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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 165 - (show annotations) (download)
Wed May 9 10:50:57 2007 UTC (7 years, 5 months ago) by ph10
File size: 127757 byte(s)
Non-longjmp heap recursion.

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

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