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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 216 - (show annotations) (download)
Wed Aug 15 14:35:57 2007 UTC (7 years, 4 months ago) by ph10
File size: 136085 byte(s)
Fixed compile-time loop for patterns like (?:[\PPa*]*){8,} (extended class 
inside group with unlimited repeat).

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

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