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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 268 - (show annotations) (download)
Thu Nov 15 10:28:09 2007 UTC (7 years, 1 month ago) by ph10
File size: 142838 byte(s)
Fix (?&) non-diagnosis bug and missing length check for (?&a) etc.

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

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