/[pcre]/code/tags/pcre-7.1/ChangeLog
ViewVC logotype

Contents of /code/tags/pcre-7.1/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 156 - (show annotations) (download)
Tue Apr 24 13:49:23 2007 UTC (6 years, 11 months ago) by ph10
File size: 126831 byte(s)
Tag the PCRE 7.1 release.

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

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