global options, as before. Thus, they show up in the info data.
1966
1967 8. Added support for the \Q...\E escape sequence. Characters in between are
1968 treated as literals. This is slightly different from Perl in that $ and @ are
1969 also handled as literals inside the quotes. In Perl, they will cause variable
1970 interpolation. Note the following examples:
1971
1972 Pattern PCRE matches Perl matches
1973
1974 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1975 \Qabc\$xyz\E abc\$xyz abc\$xyz
1976 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1977
1978 For compatibility with Perl, \Q...\E sequences are recognized inside character
1979 classes as well as outside them.
1980
1981 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1982 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1983 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1984 signed/unsigned warnings.
1985
1986 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1987 option for pcretest, so I've replaced it by a simple function that does just
1988 that job.
1989
1990 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1991 "pcregrep -".
1992
1993 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1994 Java package. This provides some syntactic sugar for simple cases of what my
1995 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1996 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1997 item, you can use this simplified notation. Note that only makes sense with
1998 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1999 greediness, whatever the setting of the PCRE_UNGREEDY option.
2000
2001 13. A change of greediness default within a pattern was not taking effect at
2002 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2003 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2004 was abstracted outside.
2005
2006 14. PCRE now supports the \G assertion. It is true when the current matching
2007 position is at the start point of the match. This differs from \A when the
2008 starting offset is non-zero. Used with the /g option of pcretest (or similar
2009 code), it works in the same way as it does for Perl's /g option. If all
2010 alternatives of a regex begin with \G, the expression is anchored to the start
2011 match position, and the "anchored" flag is set in the compiled expression.
2012
2013 15. Some bugs concerning the handling of certain option changes within patterns
2014 have been fixed. These applied to options other than (?ims). For example,
2015 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2016 been the other way round. Some of this was related to change 7 above.
2017
2018 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2019 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2020 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2021 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2022
2023 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2024 mode. Unlike ".", it always matches newline, whatever the setting of
2025 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2026 assertions. Perl allows it, but it doesn't (in general) work because it can't
2027 calculate the length of the lookbehind. At least, that's the case for Perl
2028 5.8.0 - I've been told they are going to document that it doesn't work in
2029 future.
2030
2031 18. Added an error diagnosis for escapes that PCRE does not support: these are
2032 \L, \l, \N, \P, \p, \U, \u, and \X.
2033
2034 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2035 reading past the end of the pattern in cases such as /[abcd/.
2036
2037 20. PCRE was getting more memory than necessary for patterns with classes that
2038 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2039
2040 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2041 compiling PCRE for use with Virtual Pascal.
2042
2043 22. Small fix to the Makefile to make it work properly if the build is done
2044 outside the source tree.
2045
2046 23. Added a new extension: a condition to go with recursion. If a conditional
2047 subpattern starts with (?(R) the "true" branch is used if recursion has
2048 happened, whereas the "false" branch is used only at the top level.
2049
2050 24. When there was a very long string of literal characters (over 255 bytes
2051 without UTF support, over 250 bytes with UTF support), the computation of how
2052 much memory was required could be incorrect, leading to segfaults or other
2053 strange effects.
2054
2055 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2056 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2057 there was a subsequent back reference to those brackets. This meant that, for
2058 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2059 possible to check for precisely this case. All we can do is abandon the
2060 optimization if .* occurs inside capturing brackets when there are any back
2061 references whatsoever. (See below for a better fix that came later.)
2062
2063 26. The handling of the optimization for finding the first character of a
2064 non-anchored pattern, and for finding a character that is required later in the
2065 match were failing in some cases. This didn't break the matching; it just
2066 failed to optimize when it could. The way this is done has been re-implemented.
2067
2068 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2069
2070 28. Added a new feature that provides some of the functionality that Perl
2071 provides with (?{...}). The facility is termed a "callout". The way it is done
2072 in PCRE is for the caller to provide an optional function, by setting
2073 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2074 global variable. By default it is unset, which disables all calling out. To get
2075 the function called, the regex must include (?C) at appropriate points. This
2076 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2077 This provides a means of identifying different callout points. When PCRE
2078 reaches such a point in the regex, if pcre_callout has been set, the external
2079 function is called. It is provided with data in a structure called
2080 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2081 matching continues; if it returns a non-zero value, the match at the current
2082 point fails. However, backtracking will occur if possible. [This was changed
2083 later and other features added - see item 49 below.]
2084
2085 29. pcretest is upgraded to test the callout functionality. It provides a
2086 callout function that displays information. By default, it shows the start of
2087 the match and the current position in the text. There are some new data escapes
2088 to vary what happens:
2089
2090 \C+ in addition, show current contents of captured substrings
2091 \C- do not supply a callout function
2092 \C!n return 1 when callout number n is reached
2093 \C!n!m return 1 when callout number n is reached for the mth time
2094
2095 30. If pcregrep was called with the -l option and just a single file name, it
2096 output "<stdin>" if a match was found, instead of the file name.
2097
2098 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2099 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2100 pcre_exec(). This saves a malloc/free per call. The default value of
2101 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2102 when configuring.
2103
2104 32. The default maximum size of a compiled pattern is 64K. There have been a
2105 few cases of people hitting this limit. The code now uses macros to handle the
2106 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2107 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2108 configuring. Tests 2 and 5 work only with 2-byte links because they output
2109 debugging information about compiled patterns.
2110
2111 33. Internal code re-arrangements:
2112
2113 (a) Moved the debugging function for printing out a compiled regex into
2114 its own source file (printint.c) and used #include to pull it into
2115 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2116 separate copies.
2117
2118 (b) Defined the list of op-code names for debugging as a macro in
2119 internal.h so that it is next to the definition of the opcodes.
2120
2121 (c) Defined a table of op-code lengths for simpler skipping along compiled
2122 code. This is again a macro in internal.h so that it is next to the
2123 definition of the opcodes.
2124
2125 34. Added support for recursive calls to individual subpatterns, along the
2126 lines of Robin Houston's patch (but implemented somewhat differently).
2127
2128 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2129 allow it to read and process whole directories in Win32. This code was
2130 contributed by Lionel Fourquaux; it has not been tested by me.
2131
2132 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2133 used to name a group. Names consist of alphanumerics and underscores, and must
2134 be unique. Back references use the syntax (?P=name) and recursive calls use
2135 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2136 numbers. The function pcre_fullinfo() can be used after compilation to extract
2137 a name/number map. There are three relevant calls:
2138
2139 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2140 PCRE_INFO_NAMECOUNT yields the number of entries
2141 PCRE_INFO_NAMETABLE yields a pointer to the map.
2142
2143 The map is a vector of fixed-size entries. The size of each entry depends on
2144 the length of the longest name used. The first two bytes of each entry are the
2145 group number, most significant byte first. There follows the corresponding
2146 name, zero terminated. The names are in alphabetical order.
2147
2148 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2149 case instead of 255. Making it the same both with and without UTF-8 support
2150 means that the same test output works with both.
2151
2152 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2153 calling malloc() with a zero argument.
2154
2155 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2156 optimization. I've improved things by keeping a bitmap of backreferences with
2157 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2158 fact referenced, the optimization can be applied. It is unlikely that a
2159 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2160 the match to follow \n) will appear inside brackets with a number greater than
2161 31, but if it does, any back reference > 31 suppresses the optimization.
2162
2163 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2164 of disabling numbered capturing parentheses. Any opening parenthesis that is
2165 not followed by ? behaves as if it were followed by ?: but named parentheses
2166 can still be used for capturing (and they will acquire numbers in the usual
2167 way).
2168
2169 41. Redesigned the return codes from the match() function into yes/no/error so
2170 that errors can be passed back from deep inside the nested calls. A malloc
2171 failure while inside a recursive subpattern call now causes the
2172 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2173
2174 42. It is now possible to set a limit on the number of times the match()
2175 function is called in a call to pcre_exec(). This facility makes it possible to
2176 limit the amount of recursion and backtracking, though not in a directly
2177 obvious way, because the match() function is used in a number of different
2178 circumstances. The count starts from zero for each position in the subject
2179 string (for non-anchored patterns). The default limit is, for compatibility, a
2180 large number, namely 10 000 000. You can change this in two ways:
2181
2182 (a) When configuring PCRE before making, you can use --with-match-limit=n
2183 to set a default value for the compiled library.
2184
2185 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2186 a different value is set. See 45 below.
2187
2188 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2189
2190 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2191 of things that can be changed at compile time. The first argument specifies
2192 what is wanted and the second points to where the information is to be placed.
2193 The current list of available information is:
2194
2195 PCRE_CONFIG_UTF8
2196
2197 The output is an integer that is set to one if UTF-8 support is available;
2198 otherwise it is set to zero.
2199
2200 PCRE_CONFIG_NEWLINE
2201
2202 The output is an integer that it set to the value of the code that is used for
2203 newline. It is either LF (10) or CR (13).
2204
2205 PCRE_CONFIG_LINK_SIZE
2206
2207 The output is an integer that contains the number of bytes used for internal
2208 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2209
2210 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2211
2212 The output is an integer that contains the threshold above which the POSIX
2213 interface uses malloc() for output vectors. See item 31 above.
2214
2215 PCRE_CONFIG_MATCH_LIMIT
2216
2217 The output is an unsigned integer that contains the default limit of the number
2218 of match() calls in a pcre_exec() execution. See 42 above.
2219
2220 44. pcretest has been upgraded by the addition of the -C option. This causes it
2221 to extract all the available output from the new pcre_config() function, and to
2222 output it. The program then exits immediately.
2223
2224 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2225 order to support additional features. One way would have been to define
2226 pcre_exec2() (for example) with extra arguments, but this would not have been
2227 extensible, and would also have required all calls to the original function to
2228 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2229 is used for passing in "extra" data from pcre_study().
2230
2231 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2232 contains the following fields:
2233
2234 flags a bitmap indicating which of the following fields are set
2235 study_data opaque data from pcre_study()
2236 match_limit a way of specifying a limit on match() calls for a specific
2237 call to pcre_exec()
2238 callout_data data for callouts (see 49 below)
2239
2240 The flag bits are also defined in pcre.h, and are
2241
2242 PCRE_EXTRA_STUDY_DATA
2243 PCRE_EXTRA_MATCH_LIMIT
2244 PCRE_EXTRA_CALLOUT_DATA
2245
2246 The pcre_study() function now returns one of these new pcre_extra blocks, with
2247 the actual study data pointed to by the study_data field, and the
2248 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2249 before. That is, this change is entirely upwards-compatible and requires no
2250 change to existing code.
2251
2252 If you want to pass in additional data to pcre_exec(), you can either place it
2253 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2254 block.
2255
2256 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2257 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2258 times with different match limits, until it finds the minimum value needed for
2259 pcre_exec() to complete. The value is then output. This can be instructive; for
2260 most simple matches the number is quite small, but for pathological cases it
2261 gets very large very quickly.
2262
2263 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2264 returns the size of the data block pointed to by the study_data field in a
2265 pcre_extra block, that is, the value that was passed as the argument to
2266 pcre_malloc() when PCRE was getting memory in which to place the information
2267 created by pcre_study(). The fourth argument should point to a size_t variable.
2268 pcretest has been extended so that this information is shown after a successful
2269 pcre_study() call when information about the compiled regex is being displayed.
2270
2271 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2272 because what follows is always an absolute path. (Later: it turns out that this
2273 is more than cosmetic for MinGW, because it doesn't like empty path
2274 components.)
2275
2276 49. Some changes have been made to the callout feature (see 28 above):
2277
2278 (i) A callout function now has three choices for what it returns:
2279
2280 0 => success, carry on matching
2281 > 0 => failure at this point, but backtrack if possible
2282 < 0 => serious error, return this value from pcre_exec()
2283
2284 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2285 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2286 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2287 use by callout functions. It will never be used by PCRE itself.
2288
2289 (ii) The pcre_extra structure (see 45 above) has a void * field called
2290 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2291 pcre_callout_block structure has a field of the same name. The contents of
2292 the field passed in the pcre_extra structure are passed to the callout
2293 function in the corresponding field in the callout block. This makes it
2294 easier to use the same callout-containing regex from multiple threads. For
2295 testing, the pcretest program has a new data escape
2296
2297 \C*n pass the number n (may be negative) as callout_data
2298
2299 If the callout function in pcretest receives a non-zero value as
2300 callout_data, it returns that value.
2301
2302 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2303 there were some redundant $(CFLAGS) in commands that are now specified as
2304 $(LINK), which already includes $(CFLAGS).
2305
2306 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2307 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2308 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2309 one-byte characters throughout. Note that case-insensitive matching applies
2310 only to characters whose values are less than 256. PCRE doesn't support the
2311 notion of cases for higher-valued characters.
2312
2313 (i) A character class whose characters are all within 0-255 is handled as
2314 a bit map, and the map is inverted for negative classes. Previously, a
2315 character > 255 always failed to match such a class; however it should
2316 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2317
2318 (ii) A negated character class with a single character < 255 is coded as
2319 "not this character" (OP_NOT). This wasn't working properly when the test
2320 character was multibyte, either singly or repeated.
2321
2322 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2323 mode, for example: \x{100}{2,3}.
2324
2325 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2326 singly or repeated) now correctly test multibyte characters. However,
2327 PCRE doesn't recognize any characters with values greater than 255 as
2328 digits, spaces, or word characters. Such characters always match \D, \S,
2329 and \W, and never match \d, \s, or \w.
2330
2331 (v) Classes may now contain characters and character ranges with values
2332 greater than 255. For example: [ab\x{100}-\x{400}].
2333
2334 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2335 PCRE in UTF-8 mode.
2336
2337 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2338 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2339 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2340 value.)
2341
2342 53. The single man page has become too large. I have therefore split it up into
2343 a number of separate man pages. These also give rise to individual HTML pages;
2344 these are now put in a separate directory, and there is an index.html page that
2345 lists them all. Some hyperlinking between the pages has been installed.
2346
2347 54. Added convenience functions for handling named capturing parentheses.
2348
2349 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2350 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2351 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2352 are faulted.
2353
2354 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2355 calling configure. These values are used when compiling the dftables.c program
2356 which is run to generate the source of the default character tables. They
2357 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2358 you will need to set these values.
2359
2360 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2361
2362
2363 Version 3.9 02-Jan-02
2364 ---------------------
2365
2366 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2367
2368 2. If --disable-static was given, the building process failed when trying to
2369 build pcretest and pcregrep. (For some reason it was using libtool to compile
2370 them, which is not right, as they aren't part of the library.)
2371
2372
2373 Version 3.8 18-Dec-01
2374 ---------------------
2375
2376 1. The experimental UTF-8 code was completely screwed up. It was packing the
2377 bytes in the wrong order. How dumb can you get?
2378
2379
2380 Version 3.7 29-Oct-01
2381 ---------------------
2382
2383 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2384 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2385 this didn't happen under Solaris 8, where I normally test things.
2386
2387 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2388 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2389 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2390 this fix an hour or so after the initial 3.7 release.)
2391
2392
2393 Version 3.6 23-Oct-01
2394 ---------------------
2395
2396 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2397 offsets passed as NULL with zero offset count.
2398
2399 2. The config.guess and config.sub files had not been updated when I moved to
2400 the latest autoconf.
2401
2402
2403 Version 3.5 15-Aug-01
2404 ---------------------
2405
2406 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2407 had been forgotten.
2408
2409 2. By using declared but undefined structures, we can avoid using "void"
2410 definitions in pcre.h while keeping the internal definitions of the structures
2411 private.
2412
2413 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2414 user point of view, this means that both static and shared libraries are built
2415 by default, but this can be individually controlled. More of the work of
2416 handling this static/shared cases is now inside libtool instead of PCRE's make
2417 file.
2418
2419 4. The pcretest utility is now installed along with pcregrep because it is
2420 useful for users (to test regexs) and by doing this, it automatically gets
2421 relinked by libtool. The documentation has been turned into a man page, so
2422 there are now .1, .txt, and .html versions in /doc.
2423
2424 5. Upgrades to pcregrep:
2425 (i) Added long-form option names like gnu grep.
2426 (ii) Added --help to list all options with an explanatory phrase.
2427 (iii) Added -r, --recursive to recurse into sub-directories.
2428 (iv) Added -f, --file to read patterns from a file.
2429
2430 6. pcre_exec() was referring to its "code" argument before testing that
2431 argument for NULL (and giving an error if it was NULL).
2432
2433 7. Upgraded Makefile.in to allow for compiling in a different directory from
2434 the source directory.
2435
2436 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2437 options bits, the pointer it was passed was to an int instead of to an unsigned
2438 long int. This mattered only on 64-bit systems.
2439
2440 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2441 generated) instead of pcre.in, which it its source. Also made the same change
2442 in several of the .c files.
2443
2444 10. A new release of gcc defines printf() as a macro, which broke pcretest
2445 because it had an ifdef in the middle of a string argument for printf(). Fixed
2446 by using separate calls to printf().
2447
2448 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2449 script, to force use of CR or LF instead of \n in the source. On non-Unix
2450 systems, the value can be set in config.h.
2451
2452 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2453 absolute limit. Changed the text of the error message to make this clear, and
2454 likewise updated the man page.
2455
2456 13. The limit of 99 on the number of capturing subpatterns has been removed.
2457 The new limit is 65535, which I hope will not be a "real" limit.
2458
2459
2460 Version 3.4 22-Aug-00
2461 ---------------------
2462
2463 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2464
2465 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2466
2467
2468 Version 3.3 01-Aug-00
2469 ---------------------
2470
2471 1. If an octal character was given, but the value was greater than \377, it
2472 was not getting masked to the least significant bits, as documented. This could
2473 lead to crashes in some systems.
2474
2475 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2476 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2477
2478 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2479 These just pass their arguments on to (pcre_free)(), but they are provided
2480 because some uses of PCRE bind it to non-C systems that can call its functions,
2481 but cannot call free() or pcre_free() directly.
2482
2483 4. Add "make test" as a synonym for "make check". Corrected some comments in
2484 the Makefile.
2485
2486 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2487 Makefile.
2488
2489 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2490 command called pgrep for grepping around the active processes.
2491
2492 7. Added the beginnings of support for UTF-8 character strings.
2493
2494 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2495 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2496 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2497 out for the ar command.)
2498
2499
2500 Version 3.2 12-May-00
2501 ---------------------
2502
2503 This is purely a bug fixing release.
2504
2505 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2506 of ZA. This was just one example of several cases that could provoke this bug,
2507 which was introduced by change 9 of version 2.00. The code for breaking
2508 infinite loops after an iteration that matches an empty string was't working
2509 correctly.
2510
2511 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2512 when matched against abbab (for example). After matching an empty string, it
2513 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2514 caused it to match further down the string than it should.
2515
2516 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2517 was there because it doesn't seem to be needed, and it causes trouble on some
2518 systems, as it is not a Standard C header. It has been removed.
2519
2520 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2521 were reported on the Macintosh. The changes were from
2522
2523 while ((c = *(++ptr)) != 0 && c != '\n');
2524 to
2525 while ((c = *(++ptr)) != 0 && c != '\n') ;
2526
2527 Totally extraordinary, but if that's what it takes...
2528
2529 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2530 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2531 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2532 assumes the way PCRE uses memmove() (always moving upwards).
2533
2534 6. PCRE is being used in one environment where strchr() is not available. There
2535 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2536 faster code anyway.
2537
2538
2539 Version 3.1 09-Feb-00
2540 ---------------------
2541
2542 The only change in this release is the fixing of some bugs in Makefile.in for
2543 the "install" target:
2544
2545 (1) It was failing to install pcreposix.h.
2546
2547 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2548
2549
2550 Version 3.0 01-Feb-00
2551 ---------------------
2552
2553 1. Add support for the /+ modifier to perltest (to output $` like it does in
2554 pcretest).
2555
2556 2. Add support for the /g modifier to perltest.
2557
2558 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2559 matches null strings.
2560
2561 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2562 pattern. Perl treats empty patterns specially - it reuses the most recent
2563 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2564 effect.
2565
2566 5. The POSIX interface was broken in that it was just handing over the POSIX
2567 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2568 required a bigger vector, with some working space on the end. This means that
2569 the POSIX wrapper now has to get and free some memory, and copy the results.
2570
2571 6. Added some simple autoconf support, placing the test data and the
2572 documentation in separate directories, re-organizing some of the
2573 information files, and making it build pcre-config (a GNU standard). Also added
2574 libtool support for building PCRE as a shared library, which is now the
2575 default.
2576
2577 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2578 09 are not valid octal constants. Single digits will be used for minor values
2579 less than 10.
2580
2581 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2582 existing programs that set these in the POSIX interface can use PCRE without
2583 modification.
2584
2585 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2586 return all that pcre_info() returns, plus additional data. The pcre_info()
2587 function is retained for compatibility, but is considered to be obsolete.
2588
2589 10. Added experimental recursion feature (?R) to handle one common case that
2590 Perl 5.6 will be able to do with (?p{...}).
2591
2592 11. Added support for POSIX character classes like [:alpha:], which Perl is
2593 adopting.
2594
2595
2596 Version 2.08 31-Aug-99
2597 ----------------------
2598
2599 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2600 trying to match at the startoffset position, but instead was moving forward to
2601 the next newline as if a previous match had failed.
2602
2603 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2604 and could get into a loop if a null string was matched other than at the start
2605 of the subject.
2606
2607 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2608 be distinguished at compile time, and for completeness also added PCRE_DATE.
2609
2610 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2611 in GnuWin32 environments.
2612
2613
2614 Version 2.07 29-Jul-99
2615 ----------------------
2616
2617 1. The documentation is now supplied in plain text form and HTML as well as in
2618 the form of man page sources.
2619
2620 2. C++ compilers don't like assigning (void *) values to other pointer types.
2621 In particular this affects malloc(). Although there is no problem in Standard
2622 C, I've put in casts to keep C++ compilers happy.
2623
2624 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2625 should be (const char *).
2626
2627 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2628 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2629 However, I haven't made this a standard facility. The documentation doesn't
2630 mention it, and the Makefile doesn't support it.
2631
2632 5. The Makefile now contains an "install" target, with editable destinations at
2633 the top of the file. The pcretest program is not installed.
2634
2635 6. pgrep -V now gives the PCRE version number and date.
2636
2637 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2638 causing the entire string to be ignored, instead of just the last character.
2639
2640 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2641 non-matching string, it can take a very, very long time, even for strings of
2642 quite modest length, because of the nested recursion. PCRE now does better in
2643 some of these cases. It does this by remembering the last required literal
2644 character in the pattern, and pre-searching the subject to ensure it is present
2645 before running the real match. In other words, it applies a heuristic to detect
2646 some types of certain failure quickly, and in the above example, if presented
2647 with a string that has no trailing " it gives "no match" very quickly.
2648
2649 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2650 other alternatives are tried instead.
2651
2652
2653 Version 2.06 09-Jun-99
2654 ----------------------
2655
2656 1. Change pcretest's output for amount of store used to show just the code
2657 space, because the remainder (the data block) varies in size between 32-bit and
2658 64-bit systems.
2659
2660 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2661 start matching at. This allows lookbehinds to work when searching for multiple
2662 occurrences in a string.
2663
2664 3. Added additional options to pcretest for testing multiple occurrences:
2665
2666 /+ outputs the rest of the string that follows a match
2667 /g loops for multiple occurrences, using the new startoffset argument
2668 /G loops for multiple occurrences by passing an incremented pointer
2669
2670 4. PCRE wasn't doing the "first character" optimization for patterns starting
2671 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2672 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2673 the letter 'x'. On long subject strings, this gives a significant speed-up.
2674
2675
2676 Version 2.05 21-Apr-99
2677 ----------------------
2678
2679 1. Changed the type of magic_number from int to long int so that it works
2680 properly on 16-bit systems.
2681
2682 2. Fixed a bug which caused patterns starting with .* not to work correctly
2683 when the subject string contained newline characters. PCRE was assuming
2684 anchoring for such patterns in all cases, which is not correct because .* will
2685 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2686 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2687 must be retried after every newline in the subject.
2688
2689
2690 Version 2.04 18-Feb-99
2691 ----------------------
2692
2693 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2694 computation of the store needed to hold the pattern was incorrect (too large).
2695 If such patterns were nested a few deep, this could multiply and become a real
2696 problem.
2697
2698 2. Added /M option to pcretest to show the memory requirement of a specific
2699 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2700
2701 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2702 compiled in such a way that the backtracking after subsequent failure was
2703 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2704 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2705
2706
2707 Version 2.03 02-Feb-99
2708 ----------------------
2709
2710 1. Fixed typo and small mistake in man page.
2711
2712 2. Added 4th condition (GPL supersedes if conflict) and created separate
2713 LICENCE file containing the conditions.
2714
2715 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2716 Perl, that is the internal \ allows the delimiter to be included in the
2717 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2718 the final delimiter, add \ to the end of the pattern (to test the error).
2719
2720 4. Added the convenience functions for extracting substrings after a successful
2721 match. Updated pcretest to make it able to test these functions.
2722
2723
2724 Version 2.02 14-Jan-99
2725 ----------------------
2726
2727 1. Initialized the working variables associated with each extraction so that
2728 their saving and restoring doesn't refer to uninitialized store.
2729
2730 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2731 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2732 fix the problem.
2733
2734 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2735 calls, and wasn't printing the correct value for compiling calls. Increased the
2736 default value of LOOPREPEAT, and the number of significant figures in the
2737 times.
2738
2739 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2740
2741 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2742 a building problem on Windows NT with a FAT file system.
2743
2744
2745 Version 2.01 21-Oct-98
2746 ----------------------
2747
2748 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2749 to character tables built by pcre_maketables() in the current locale. If NULL
2750 is passed, the default tables are used.
2751
2752
2753 Version 2.00 24-Sep-98
2754 ----------------------
2755
2756 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2757 it any more.
2758
2759 2. Allow quantification of (?>) groups, and make it work correctly.
2760
2761 3. The first character computation wasn't working for (?>) groups.
2762
2763 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2764 end of the subject) and add 5.005's \z, which really does match only at the
2765 very end of the subject.
2766
2767 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2768
2769 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2770 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2771 localized options. All options to pcre_study() were also removed.
2772
2773 7. Add other new features from 5.005:
2774
2775 $(?<= positive lookbehind
2776 $(?<! negative lookbehind
2777 (?imsx-imsx) added the unsetting capability
2778 such a setting is global if at outer level; local otherwise
2779 (?imsx-imsx:) non-capturing groups with option setting
2780 (?(cond)re|re) conditional pattern matching
2781
2782 A backreference to itself in a repeated group matches the previous
2783 captured string.
2784
2785 8. General tidying up of studying (both automatic and via "study")
2786 consequential on the addition of new assertions.
2787
2788 9. As in 5.005, unlimited repeated groups that could match an empty substring
2789 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2790 runtime if any iteration does actually match an empty substring.
2791
2792 10. Include the RunTest script in the distribution.
2793
2794 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2795 discrepancies, some of which were old and were also with respect to 5.004. They
2796 have now been fixed.
2797
2798
2799 Version 1.09 28-Apr-98
2800 ----------------------
2801
2802 1. A negated single character class followed by a quantifier with a minimum
2803 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2804 program crashes, or just wrong answers. This did not apply to negated classes
2805 containing more than one character, or to minima other than one.
2806
2807
2808 Version 1.08 27-Mar-98
2809 ----------------------
2810
2811 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2812
2813 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2814 latter must appear before anything that relies on it in the pattern.
2815
2816
2817 Version 1.07 16-Feb-98
2818 ----------------------
2819
2820 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2821 repeat of a potentially empty string).
2822
2823
2824 Version 1.06 23-Jan-98
2825 ----------------------
2826
2827 1. Added Markus Oberhumer's little patches for C++.
2828
2829 2. Literal strings longer than 255 characters were broken.
2830
2831
2832 Version 1.05 23-Dec-97
2833 ----------------------
2834
2835 1. Negated character classes containing more than one character were failing if
2836 PCRE_CASELESS was set at run time.
2837
2838
2839 Version 1.04 19-Dec-97
2840 ----------------------
2841
2842 1. Corrected the man page, where some "const" qualifiers had been omitted.
2843
2844 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2845 input syntax.
2846
2847 3. Fixed memory leak which occurred when a regex with back references was
2848 matched with an offsets vector that wasn't big enough. The temporary memory
2849 that is used in this case wasn't being freed if the match failed.
2850
2851 4. Tidied pcretest to ensure it frees memory that it gets.
2852
2853 5. Temporary memory was being obtained in the case where the passed offsets
2854 vector was exactly big enough.
2855
2856 6. Corrected definition of offsetof() from change 5 below.
2857
2858 7. I had screwed up change 6 below and broken the rules for the use of
2859 setjmp(). Now fixed.
2860
2861
2862 Version 1.03 18-Dec-97
2863 ----------------------
2864
2865 1. A erroneous regex with a missing opening parenthesis was correctly
2866 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2867 on some systems.
2868
2869 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2870 it was reported that one broken compiler failed on the former because "code" is
2871 also an independent variable.
2872
2873 3. The erroneous regex a[]b caused an array overrun reference.
2874
2875 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2876 fail on data ending with that character. (It was going on too far, and checking
2877 the next character, typically a binary zero.) This was specific to the
2878 optimized code for single-character negative classes.
2879
2880 5. Added a contributed patch from the TIN world which does the following:
2881
2882 + Add an undef for memmove, in case the the system defines a macro for it.
2883
2884 + Add a definition of offsetof(), in case there isn't one. (I don't know
2885 the reason behind this - offsetof() is part of the ANSI standard - but
2886 it does no harm).
2887
2888 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2889 most of the places where whitespace preceded '#'. I have given up and
2890 allowed the remaining 2 cases to be at the margin.
2891
2892 + Rename some variables in pcre to eliminate shadowing. This seems very
2893 pedantic, but does no harm, of course.
2894
2895 6. Moved the call to setjmp() into its own function, to get rid of warnings
2896 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2897
2898 7. Constructs such as \d{8,} were compiling into the equivalent of
2899 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2900 outcome, but in this particular case used more store than had been allocated,
2901 which caused the bug to be discovered because it threw up an internal error.
2902
2903 8. The debugging code in both pcre and pcretest for outputting the compiled
2904 form of a regex was going wrong in the case of back references followed by
2905 curly-bracketed repeats.
2906
2907
2908 Version 1.02 12-Dec-97
2909 ----------------------
2910
2911 1. Typos in pcre.3 and comments in the source fixed.
2912
2913 2. Applied a contributed patch to get rid of places where it used to remove
2914 'const' from variables, and fixed some signed/unsigned and uninitialized
2915 variable warnings.
2916
2917 3. Added the "runtest" target to Makefile.
2918
2919 4. Set default compiler flag to -O2 rather than just -O.
2920
2921
2922 Version 1.01 19-Nov-97
2923 ----------------------
2924
2925 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2926 like /([ab]*)*/, that is, for classes with more than one character in them.
2927
2928 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2929 as /((?>a*))*/ (a PCRE_EXTRA facility).
2930
2931
2932 Version 1.00 18-Nov-97
2933 ----------------------
2934
2935 1. Added compile-time macros to support systems such as SunOS4 which don't have
2936 memmove() or strerror() but have other things that can be used instead.
2937
2938 2. Arranged that "make clean" removes the executables.
2939
2940
2941 Version 0.99 27-Oct-97
2942 ----------------------
2943
2944 1. Fixed bug in code for optimizing classes with only one character. It was
2945 initializing a 32-byte map regardless, which could cause it to run off the end
2946 of the memory it had got.
2947
2948 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2949
2950
2951 Version 0.98 22-Oct-97
2952 ----------------------
2953
2954 1. Fixed bug in code for handling temporary memory usage when there are more
2955 back references than supplied space in the ovector. This could cause segfaults.
2956
2957
2958 Version 0.97 21-Oct-97
2959 ----------------------
2960
2961 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2962
2963 2. Optimized negated single characters not to use a bit map.
2964
2965 3. Brought error texts together as macro definitions; clarified some of them;
2966 fixed one that was wrong - it said "range out of order" when it meant "invalid
2967 escape sequence".
2968
2969 4. Changed some char * arguments to const char *.
2970
2971 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2972
2973 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2974 pcretest.
2975
2976
2977 Version 0.96 16-Oct-97
2978 ----------------------
2979
2980 1. Added a simple "pgrep" utility to the distribution.
2981
2982 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2983 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2984 where "ddd" means "one or more decimal digits".
2985
2986 3. Fixed serious bug. If a pattern had a back reference, but the call to
2987 pcre_exec() didn't supply a large enough ovector to record the related
2988 identifying subpattern, the match always failed. PCRE now remembers the number
2989 of the largest back reference, and gets some temporary memory in which to save
2990 the offsets during matching if necessary, in order to ensure that
2991 backreferences always work.
2992
2993 4. Increased the compatibility with Perl in a number of ways:
2994
2995 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2996 to request this handling. The option can be set at compile or exec time.
2997
2998 (b) $ matches before a terminating newline by default; an option
2999 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3000 mode). The option can be set at compile or exec time.
3001
3002 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3003 the same as Perl's. If the decimal number it represents is less than 10
3004 or there aren't that many previous left capturing parentheses, an octal
3005 escape is read. Inside a character class, it's always an octal escape,
3006 even if it is a single digit.
3007
3008 (d) An escaped but undefined alphabetic character is taken as a literal,
3009 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3010 escapes.
3011
3012 (e) {0} is now permitted. (The previous item is removed from the compiled
3013 pattern).
3014
3015 5. Changed all the names of code files so that the basic parts are no longer
3016 than 10 characters, and abolished the teeny "globals.c" file.
3017
3018 6. Changed the handling of character classes; they are now done with a 32-byte
3019 bit map always.
3020
3021 7. Added the -d and /D options to pcretest to make it possible to look at the
3022 internals of compilation without having to recompile pcre.
3023
3024
3025 Version 0.95 23-Sep-97
3026 ----------------------
3027
3028 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3029 \x20 at the start of a run of normal characters. These were being treated as
3030 real characters, instead of the source characters being re-checked.
3031
3032
3033 Version 0.94 18-Sep-97
3034 ----------------------
3035
3036 1. The functions are now thread-safe, with the caveat that the global variables
3037 containing pointers to malloc() and free() or alternative functions are the
3038 same for all threads.
3039
3040 2. Get pcre_study() to generate a bitmap of initial characters for non-
3041 anchored patterns when this is possible, and use it if passed to pcre_exec().
3042
3043
3044 Version 0.93 15-Sep-97
3045 ----------------------
3046
3047 1. /(b)|(:+)/ was computing an incorrect first character.
3048
3049 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3050 but not actually doing anything yet.
3051
3052 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3053 as Perl does (e.g. [-az] or [az-]).
3054
3055 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3056 all possible positions.
3057
3058 5. Split up into different modules to avoid including unneeded functions in a
3059 compiled binary. However, compile and exec are still in one module. The "study"
3060 function is split off.
3061
3062 6. The character tables are now in a separate module whose source is generated
3063 by an auxiliary program - but can then be edited by hand if required. There are
3064 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3065 toupper() in the code.
3066
3067 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3068 make them global. Abolish the function for setting them, as the caller can now
3069 set them directly.
3070
3071
3072 Version 0.92 11-Sep-97
3073 ----------------------
3074
3075 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3076 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3077
3078 2. Caseless matching was not working in character classes if the characters in
3079 the pattern were in upper case.
3080
3081 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3082
3083 4. Make PCRE_ANCHORED public and accept as a compile option.
3084
3085 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3086 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3087 pass them.
3088
3089 6. Give an error if bad option bits passed at compile or run time.
3090
3091 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3092 pcretest to cause it to pass that flag.
3093
3094 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3095 options, and the first character, if set.
3096
3097 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3098
3099
3100 Version 0.91 10-Sep-97
3101 ----------------------
3102
3103 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3104 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3105
3106 2. PCRE was looping on encountering an indefinitely repeated back reference to
3107 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3108 Perl does - treats the match as successful.
3109
3110 ****

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