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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 265 - (show annotations) (download)
Wed Nov 14 11:35:48 2007 UTC (6 years, 10 months ago) by ph10
File size: 142385 byte(s)
Fix negative POSIX class bug with Unicode characters.

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

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