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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 161 - (show annotations) (download)
Thu May 3 10:28:07 2007 UTC (7 years, 2 months ago) by ph10
File size: 127508 byte(s)
Use the pcretest "Z" option so all tests can be run with all link sizes.

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

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