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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 230 - (show annotations) (download)
Mon Sep 10 13:23:56 2007 UTC (7 years, 2 months ago) by ph10
File size: 139676 byte(s)
(1) Move internal flags out of the options field, to make room.
(2) \r and \n must be explicit to trigger the special CRLF handline exception.
(3) (?J) at the start now sets JCHANGED as well as DUPNAMES.

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

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