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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 131 - (show annotations) (download)
Mon Mar 26 15:10:12 2007 UTC (7 years, 7 months ago) by ph10
File size: 125770 byte(s)
Various new tidies.

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

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