ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 193 - (show annotations)
Mon Jul 30 11:34:18 2007 UTC (14 years, 3 months ago) by ph10
File size: 131384 byte(s)
Arrange to use "%I64d" instead of "%lld" for long printing in the pcrecpp 
unittest when running under MinGW.
1 ChangeLog for PCRE
2 ------------------
4 Version 7.3 05-Jul-07
5 ---------------------
7 1. In the rejigging of the build system that eventually resulted in 7.1, the
8 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
9 brackets there is not right, since it causes compilers to look for an
10 installed pcre.h, not the version that is in the source that is being
11 compiled (which of course may be different). I have changed it back to:
13 #include "pcre.h"
15 I have a vague recollection that the change was concerned with compiling in
16 different directories, but in the new build system, that is taken care of
17 by the VPATH setting the Makefile.
19 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
20 when the subject happened to end in the byte 0x85 (e.g. if the last
21 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
22 characters but of course it shouldn't be taken as a newline when it is part
23 of another character. The bug was that, for an unlimited repeat of . in
24 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
25 characters when looking for a newline.
27 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
29 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
30 in debug output.
32 5. Arrange to use "%I64d" instead of "%lld" for long printing in the pcrecpp
33 unittest when running under MinGW.
37 Version 7.2 19-Jun-07
38 ---------------------
40 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
41 which is apparently normally available under Windows.
43 2. Re-jig the pcregrep tests with different newline settings in an attempt
44 to make them independent of the local environment's newline setting.
46 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
48 4. Some of the "internals" tests were previously cut out when the link size
49 was not 2, because the output contained actual offsets. The recent new
50 "Z" feature of pcretest means that these can be cut out, making the tests
51 usable with all link sizes.
53 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
54 stack recursion. This gives a massive performance boost under BSD, but just
55 a small improvement under Linux. However, it saves one field in the frame
56 in all cases.
58 6. Added more features from the forthcoming Perl 5.10:
60 (a) (?-n) (where n is a string of digits) is a relative subroutine or
61 recursion call. It refers to the nth most recently opened parentheses.
63 (b) (?+n) is also a relative subroutine call; it refers to the nth next
64 to be opened parentheses.
66 (c) Conditions that refer to capturing parentheses can be specified
67 relatively, for example, (?(-2)... or (?(+3)...
69 (d) \K resets the start of the current match so that everything before
70 is not part of it.
72 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
74 (f) \g{name} is another synonym - part of Perl 5.10's unification of
75 reference syntax.
77 (g) (?| introduces a group in which the numbering of parentheses in each
78 alternative starts with the same number.
80 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
82 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
85 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
86 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
87 for detecting groups that can match an empty string.
89 9. A pattern with a very large number of alternatives (more than several
90 hundred) was running out of internal workspace during the pre-compile
91 phase, where pcre_compile() figures out how much memory will be needed. A
92 bit of new cunning has reduced the workspace needed for groups with
93 alternatives. The 1000-alternative test pattern now uses 12 bytes of
94 workspace instead of running out of the 4096 that are available.
96 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
98 11. Applied patch from Google to remove an optimization that didn't quite work.
99 The report of the bug said:
101 pcrecpp::RE("a*").FullMatch("aaa") matches, while
102 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
103 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
105 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
106 it matched the wrong number of bytes.
109 Version 7.1 24-Apr-07
110 ---------------------
112 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
113 that is more "standard", making use of automake and other Autotools. There
114 is some re-arrangement of the files and adjustment of comments consequent
115 on this.
117 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
118 for recursive directory scanning broke on some systems because the files
119 are not scanned in any specific order and on different systems the order
120 was different. A call to "sort" has been inserted into RunGrepTest for the
121 approprate test as a short-term fix. In the longer term there may be an
122 alternative.
124 3. I had an email from Eric Raymond about problems translating some of PCRE's
125 man pages to HTML (despite the fact that I distribute HTML pages, some
126 people do their own conversions for various reasons). The problems
127 concerned the use of low-level troff macros .br and .in. I have therefore
128 removed all such uses from the man pages (some were redundant, some could
129 be replaced by .nf/.fi pairs). The 132html script that I use to generate
130 HTML has been updated to handle .nf/.fi and to complain if it encounters
131 .br or .in.
133 4. Updated comments in configure.ac that get placed in config.h.in and also
134 arranged for config.h to be included in the distribution, with the name
135 config.h.generic, for the benefit of those who have to compile without
136 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
138 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
139 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
140 makevp.bat for the current PCRE, using the additional files
141 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
143 6. A Windows user reported a minor discrepancy with test 2, which turned out
144 to be caused by a trailing space on an input line that had got lost in his
145 copy. The trailing space was an accident, so I've just removed it.
147 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
148 that is needed.
150 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
151 as "const" (a) because they are and (b) because it helps the PHP
152 maintainers who have recently made a script to detect big data structures
153 in the php code that should be moved to the .rodata section. I remembered
154 to update Builducptable as well, so it won't revert if ucptable.h is ever
155 re-created.
157 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
158 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
159 order to be able to cut out the UTF-8 tables in the latter when UTF-8
160 support is not required. This saves 1.5-2K of code, which is important in
161 some applications.
163 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
164 so as not to refer to the tables, even though these functions will never be
165 called when UTF-8 support is disabled. Otherwise there are problems with a
166 shared library.
168 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
170 (a) It was defining its arguments as char * instead of void *.
172 (b) It was assuming that all moves were upwards in memory; this was true
173 a long time ago when I wrote it, but is no longer the case.
175 The emulated memove() is provided for those environments that have neither
176 memmove() nor bcopy(). I didn't think anyone used it these days, but that
177 is clearly not the case, as these two bugs were recently reported.
179 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
180 and Detrail to create the HTML documentation, the .txt form of the man
181 pages, and it removes trailing spaces from listed files. It also creates
182 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
183 case, it wraps all the #defines with #ifndefs. This script should be run
184 before "make dist".
186 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
187 with Unicode property support.
189 (a) For a maximizing quantifier, if the two different cases of the
190 character were of different lengths in their UTF-8 codings (there are
191 some cases like this - I found 11), and the matching function had to
192 back up over a mixture of the two cases, it incorrectly assumed they
193 were both the same length.
195 (b) When PCRE was configured to use the heap rather than the stack for
196 recursion during matching, it was not correctly preserving the data for
197 the other case of a UTF-8 character when checking ahead for a match
198 while processing a minimizing repeat. If the check also involved
199 matching a wide character, but failed, corruption could cause an
200 erroneous result when trying to check for a repeat of the original
201 character.
203 13. Some tidying changes to the testing mechanism:
205 (a) The RunTest script now detects the internal link size and whether there
206 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
207 values substituted by "configure". (The RunGrepTest script already did
208 this for UTF-8.) The configure.ac script no longer substitutes the
209 relevant variables.
211 (b) The debugging options /B and /D in pcretest show the compiled bytecode
212 with length and offset values. This means that the output is different
213 for different internal link sizes. Test 2 is skipped for link sizes
214 other than 2 because of this, bypassing the problem. Unfortunately,
215 there was also a test in test 3 (the locale tests) that used /B and
216 failed for link sizes other than 2. Rather than cut the whole test out,
217 I have added a new /Z option to pcretest that replaces the length and
218 offset values with spaces. This is now used to make test 3 independent
219 of link size. (Test 2 will be tidied up later.)
221 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
222 segmentation fault instead of returning the appropriate error message.
224 15. In multiline mode when the newline sequence was set to "any", the pattern
225 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
226 This doesn't seem right; it now treats the CRLF combination as the line
227 ending, and so does not match in that case. It's only a pattern such as ^$
228 that would hit this one: something like ^ABC$ would have failed after \r
229 and then tried again after \r\n.
231 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
232 in an attempt to make files that differ only in their line terminators
233 compare equal. This works on Linux.
235 17. Under certain error circumstances pcregrep might try to free random memory
236 as it exited. This is now fixed, thanks to valgrind.
238 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
239 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
240 was because its rules for how to advance for /g after matching an empty
241 string at the end of a line did not allow for this case. They now check for
242 it specially.
244 20. pcretest is supposed to handle patterns and data of any length, by
245 extending its buffers when necessary. It was getting this wrong when the
246 buffer for a data line had to be extended.
248 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
249 CRLF as a newline sequence.
251 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
252 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
253 I have nevertheless tidied it up.
255 23. Added some casts to kill warnings from HP-UX ia64 compiler.
257 24. Added a man page for pcre-config.
260 Version 7.0 19-Dec-06
261 ---------------------
263 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
264 moving to gcc 4.1.1.
266 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
267 sys/time.h, which is documented as needed for this function. It doesn't
268 seem to matter on Linux, but it showed up on some releases of OS X.
270 3. It seems that there are systems where bytes whose values are greater than
271 127 match isprint() in the "C" locale. The "C" locale should be the
272 default when a C program starts up. In most systems, only ASCII printing
273 characters match isprint(). This difference caused the output from pcretest
274 to vary, making some of the tests fail. I have changed pcretest so that:
276 (a) When it is outputting text in the compiled version of a pattern, bytes
277 other than 32-126 are always shown as hex escapes.
279 (b) When it is outputting text that is a matched part of a subject string,
280 it does the same, unless a different locale has been set for the match
281 (using the /L modifier). In this case, it uses isprint() to decide.
283 4. Fixed a major bug that caused incorrect computation of the amount of memory
284 required for a compiled pattern when options that changed within the
285 pattern affected the logic of the preliminary scan that determines the
286 length. The relevant options are -x, and -i in UTF-8 mode. The result was
287 that the computed length was too small. The symptoms of this bug were
288 either the PCRE error "internal error: code overflow" from pcre_compile(),
289 or a glibc crash with a message such as "pcretest: free(): invalid next
290 size (fast)". Examples of patterns that provoked this bug (shown in
291 pcretest format) are:
293 /(?-x: )/x
294 /(?x)(?-x: \s*#\s*)/
295 /((?i)[\x{c0}])/8
296 /(?i:[\x{c0}])/8
298 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
299 is now done differently.
301 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
302 wrapper classes; (b) implement a new function in the C++ scanner that is
303 more efficient than the old way of doing things because it avoids levels of
304 recursion in the regex matching; (c) add a paragraph to the documentation
305 for the FullMatch() function.
307 6. The escape sequence \n was being treated as whatever was defined as
308 "newline". Not only was this contrary to the documentation, which states
309 that \n is character 10 (hex 0A), but it also went horribly wrong when
310 "newline" was defined as CRLF. This has been fixed.
312 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
313 was being set to -1 for the "end of line" case (supposedly a value that no
314 character can have). Though this value is never used (the check for end of
315 line is "zero bytes in current character"), it caused compiler complaints.
316 I've changed it to 0xffffffff.
318 8. In pcre_version.c, the version string was being built by a sequence of
319 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
320 string (as it is for production releases) called a macro with an empty
321 argument. The C standard says the result of this is undefined. The gcc
322 compiler treats it as an empty string (which was what was wanted) but it is
323 reported that Visual C gives an error. The source has been hacked around to
324 avoid this problem.
326 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
327 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
328 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
329 of them did).
331 10. Originally, pcretest opened its input and output without "b"; then I was
332 told that "b" was needed in some environments, so it was added for release
333 5.0 to both the input and output. (It makes no difference on Unix-like
334 systems.) Later I was told that it is wrong for the input on Windows. I've
335 now abstracted the modes into two macros, to make it easier to fiddle with
336 them, and removed "b" from the input mode under Windows.
338 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
340 12. Added -help and --help to pcretest as an official way of being reminded
341 of the options.
343 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
344 and pcrecpp.cc because they annoy compilers at high warning levels.
346 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
348 15. Fixed an occurrence of == in configure.ac that should have been = (shell
349 scripts are not C programs :-) and which was not noticed because it works
350 on Linux.
352 16. pcretest is supposed to handle any length of pattern and data line (as one
353 line or as a continued sequence of lines) by extending its input buffer if
354 necessary. This feature was broken for very long pattern lines, leading to
355 a string of junk being passed to pcre_compile() if the pattern was longer
356 than about 50K.
358 17. I have done a major re-factoring of the way pcre_compile() computes the
359 amount of memory needed for a compiled pattern. Previously, there was code
360 that made a preliminary scan of the pattern in order to do this. That was
361 OK when PCRE was new, but as the facilities have expanded, it has become
362 harder and harder to keep it in step with the real compile phase, and there
363 have been a number of bugs (see for example, 4 above). I have now found a
364 cunning way of running the real compile function in a "fake" mode that
365 enables it to compute how much memory it would need, while actually only
366 ever using a few hundred bytes of working memory and without too many
367 tests of the mode. This should make future maintenance and development
368 easier. A side effect of this work is that the limit of 200 on the nesting
369 depth of parentheses has been removed (though this was never a serious
370 limitation, I suspect). However, there is a downside: pcre_compile() now
371 runs more slowly than before (30% or more, depending on the pattern). I
372 hope this isn't a big issue. There is no effect on runtime performance.
374 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
375 newline (only possible for the last line of a file) and it was a
376 pattern that set a locale (followed by /Lsomething), pcretest crashed.
378 19. Added additional timing features to pcretest. (1) The -tm option now times
379 matching only, not compiling. (2) Both -t and -tm can be followed, as a
380 separate command line item, by a number that specifies the number of
381 repeats to use when timing. The default is 50000; this gives better
382 precision, but takes uncomfortably long for very large patterns.
384 20. Extended pcre_study() to be more clever in cases where a branch of a
385 subpattern has no definite first character. For example, (a*|b*)[cd] would
386 previously give no result from pcre_study(). Now it recognizes that the
387 first character must be a, b, c, or d.
389 21. There was an incorrect error "recursive call could loop indefinitely" if
390 a subpattern (or the entire pattern) that was being tested for matching an
391 empty string contained only one non-empty item after a nested subpattern.
392 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
393 incorrectly, because the \d was being skipped in the check.
395 22. The pcretest program now has a new pattern option /B and a command line
396 option -b, which is equivalent to adding /B to every pattern. This causes
397 it to show the compiled bytecode, without the additional information that
398 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
399 is the same as /B/I).
401 23. A new optimization is now able automatically to treat some sequences such
402 as a*b as a*+b. More specifically, if something simple (such as a character
403 or a simple class like \d) has an unlimited quantifier, and is followed by
404 something that cannot possibly match the quantified thing, the quantifier
405 is automatically "possessified".
407 24. A recursive reference to a subpattern whose number was greater than 39
408 went wrong under certain circumstances in UTF-8 mode. This bug could also
409 have affected the operation of pcre_study().
411 25. Realized that a little bit of performance could be had by replacing
412 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
414 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
416 27. Possessive quantifiers such as a++ were previously implemented by turning
417 them into atomic groups such as ($>a+). Now they have their own opcodes,
418 which improves performance. This includes the automatically created ones
419 from 23 above.
421 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
422 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
423 the first matched character to be a colon. This applied both to named and
424 numbered groups.
426 29. The ucpinternal.h header file was missing its idempotency #ifdef.
428 30. I was sent a "project" file called libpcre.a.dev which I understand makes
429 building PCRE on Windows easier, so I have included it in the distribution.
431 31. There is now a check in pcretest against a ridiculously large number being
432 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
433 loop, the loop is abandoned.
435 32. Forward references to subpatterns in conditions such as (?(2)...) where
436 subpattern 2 is defined later cause pcre_compile() to search forwards in
437 the pattern for the relevant set of parentheses. This search went wrong
438 when there were unescaped parentheses in a character class, parentheses
439 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
441 33. "Subroutine" calls and backreferences were previously restricted to
442 referencing subpatterns earlier in the regex. This restriction has now
443 been removed.
445 34. Added a number of extra features that are going to be in Perl 5.10. On the
446 whole, these are just syntactic alternatives for features that PCRE had
447 previously implemented using the Python syntax or my own invention. The
448 other formats are all retained for compatibility.
450 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
451 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
452 also .NET compatible.
454 (b) A recursion or subroutine call to a named group can now be defined as
455 (?&name) as well as (?P>name).
457 (c) A backreference to a named group can now be defined as \k<name> or
458 \k'name' as well as (?P=name). The new forms, as well as being in Perl
459 5.10, are also .NET compatible.
461 (d) A conditional reference to a named group can now use the syntax
462 (?(<name>) or (?('name') as well as (?(name).
464 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
465 groups (named and numbered) that are never evaluated inline, but can be
466 called as "subroutines" from elsewhere. In effect, the DEFINE condition
467 is always false. There may be only one alternative in such a group.
469 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
470 as the simple (?(R). The condition is true only if the most recent
471 recursion is that of the given number or name. It does not search out
472 through the entire recursion stack.
474 (g) The escape \gN or \g{N} has been added, where N is a positive or
475 negative number, specifying an absolute or relative reference.
477 35. Tidied to get rid of some further signed/unsigned compiler warnings and
478 some "unreachable code" warnings.
480 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
481 things, this adds five new scripts.
483 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
484 There were also incompatibilities regarding the handling of \Q..\E inside
485 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
486 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
488 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
489 matches an empty string, and forcibly breaks the loop. There were bugs in
490 this code in non-simple cases. For a pattern such as ^(a()*)* matched
491 against aaaa the result was just "a" rather than "aaaa", for example. Two
492 separate and independent bugs (that affected different cases) have been
493 fixed.
495 39. Refactored the code to abolish the use of different opcodes for small
496 capturing bracket numbers. This is a tidy that I avoided doing when I
497 removed the limit on the number of capturing brackets for 3.5 back in 2001.
498 The new approach is not only tidier, it makes it possible to reduce the
499 memory needed to fix the previous bug (38).
501 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
502 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
503 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
504 mode.
506 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
507 report.
509 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
510 copy construction and assignment in the C++ wrapper.
512 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
513 couple of bugs that could have given wrong results in the "--newline=crlf"
514 case.
516 44. Added a number of casts and did some reorganization of signed/unsigned int
517 variables following suggestions from Dair Grant. Also renamed the variable
518 "this" as "item" because it is a C++ keyword.
520 45. Arranged for dftables to add
522 #include "pcre_internal.h"
524 to pcre_chartables.c because without it, gcc 4.x may remove the array
525 definition from the final binary if PCRE is built into a static library and
526 dead code stripping is activated.
528 46. For an unanchored pattern, if a match attempt fails at the start of a
529 newline sequence, and the newline setting is CRLF or ANY, and the next two
530 characters are CRLF, advance by two characters instead of one.
533 Version 6.7 04-Jul-06
534 ---------------------
536 1. In order to handle tests when input lines are enormously long, pcretest has
537 been re-factored so that it automatically extends its buffers when
538 necessary. The code is crude, but this _is_ just a test program. The
539 default size has been increased from 32K to 50K.
541 2. The code in pcre_study() was using the value of the re argument before
542 testing it for NULL. (Of course, in any sensible call of the function, it
543 won't be NULL.)
545 3. The memmove() emulation function in pcre_internal.h, which is used on
546 systems that lack both memmove() and bcopy() - that is, hardly ever -
547 was missing a "static" storage class specifier.
549 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
550 containing an extended class (one that cannot be represented by a bitmap
551 because it contains high-valued characters or Unicode property items, e.g.
552 [\pZ]). Almost always one would set UTF-8 mode when processing such a
553 pattern, but PCRE should not loop if you do not (it no longer does).
554 [Detail: two cases were found: (a) a repeated subpattern containing an
555 extended class; (b) a recursive reference to a subpattern that followed a
556 previous extended class. It wasn't skipping over the extended class
557 correctly when UTF-8 mode was not set.]
559 5. A negated single-character class was not being recognized as fixed-length
560 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
561 compile error "lookbehind assertion is not fixed length".
563 6. The RunPerlTest auxiliary script was showing an unexpected difference
564 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
565 write a Perl script that can interpret lines of an input file either as
566 byte characters or as UTF-8, which is what "perltest" was being required to
567 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
568 can't do is switch easily at run time between having the "use utf8;" pragma
569 or not. In the end, I fudged it by using the RunPerlTest script to insert
570 "use utf8;" explicitly for the UTF-8 tests.
572 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
573 the end of the subject string, contrary to the documentation and to what
574 Perl does. This was true of both matching functions. Now it matches only at
575 the start of the subject and immediately after *internal* newlines.
577 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
578 a pointer to an int instead of a pointer to an unsigned long int. This
579 caused problems on 64-bit systems.
581 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
582 instance of the 'standard' template library not being so standard".
584 10. There was no check on the number of named subpatterns nor the maximum
585 length of a subpattern name. The product of these values is used to compute
586 the size of the memory block for a compiled pattern. By supplying a very
587 long subpattern name and a large number of named subpatterns, the size
588 computation could be caused to overflow. This is now prevented by limiting
589 the length of names to 32 characters, and the number of named subpatterns
590 to 10,000.
592 11. Subpatterns that are repeated with specific counts have to be replicated in
593 the compiled pattern. The size of memory for this was computed from the
594 length of the subpattern and the repeat count. The latter is limited to
595 65535, but there was no limit on the former, meaning that integer overflow
596 could in principle occur. The compiled length of a repeated subpattern is
597 now limited to 30,000 bytes in order to prevent this.
599 12. Added the optional facility to have named substrings with the same name.
601 13. Added the ability to use a named substring as a condition, using the
602 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
603 are numbers (not recommended). Forward references are permitted.
605 14. Added forward references in named backreferences (if you see what I mean).
607 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
608 pattern could run off the end of the subject. For example, the pattern
609 "(?s)(.{1,5})"8 did this with the subject "ab".
611 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
612 PCRE_CASELESS was set when matching characters that were quantified with ?
613 or *.
615 17. A character class other than a single negated character that had a minimum
616 but no maximum quantifier - for example [ab]{6,} - was not handled
617 correctly by pce_dfa_exec(). It would match only one character.
619 18. A valid (though odd) pattern that looked like a POSIX character
620 class but used an invalid character after [ (for example [[,abc,]]) caused
621 pcre_compile() to give the error "Failed: internal error: code overflow" or
622 in some cases to crash with a glibc free() error. This could even happen if
623 the pattern terminated after [[ but there just happened to be a sequence of
624 letters, a binary zero, and a closing ] in the memory that followed.
626 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
627 over the years. Originally (before any Unicode support), just the bottom 8
628 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
629 output from "man perlunicode" includes this:
631 The regular expression compiler produces polymorphic opcodes. That
632 is, the pattern adapts to the data and automatically switches to
633 the Unicode character scheme when presented with Unicode data--or
634 instead uses a traditional byte scheme when presented with byte
635 data.
637 Sadly, a wide octal escape does not cause a switch, and in a string with
638 no other multibyte characters, these octal escapes are treated as before.
639 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
640 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
641 Unicode string.
643 I have not perpetrated such confusion in PCRE. Up till now, it took just
644 the bottom 8 bits, as in old Perl. I have now made octal escapes with
645 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
646 translate to the appropriate multibyte character.
648 29. Applied some refactoring to reduce the number of warnings from Microsoft
649 and Borland compilers. This has included removing the fudge introduced
650 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
651 a warning about an unused variable.
653 21. PCRE has not included VT (character 0x0b) in the set of whitespace
654 characters since release 4.0, because Perl (from release 5.004) does not.
655 [Or at least, is documented not to: some releases seem to be in conflict
656 with the documentation.] However, when a pattern was studied with
657 pcre_study() and all its branches started with \s, PCRE still included VT
658 as a possible starting character. Of course, this did no harm; it just
659 caused an unnecessary match attempt.
661 22. Removed a now-redundant internal flag bit that recorded the fact that case
662 dependency changed within the pattern. This was once needed for "required
663 byte" processing, but is no longer used. This recovers a now-scarce options
664 bit. Also moved the least significant internal flag bit to the most-
665 significant bit of the word, which was not previously used (hangover from
666 the days when it was an int rather than a uint) to free up another bit for
667 the future.
669 23. Added support for CRLF line endings as well as CR and LF. As well as the
670 default being selectable at build time, it can now be changed at runtime
671 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
672 specify that it is scanning data with non-default line endings.
674 24. Changed the definition of CXXLINK to make it agree with the definition of
675 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
677 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
678 recursions. This makes a big different to stack usage for some patterns.
680 26. If a subpattern containing a named recursion or subroutine reference such
681 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
682 the space required for the compiled pattern went wrong and gave too small a
683 value. Depending on the environment, this could lead to "Failed: internal
684 error: code overflow at offset 49" or "glibc detected double free or
685 corruption" errors.
687 27. Applied patches from Google (a) to support the new newline modes and (b) to
688 advance over multibyte UTF-8 characters in GlobalReplace.
690 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
691 difference for some implementation of PCRE in some Windows version.
693 29. Added some extra testing facilities to pcretest:
695 \q<number> in a data line sets the "match limit" value
696 \Q<number> in a data line sets the "match recursion limt" value
697 -S <number> sets the stack size, where <number> is in megabytes
699 The -S option isn't available for Windows.
702 Version 6.6 06-Feb-06
703 ---------------------
705 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
706 in pcreposix.h. I have copied the definition from pcre.h.
708 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
709 because pcre.h is no longer a built file.
711 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
712 not normally included in the compiled code.
715 Version 6.5 01-Feb-06
716 ---------------------
718 1. When using the partial match feature with pcre_dfa_exec(), it was not
719 anchoring the second and subsequent partial matches at the new starting
720 point. This could lead to incorrect results. For example, with the pattern
721 /1234/, partially matching against "123" and then "a4" gave a match.
723 2. Changes to pcregrep:
725 (a) All non-match returns from pcre_exec() were being treated as failures
726 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
727 error message is output. Some extra information is given for the
729 probably the only errors that are likely to be caused by users (by
730 specifying a regex that has nested indefinite repeats, for instance).
731 If there are more than 20 of these errors, pcregrep is abandoned.
733 (b) A binary zero was treated as data while matching, but terminated the
734 output line if it was written out. This has been fixed: binary zeroes
735 are now no different to any other data bytes.
737 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
738 used to set a locale for matching. The --locale=xxxx long option has
739 been added (no short equivalent) to specify a locale explicitly on the
740 pcregrep command, overriding the environment variables.
742 (d) When -B was used with -n, some line numbers in the output were one less
743 than they should have been.
745 (e) Added the -o (--only-matching) option.
747 (f) If -A or -C was used with -c (count only), some lines of context were
748 accidentally printed for the final match.
750 (g) Added the -H (--with-filename) option.
752 (h) The combination of options -rh failed to suppress file names for files
753 that were found from directory arguments.
755 (i) Added the -D (--devices) and -d (--directories) options.
757 (j) Added the -F (--fixed-strings) option.
759 (k) Allow "-" to be used as a file name for -f as well as for a data file.
761 (l) Added the --colo(u)r option.
763 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
764 is not present by default.
766 3. A nasty bug was discovered in the handling of recursive patterns, that is,
767 items such as (?R) or (?1), when the recursion could match a number of
768 alternatives. If it matched one of the alternatives, but subsequently,
769 outside the recursion, there was a failure, the code tried to back up into
770 the recursion. However, because of the way PCRE is implemented, this is not
771 possible, and the result was an incorrect result from the match.
773 In order to prevent this happening, the specification of recursion has
774 been changed so that all such subpatterns are automatically treated as
775 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
777 4. I had overlooked the fact that, in some locales, there are characters for
778 which isalpha() is true but neither isupper() nor islower() are true. In
779 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
780 and ordfeminine) are like this. This affected the treatment of \w and \W
781 when they appeared in character classes, but not when they appeared outside
782 a character class. The bit map for "word" characters is now created
783 separately from the results of isalnum() instead of just taking it from the
784 upper, lower, and digit maps. (Plus the underscore character, of course.)
786 5. The above bug also affected the handling of POSIX character classes such as
787 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
788 permanent tables. Instead, the bit maps for such a class were previously
789 created as the appropriate unions of the upper, lower, and digit bitmaps.
790 Now they are created by subtraction from the [[:word:]] class, which has
791 its own bitmap.
793 6. The [[:blank:]] character class matches horizontal, but not vertical space.
794 It is created by subtracting the vertical space characters (\x09, \x0a,
795 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
796 subtraction was done in the overall bitmap for a character class, meaning
797 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
798 be recognized. This bug has been fixed.
800 7. Patches from the folks at Google:
802 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
803 real life, but is still worth protecting against".
805 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
806 regular expressions".
808 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
809 have it.
811 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
812 "configure" and the latter not, in order to fix a problem somebody had
813 with compiling the Arg class on HP-UX.
815 (e) Improve the error-handling of the C++ wrapper a little bit.
817 (f) New tests for checking recursion limiting.
819 8. The pcre_memmove() function, which is used only if the environment does not
820 have a standard memmove() function (and is therefore rarely compiled),
821 contained two bugs: (a) use of int instead of size_t, and (b) it was not
822 returning a result (though PCRE never actually uses the result).
824 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
825 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
826 returned instead of calling malloc() with an overflowing number that would
827 most likely cause subsequent chaos.
829 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
831 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
832 with this option is matched, the nmatch and pmatch options of regexec() are
833 ignored.
835 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
836 provided in case anyone wants to the the POSIX interface with UTF-8
837 strings.
839 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
840 C++ linking (needed for some HP-UX environments).
842 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
843 (unused parameter) and in the pcre_printint() function (omitted "default"
844 switch label when the default is to do nothing).
846 15. Added some code to make it possible, when PCRE is compiled as a C++
847 library, to replace subject pointers for pcre_exec() with a smart pointer
848 class, thus making it possible to process discontinuous strings.
850 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
851 much the same function. They were added by different people who were trying
852 to make PCRE easy to compile on non-Unix systems. It has been suggested
853 that PCRE_EXPORT be abolished now that there is more automatic apparatus
854 for compiling on Windows systems. I have therefore replaced it with
855 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
856 defaults to "extern" for C or "extern C" for C++, which works fine on
857 Unix-like systems. It is now possible to override the value of PCRE_DATA_
858 SCOPE with something explicit in config.h. In addition:
860 (a) pcreposix.h still had just "extern" instead of either of these macros;
861 I have replaced it with PCRE_DATA_SCOPE.
863 (b) Functions such as _pcre_xclass(), which are internal to the library,
864 but external in the C sense, all had PCRE_EXPORT in their definitions.
865 This is apparently wrong for the Windows case, so I have removed it.
866 (It makes no difference on Unix-like systems.)
868 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
869 of recursive calls to match(). This is different to MATCH_LIMIT because
870 that limits the total number of calls to match(), not all of which increase
871 the depth of recursion. Limiting the recursion depth limits the amount of
872 stack (or heap if NO_RECURSE is set) that is used. The default can be set
873 when PCRE is compiled, and changed at run time. A patch from Google adds
874 this functionality to the C++ interface.
876 18. Changes to the handling of Unicode character properties:
878 (a) Updated the table to Unicode 4.1.0.
880 (b) Recognize characters that are not in the table as "Cn" (undefined).
882 (c) I revised the way the table is implemented to a much improved format
883 which includes recognition of ranges. It now supports the ranges that
884 are defined in UnicodeData.txt, and it also amalgamates other
885 characters into ranges. This has reduced the number of entries in the
886 table from around 16,000 to around 3,000, thus reducing its size
887 considerably. I realized I did not need to use a tree structure after
888 all - a binary chop search is just as efficient. Having reduced the
889 number of entries, I extended their size from 6 bytes to 8 bytes to
890 allow for more data.
892 (d) Added support for Unicode script names via properties such as \p{Han}.
894 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
895 matching that character.
897 20. When matching a repeated Unicode property with a minimum greater than zero,
898 (for example \pL{2,}), PCRE could look past the end of the subject if it
899 reached it while seeking the minimum number of characters. This could
900 happen only if some of the characters were more than one byte long, because
901 there is a check for at least the minimum number of bytes.
903 21. Refactored the implementation of \p and \P so as to be more general, to
904 allow for more different types of property in future. This has changed the
905 compiled form incompatibly. Anybody with saved compiled patterns that use
906 \p or \P will have to recompile them.
908 22. Added "Any" and "L&" to the supported property types.
910 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
911 but give a compile time error if the value is greater than 0xff.
913 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
914 accidentally not being installed or uninstalled.
916 25. The pcre.h file was built from pcre.h.in, but the only changes that were
917 made were to insert the current release number. This seemed silly, because
918 it made things harder for people building PCRE on systems that don't run
919 "configure". I have turned pcre.h into a distributed file, no longer built
920 by "configure", with the version identification directly included. There is
921 no longer a pcre.h.in file.
923 However, this change necessitated a change to the pcre-config script as
924 well. It is built from pcre-config.in, and one of the substitutions was the
925 release number. I have updated configure.ac so that ./configure now finds
926 the release number by grepping pcre.h.
928 26. Added the ability to run the tests under valgrind.
931 Version 6.4 05-Sep-05
932 ---------------------
934 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
935 "--" to be printed when multiple files were scanned, even when none of the
936 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
937 consider it to be a bug, and have restored the previous behaviour.
939 2. A couple of code tidies to get rid of compiler warnings.
941 3. The pcretest program used to cheat by referring to symbols in the library
942 whose names begin with _pcre_. These are internal symbols that are not
943 really supposed to be visible externally, and in some environments it is
944 possible to suppress them. The cheating is now confined to including
945 certain files from the library's source, which is a bit cleaner.
947 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
948 file's purpose clearer.
950 5. Reorganized pcre_ucp_findchar().
953 Version 6.3 15-Aug-05
954 ---------------------
956 1. The file libpcre.pc.in did not have general read permission in the tarball.
958 2. There were some problems when building without C++ support:
960 (a) If C++ support was not built, "make install" and "make test" still
961 tried to test it.
963 (b) There were problems when the value of CXX was explicitly set. Some
964 changes have been made to try to fix these, and ...
966 (c) --disable-cpp can now be used to explicitly disable C++ support.
968 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
969 backslash in a target when C++ was disabled. This confuses some
970 versions of "make", apparently. Using an intermediate variable solves
971 this. (Same for CPP_LOBJ.)
973 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
974 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
975 necessary on certain architectures.
977 4. Added a setting of -export-symbols-regex to the link command to remove
978 those symbols that are exported in the C sense, but actually are local
979 within the library, and not documented. Their names all begin with
980 "_pcre_". This is not a perfect job, because (a) we have to except some
981 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
982 available (and never for static libraries). I have made a note to try to
983 find a way round (a) in the future.
986 Version 6.2 01-Aug-05
987 ---------------------
989 1. There was no test for integer overflow of quantifier values. A construction
990 such as {1111111111111111} would give undefined results. What is worse, if
991 a minimum quantifier for a parenthesized subpattern overflowed and became
992 negative, the calculation of the memory size went wrong. This could have
993 led to memory overwriting.
995 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
997 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
998 operating environments where this matters.
1000 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1001 PCRE options from within the C++ wrapper.
1003 5. Named capturing subpatterns were not being correctly counted when a pattern
1004 was compiled. This caused two problems: (a) If there were more than 100
1005 such subpatterns, the calculation of the memory needed for the whole
1006 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1007 back references of the form \12, where the number was greater than 9, were
1008 not recognized as back references, even though there were sufficient
1009 previous subpatterns.
1011 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1012 versions of gcc, e.g. 2.95.4.
1015 Version 6.1 21-Jun-05
1016 ---------------------
1018 1. There was one reference to the variable "posix" in pcretest.c that was not
1019 surrounded by "#if !defined NOPOSIX".
1021 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1022 the cross-check on the old pcre_info() function, for the benefit of the
1023 cut-down version of PCRE that is currently imported into Exim.
1025 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1026 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1027 patterns that start (?i) but I don't think that matters. The use of (?i) is
1028 just an example; this all applies to the other options as well.
1030 4. Since libtool seems to echo the compile commands it is issuing, the output
1031 from "make" can be reduced a bit by putting "@" in front of each libtool
1032 compile command.
1034 5. Patch from the folks at Google for configure.in to be a bit more thorough
1035 in checking for a suitable C++ installation before trying to compile the
1036 C++ stuff. This should fix a reported problem when a compiler was present,
1037 but no suitable headers.
1039 6. The man pages all had just "PCRE" as their title. I have changed them to
1040 be the relevant file name. I have also arranged that these names are
1041 retained in the file doc/pcre.txt, which is a concatenation in text format
1042 of all the man pages except the little individual ones for each function.
1044 7. The NON-UNIX-USE file had not been updated for the different set of source
1045 files that come with release 6. I also added a few comments about the C++
1046 wrapper.
1049 Version 6.0 07-Jun-05
1050 ---------------------
1052 1. Some minor internal re-organization to help with my DFA experiments.
1054 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1055 didn't matter for the library itself when fully configured, but did matter
1056 when compiling without UCP support, or within Exim, where the ucp files are
1057 not imported.
1059 3. Refactoring of the library code to split up the various functions into
1060 different source modules. The addition of the new DFA matching code (see
1061 below) to a single monolithic source would have made it really too
1062 unwieldy, quite apart from causing all the code to be include in a
1063 statically linked application, when only some functions are used. This is
1064 relevant even without the DFA addition now that patterns can be compiled in
1065 one application and matched in another.
1067 The downside of splitting up is that there have to be some external
1068 functions and data tables that are used internally in different modules of
1069 the library but which are not part of the API. These have all had their
1070 names changed to start with "_pcre_" so that they are unlikely to clash
1071 with other external names.
1073 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1074 a different (DFA) algorithm. Although it is slower than the original
1075 function, it does have some advantages for certain types of matching
1076 problem.
1078 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1079 including restarting after a partial match.
1081 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1082 defined when compiling for Windows was sent to me. I have put it into the
1083 code, though I have no means of testing or verifying it.
1085 7. Added the pcre_refcount() auxiliary function.
1087 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1088 match before or at the first newline in the subject string. In pcretest,
1089 the /f option on a pattern can be used to set this.
1091 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1092 would behave wrongly. This has been present in PCRE since release 4.0.
1094 10. A number of changes to the pcregrep command:
1096 (a) Refactored how -x works; insert ^(...)$ instead of setting
1097 PCRE_ANCHORED and checking the length, in preparation for adding
1098 something similar for -w.
1100 (b) Added the -w (match as a word) option.
1102 (c) Refactored the way lines are read and buffered so as to have more
1103 than one at a time available.
1105 (d) Implemented a pcregrep test script.
1107 (e) Added the -M (multiline match) option. This allows patterns to match
1108 over several lines of the subject. The buffering ensures that at least
1109 8K, or the rest of the document (whichever is the shorter) is available
1110 for matching (and similarly the previous 8K for lookbehind assertions).
1112 (f) Changed the --help output so that it now says
1114 -w, --word-regex(p)
1116 instead of two lines, one with "regex" and the other with "regexp"
1117 because that confused at least one person since the short forms are the
1118 same. (This required a bit of code, as the output is generated
1119 automatically from a table. It wasn't just a text change.)
1121 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1122 option but starts with a hyphen. Could be a pattern or a path name
1123 starting with a hyphen, for instance.
1125 (h) "-" can be given as a file name to represent stdin.
1127 (i) When file names are being printed, "(standard input)" is used for
1128 the standard input, for compatibility with GNU grep. Previously
1129 "<stdin>" was used.
1131 (j) The option --label=xxx can be used to supply a name to be used for
1132 stdin when file names are being printed. There is no short form.
1134 (k) Re-factored the options decoding logic because we are going to add
1135 two more options that take data. Such options can now be given in four
1136 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1138 (l) Added the -A, -B, and -C options for requesting that lines of context
1139 around matches be printed.
1141 (m) Added the -L option to print the names of files that do not contain
1142 any matching lines, that is, the complement of -l.
1144 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1145 continue to scan other files.
1147 (o) The -s option was incorrectly implemented. For compatibility with other
1148 greps, it now suppresses the error message for a non-existent or non-
1149 accessible file (but not the return code). There is a new option called
1150 -q that suppresses the output of matching lines, which was what -s was
1151 previously doing.
1153 (p) Added --include and --exclude options to specify files for inclusion
1154 and exclusion when recursing.
1156 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1157 Hopefully, it now does.
1159 12. Missing cast in pcre_study().
1161 13. Added an "uninstall" target to the makefile.
1163 14. Replaced "extern" in the function prototypes in Makefile.in with
1164 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1165 world, but is set differently for Windows.
1167 15. Added a second compiling function called pcre_compile2(). The only
1168 difference is that it has an extra argument, which is a pointer to an
1169 integer error code. When there is a compile-time failure, this is set
1170 non-zero, in addition to the error test pointer being set to point to an
1171 error message. The new argument may be NULL if no error number is required
1172 (but then you may as well call pcre_compile(), which is now just a
1173 wrapper). This facility is provided because some applications need a
1174 numeric error indication, but it has also enabled me to tidy up the way
1175 compile-time errors are handled in the POSIX wrapper.
1177 16. Added VPATH=.libs to the makefile; this should help when building with one
1178 prefix path and installing with another. (Or so I'm told by someone who
1179 knows more about this stuff than I do.)
1181 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1182 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1183 match everything, including newlines. This is not POSIX-compatible, but
1184 somebody wanted the feature. From pcretest it can be activated by using
1185 both the P and the s flags.
1187 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1189 19. libpcre.pc was being incorrectly installed as executable.
1191 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1192 it now also looks for '\r' so that it will work unmodified on Windows.
1194 21. Added Google's contributed C++ wrapper to the distribution.
1196 22. Added some untidy missing memory free() calls in pcretest, to keep
1197 Electric Fence happy when testing.
1201 Version 5.0 13-Sep-04
1202 ---------------------
1204 1. Internal change: literal characters are no longer packed up into items
1205 containing multiple characters in a single byte-string. Each character
1206 is now matched using a separate opcode. However, there may be more than one
1207 byte in the character in UTF-8 mode.
1209 2. The pcre_callout_block structure has two new fields: pattern_position and
1210 next_item_length. These contain the offset in the pattern to the next match
1211 item, and its length, respectively.
1213 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1214 insertion of callouts before each pattern item. Added the /C option to
1215 pcretest to make use of this.
1217 4. On the advice of a Windows user, the lines
1219 #if defined(_WIN32) || defined(WIN32)
1220 _setmode( _fileno( stdout ), 0x8000 );
1221 #endif /* defined(_WIN32) || defined(WIN32) */
1223 have been added to the source of pcretest. This apparently does useful
1224 magic in relation to line terminators.
1226 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1227 for the benefit of those environments where the "b" makes a difference.
1229 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1230 to know about it. I have put a hack into configure.in that adds in code
1231 to set GCC=yes if CC=icc. This seems to end up at a point in the
1232 generated configure script that is early enough to affect the setting of
1233 compiler options, which is what is needed, but I have no means of testing
1234 whether it really works. (The user who reported this had patched the
1235 generated configure script, which of course I cannot do.)
1237 LATER: After change 22 below (new libtool files), the configure script
1238 seems to know about icc (and also ecc). Therefore, I have commented out
1239 this hack in configure.in.
1241 7. Added support for pkg-config (2 patches were sent in).
1243 8. Negated POSIX character classes that used a combination of internal tables
1244 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1245 [[:^ascii]]. Typically, they would match almost any characters. The other
1246 POSIX classes were not broken in this way.
1248 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1249 to find the match, as PCRE was deluded into thinking that the match had to
1250 start at the start point or following a newline. The same bug applied to
1251 patterns with negative forward assertions or any backward assertions
1252 preceding ".*" at the start, unless the pattern required a fixed first
1253 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1255 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1256 starting at the last subject character, bytes beyond the end of the subject
1257 string were read.
1259 11. Renamed the variable "class" as "classbits" to make life easier for C++
1260 users. (Previously there was a macro definition, but it apparently wasn't
1261 enough.)
1263 12. Added the new field "tables" to the extra data so that tables can be passed
1264 in at exec time, or the internal tables can be re-selected. This allows
1265 a compiled regex to be saved and re-used at a later time by a different
1266 program that might have everything at different addresses.
1268 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1269 -R library as well as a -L library.
1271 14. The debugging options of pcretest (-d on the command line or D on a
1272 pattern) showed incorrect output for anything following an extended class
1273 that contained multibyte characters and which was followed by a quantifier.
1275 15. Added optional support for general category Unicode character properties
1276 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1277 support. It adds about 90K to the size of the library. The meanings of the
1278 inbuilt class escapes such as \d and \s have NOT been changed.
1280 16. Updated pcredemo.c to include calls to free() to release the memory for the
1281 compiled pattern.
1283 17. The generated file chartables.c was being created in the source directory
1284 instead of in the building directory. This caused the build to fail if the
1285 source directory was different from the building directory, and was
1286 read-only.
1288 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1289 file. No doubt somebody will tell me if they don't make sense... Also added
1290 Dan Mooney's comments about building on OpenVMS.
1292 19. Added support for partial matching via the PCRE_PARTIAL option for
1293 pcre_exec() and the \P data escape in pcretest.
1295 20. Extended pcretest with 3 new pattern features:
1297 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1298 write the compiled pattern to the file whose name is "rest-of-line".
1299 This is a straight binary dump of the data, with the saved pointer to
1300 the character tables forced to be NULL. The study data, if any, is
1301 written too. After writing, pcretest reads a new pattern.
1303 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1304 compiled pattern from the given file. There must not be any
1305 occurrences of "<" in the file name (pretty unlikely); if there are,
1306 pcretest will instead treat the initial "<" as a pattern delimiter.
1307 After reading in the pattern, pcretest goes on to read data lines as
1308 usual.
1310 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1311 and 16-bit fields in a compiled pattern, to simulate a pattern that
1312 was compiled on a host of opposite endianness.
1314 21. The pcre-exec() function can now cope with patterns that were compiled on
1315 hosts of opposite endianness, with this restriction:
1317 As for any compiled expression that is saved and used later, the tables
1318 pointer field cannot be preserved; the extra_data field in the arguments
1319 to pcre_exec() should be used to pass in a tables address if a value
1320 other than the default internal tables were used at compile time.
1322 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1323 now diagnosed as an error. Previously, most of the time, a negative number
1324 would have been treated as zero, but if in addition "ovector" was passed as
1325 NULL, a crash could occur.
1327 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1328 new versions from the libtool 1.5 distribution (the last one is a copy of
1329 a file called libtool.m4). This seems to have fixed the need to patch
1330 "configure" to support Darwin 1.3 (which I used to do). However, I still
1331 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1332 workstation).
1334 24. Changed the PCRE licence to be the more standard "BSD" licence.
1337 Version 4.5 01-Dec-03
1338 ---------------------
1340 1. There has been some re-arrangement of the code for the match() function so
1341 that it can be compiled in a version that does not call itself recursively.
1342 Instead, it keeps those local variables that need separate instances for
1343 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1344 needs to "recurse". Keeping track of where control must go is done by means
1345 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1346 hide most of the details from the main code, and operates only if
1347 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1348 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1349 operating.
1351 To make it easier for callers to provide specially tailored get/free
1352 functions for this usage, two new functions, pcre_stack_malloc, and
1353 pcre_stack_free, are used. They are always called in strict stacking order,
1354 and the size of block requested is always the same.
1356 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1357 PCRE has been compiled to use the stack or the heap for recursion. The
1358 -C option of pcretest uses this to show which version is compiled.
1360 A new data escape \S, is added to pcretest; it causes the amounts of store
1361 obtained and freed by both kinds of malloc/free at match time to be added
1362 to the output.
1364 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1365 what's available on my current Linux desktop machine.
1367 3. When matching a UTF-8 string, the test for a valid string at the start has
1368 been extended. If start_offset is not zero, PCRE now checks that it points
1369 to a byte that is the start of a UTF-8 character. If not, it returns
1370 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1371 this is necessary because there may be backward assertions in the pattern.
1372 When matching the same subject several times, it may save resources to use
1373 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1375 4. The code for checking the validity of UTF-8 strings has been tightened so
1376 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1377 containing "overlong sequences".
1379 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1380 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1381 should have been "&", but it just so happened that all the cases this let
1382 through by mistake were picked up later in the function.
1384 6. I had used a variable called "isblank" - this is a C99 function, causing
1385 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1387 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1388 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1389 so the version is shown; (c) stop "make test" echoing "./RunTest".
1391 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1393 9. The prototype for memmove() for systems that don't have it was using
1394 size_t, but the inclusion of the header that defines size_t was later. I've
1395 moved the #includes for the C headers earlier to avoid this.
1397 10. Added some adjustments to the code to make it easier to compiler on certain
1398 special systems:
1400 (a) Some "const" qualifiers were missing.
1401 (b) Added the macro EXPORT before all exported functions; by default this
1402 is defined to be empty.
1403 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1404 that it reads its output file name as an argument instead of writing
1405 to the standard output and assuming this can be redirected.
1407 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1408 class containing characters with values greater than 255, PCRE compilation
1409 went into a loop.
1411 12. A recursive reference to a subpattern that was within another subpattern
1412 that had a minimum quantifier of zero caused PCRE to crash. For example,
1413 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1414 recursion. If the recursively-called subpattern itself had a zero repeat,
1415 that was OK.
1417 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1418 buffer into which it was copied (for escape processing) was still set at
1419 1024, so long lines caused crashes.
1421 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1422 "internal error: code overflow...". This applied to any character class
1423 that was followed by a possessive quantifier.
1425 15. Modified the Makefile to add libpcre.la as a prerequisite for
1426 libpcreposix.la because I was told this is needed for a parallel build to
1427 work.
1429 16. If a pattern that contained .* following optional items at the start was
1430 studied, the wrong optimizing data was generated, leading to matching
1431 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1432 matching string must start with a or b or c. The correct conclusion for
1433 this pattern is that a match can start with any character.
1436 Version 4.4 13-Aug-03
1437 ---------------------
1439 1. In UTF-8 mode, a character class containing characters with values between
1440 127 and 255 was not handled correctly if the compiled pattern was studied.
1441 In fixing this, I have also improved the studying algorithm for such
1442 classes (slightly).
1444 2. Three internal functions had redundant arguments passed to them. Removal
1445 might give a very teeny performance improvement.
1447 3. Documentation bug: the value of the capture_top field in a callout is *one
1448 more than* the number of the hightest numbered captured substring.
1450 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1451 in incorrectly linking with a previously installed version. They now link
1452 explicitly with libpcre.la.
1454 5. configure.in no longer needs to recognize Cygwin specially.
1456 6. A problem in pcre.in for Windows platforms is fixed.
1458 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1459 pcretest, it used to include the size of the study block as part of its
1460 output. Unfortunately, the structure contains a field that has a different
1461 size on different hardware architectures. This meant that the tests that
1462 showed this size failed. As the block is currently always of a fixed size,
1463 this information isn't actually particularly useful in pcretest output, so
1464 I have just removed it.
1466 8. Three pre-processor statements accidentally did not start in column 1.
1467 Sadly, there are *still* compilers around that complain, even though
1468 standard C has not required this for well over a decade. Sigh.
1470 9. In pcretest, the code for checking callouts passed small integers in the
1471 callout_data field, which is a void * field. However, some picky compilers
1472 complained about the casts involved for this on 64-bit systems. Now
1473 pcretest passes the address of the small integer instead, which should get
1474 rid of the warnings.
1476 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1477 both compile and run time, and gives an error if an invalid UTF-8 sequence
1478 is found. There is a option for disabling this check in cases where the
1479 string is known to be correct and/or the maximum performance is wanted.
1481 11. In response to a bug report, I changed one line in Makefile.in from
1483 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1484 to
1485 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1487 to look similar to other lines, but I have no way of telling whether this
1488 is the right thing to do, as I do not use Windows. No doubt I'll get told
1489 if it's wrong...
1492 Version 4.3 21-May-03
1493 ---------------------
1495 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1496 Makefile.
1498 2. Some refactoring to improve the quality of the code:
1500 (i) The utf8_table... variables are now declared "const".
1502 (ii) The code for \cx, which used the "case flipping" table to upper case
1503 lower case letters, now just substracts 32. This is ASCII-specific,
1504 but the whole concept of \cx is ASCII-specific, so it seems
1505 reasonable.
1507 (iii) PCRE was using its character types table to recognize decimal and
1508 hexadecimal digits in the pattern. This is silly, because it handles
1509 only 0-9, a-f, and A-F, but the character types table is locale-
1510 specific, which means strange things might happen. A private
1511 table is now used for this - though it costs 256 bytes, a table is
1512 much faster than multiple explicit tests. Of course, the standard
1513 character types table is still used for matching digits in subject
1514 strings against \d.
1516 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1517 ending in _t are). So I've renamed it as ESC_tee.
1519 3. The first argument for regexec() in the POSIX wrapper should have been
1520 defined as "const".
1522 4. Changed pcretest to use malloc() for its buffers so that they can be
1523 Electric Fenced for debugging.
1525 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1526 to read one or more bytes before the start of the subject string. Often this
1527 had no effect on PCRE's behaviour, but in some circumstances it could
1528 provoke a segmentation fault.
1530 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1531 to try to read one or more bytes before the start of the subject string.
1533 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1534 UTF-8 support could misbehave in various ways if the subject string
1535 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1536 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1537 back over UTF-8 characters.)
1540 Version 4.2 14-Apr-03
1541 ---------------------
1543 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1545 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1546 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1547 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1548 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1549 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1551 Note: automatic setting of the BUILD variables is not yet working
1552 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1553 compile-time but not at link-time
1554 [LINK]: use for linking executables only
1555 make different versions for Windows and non-Windows
1556 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1557 libraries
1558 [LINK_FOR_BUILD]: new variable
1559 [OBJEXT]: use throughout
1560 [EXEEXT]: use throughout
1561 <winshared>: new target
1562 <wininstall>: new target
1563 <dftables.o>: use native compiler
1564 <dftables>: use native linker
1565 <install>: handle Windows platform correctly
1566 <clean>: ditto
1567 <check>: ditto
1568 copy DLL to top builddir before testing
1570 As part of these changes, -no-undefined was removed again. This was reported
1571 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1572 in any case.
1574 3. Some tidies to get rid of compiler warnings:
1576 . In the match_data structure, match_limit was an unsigned long int, whereas
1577 match_call_count was an int. I've made them both unsigned long ints.
1579 . In pcretest the fact that a const uschar * doesn't automatically cast to
1580 a void * provoked a warning.
1582 . Turning on some more compiler warnings threw up some "shadow" variables
1583 and a few more missing casts.
1585 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1586 option, a class that contained a single character with a value between 128
1587 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1589 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1590 option, a class that contained several characters, but with at least one
1591 whose value was between 128 and 255 caused PCRE to crash.
1594 Version 4.1 12-Mar-03
1595 ---------------------
1597 1. Compiling with gcc -pedantic found a couple of places where casts were
1598 needed, and a string in dftables.c that was longer than standard compilers are
1599 required to support.
1601 2. Compiling with Sun's compiler found a few more places where the code could
1602 be tidied up in order to avoid warnings.
1604 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1605 first of these names is deprecated in the latest Autoconf in favour of the name
1606 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1607 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1608 analogy I have changed it to CFLAGS_FOR_BUILD.
1610 4. Added -no-undefined to the linking command in the Makefile, because this is
1611 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1612 linking step for the pcreposix library.
1614 5. PCRE was failing to diagnose the case of two named groups with the same
1615 name.
1617 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1618 literal character that is needed in the subject for a match, and scans along to
1619 ensure that it is present before embarking on the full matching process. This
1620 saves time in cases of nested unlimited repeats that are never going to match.
1621 Problem: the scan can take a lot of time if the subject is very long (e.g.
1622 megabytes), thus penalizing straightforward matches. It is now done only if the
1623 amount of subject to be scanned is less than 1000 bytes.
1625 7. A lesser problem with the same optimization is that it was recording the
1626 first character of an anchored pattern as "needed", thus provoking a search
1627 right along the subject, even when the first match of the pattern was going to
1628 fail. The "needed" character is now not set for anchored patterns, unless it
1629 follows something in the pattern that is of non-fixed length. Thus, it still
1630 fulfils its original purpose of finding quick non-matches in cases of nested
1631 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1634 Version 4.0 17-Feb-03
1635 ---------------------
1637 1. If a comment in an extended regex that started immediately after a meta-item
1638 extended to the end of string, PCRE compiled incorrect data. This could lead to
1639 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1641 2. Moved to autoconf 2.53 and libtool 1.4.2.
1643 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1644 the special perltest8 script is no longer needed - all the tests can be run
1645 from a single perltest script.
1647 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1648 by \s. It has now been removed in PCRE. This means it isn't recognized as
1649 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1650 class [:space:] *does* include VT, thereby creating a mess.
1652 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1653 space and tab.
1655 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1656 its new features into the main test script, reducing the number of scripts.
1658 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1659 were backward compatible, and made the (?i) apply to the whole pattern, as if
1660 /i were given. Now it behaves more logically, and applies the option setting
1661 only to what follows. PCRE has been changed to follow suit. However, if it
1662 finds options settings right at the start of the pattern, it extracts them into
1663 the global options, as before. Thus, they show up in the info data.
1665 8. Added support for the \Q...\E escape sequence. Characters in between are
1666 treated as literals. This is slightly different from Perl in that $ and @ are
1667 also handled as literals inside the quotes. In Perl, they will cause variable
1668 interpolation. Note the following examples:
1670 Pattern PCRE matches Perl matches
1672 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1673 \Qabc\$xyz\E abc\$xyz abc\$xyz
1674 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1676 For compatibility with Perl, \Q...\E sequences are recognized inside character
1677 classes as well as outside them.
1679 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1680 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1681 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1682 signed/unsigned warnings.
1684 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1685 option for pcretest, so I've replaced it by a simple function that does just
1686 that job.
1688 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1689 "pcregrep -".
1691 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1692 Java package. This provides some syntactic sugar for simple cases of what my
1693 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1694 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1695 item, you can use this simplified notation. Note that only makes sense with
1696 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1697 greediness, whatever the setting of the PCRE_UNGREEDY option.
1699 13. A change of greediness default within a pattern was not taking effect at
1700 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1701 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1702 was abstracted outside.
1704 14. PCRE now supports the \G assertion. It is true when the current matching
1705 position is at the start point of the match. This differs from \A when the
1706 starting offset is non-zero. Used with the /g option of pcretest (or similar
1707 code), it works in the same way as it does for Perl's /g option. If all
1708 alternatives of a regex begin with \G, the expression is anchored to the start
1709 match position, and the "anchored" flag is set in the compiled expression.
1711 15. Some bugs concerning the handling of certain option changes within patterns
1712 have been fixed. These applied to options other than (?ims). For example,
1713 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1714 been the other way round. Some of this was related to change 7 above.
1716 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1717 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1718 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1719 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1721 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1722 mode. Unlike ".", it always matches newline, whatever the setting of
1723 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1724 assertions. Perl allows it, but it doesn't (in general) work because it can't
1725 calculate the length of the lookbehind. At least, that's the case for Perl
1726 5.8.0 - I've been told they are going to document that it doesn't work in
1727 future.
1729 18. Added an error diagnosis for escapes that PCRE does not support: these are
1730 \L, \l, \N, \P, \p, \U, \u, and \X.
1732 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1733 reading past the end of the pattern in cases such as /[abcd/.
1735 20. PCRE was getting more memory than necessary for patterns with classes that
1736 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1738 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1739 compiling PCRE for use with Virtual Pascal.
1741 22. Small fix to the Makefile to make it work properly if the build is done
1742 outside the source tree.
1744 23. Added a new extension: a condition to go with recursion. If a conditional
1745 subpattern starts with (?(R) the "true" branch is used if recursion has
1746 happened, whereas the "false" branch is used only at the top level.
1748 24. When there was a very long string of literal characters (over 255 bytes
1749 without UTF support, over 250 bytes with UTF support), the computation of how
1750 much memory was required could be incorrect, leading to segfaults or other
1751 strange effects.
1753 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1754 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1755 there was a subsequent back reference to those brackets. This meant that, for
1756 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1757 possible to check for precisely this case. All we can do is abandon the
1758 optimization if .* occurs inside capturing brackets when there are any back
1759 references whatsoever. (See below for a better fix that came later.)
1761 26. The handling of the optimization for finding the first character of a
1762 non-anchored pattern, and for finding a character that is required later in the
1763 match were failing in some cases. This didn't break the matching; it just
1764 failed to optimize when it could. The way this is done has been re-implemented.
1766 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1768 28. Added a new feature that provides some of the functionality that Perl
1769 provides with (?{...}). The facility is termed a "callout". The way it is done
1770 in PCRE is for the caller to provide an optional function, by setting
1771 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1772 global variable. By default it is unset, which disables all calling out. To get
1773 the function called, the regex must include (?C) at appropriate points. This
1774 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1775 This provides a means of identifying different callout points. When PCRE
1776 reaches such a point in the regex, if pcre_callout has been set, the external
1777 function is called. It is provided with data in a structure called
1778 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1779 matching continues; if it returns a non-zero value, the match at the current
1780 point fails. However, backtracking will occur if possible. [This was changed
1781 later and other features added - see item 49 below.]
1783 29. pcretest is upgraded to test the callout functionality. It provides a
1784 callout function that displays information. By default, it shows the start of
1785 the match and the current position in the text. There are some new data escapes
1786 to vary what happens:
1788 \C+ in addition, show current contents of captured substrings
1789 \C- do not supply a callout function
1790 \C!n return 1 when callout number n is reached
1791 \C!n!m return 1 when callout number n is reached for the mth time
1793 30. If pcregrep was called with the -l option and just a single file name, it
1794 output "<stdin>" if a match was found, instead of the file name.
1796 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1797 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1798 pcre_exec(). This saves a malloc/free per call. The default value of
1799 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1800 when configuring.
1802 32. The default maximum size of a compiled pattern is 64K. There have been a
1803 few cases of people hitting this limit. The code now uses macros to handle the
1804 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1805 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1806 configuring. Tests 2 and 5 work only with 2-byte links because they output
1807 debugging information about compiled patterns.
1809 33. Internal code re-arrangements:
1811 (a) Moved the debugging function for printing out a compiled regex into
1812 its own source file (printint.c) and used #include to pull it into
1813 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1814 separate copies.
1816 (b) Defined the list of op-code names for debugging as a macro in
1817 internal.h so that it is next to the definition of the opcodes.
1819 (c) Defined a table of op-code lengths for simpler skipping along compiled
1820 code. This is again a macro in internal.h so that it is next to the
1821 definition of the opcodes.
1823 34. Added support for recursive calls to individual subpatterns, along the
1824 lines of Robin Houston's patch (but implemented somewhat differently).
1826 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1827 allow it to read and process whole directories in Win32. This code was
1828 contributed by Lionel Fourquaux; it has not been tested by me.
1830 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1831 used to name a group. Names consist of alphanumerics and underscores, and must
1832 be unique. Back references use the syntax (?P=name) and recursive calls use
1833 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1834 numbers. The function pcre_fullinfo() can be used after compilation to extract
1835 a name/number map. There are three relevant calls:
1837 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1838 PCRE_INFO_NAMECOUNT yields the number of entries
1839 PCRE_INFO_NAMETABLE yields a pointer to the map.
1841 The map is a vector of fixed-size entries. The size of each entry depends on
1842 the length of the longest name used. The first two bytes of each entry are the
1843 group number, most significant byte first. There follows the corresponding
1844 name, zero terminated. The names are in alphabetical order.
1846 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1847 case instead of 255. Making it the same both with and without UTF-8 support
1848 means that the same test output works with both.
1850 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1851 calling malloc() with a zero argument.
1853 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1854 optimization. I've improved things by keeping a bitmap of backreferences with
1855 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1856 fact referenced, the optimization can be applied. It is unlikely that a
1857 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1858 the match to follow \n) will appear inside brackets with a number greater than
1859 31, but if it does, any back reference > 31 suppresses the optimization.
1861 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1862 of disabling numbered capturing parentheses. Any opening parenthesis that is
1863 not followed by ? behaves as if it were followed by ?: but named parentheses
1864 can still be used for capturing (and they will acquire numbers in the usual
1865 way).
1867 41. Redesigned the return codes from the match() function into yes/no/error so
1868 that errors can be passed back from deep inside the nested calls. A malloc
1869 failure while inside a recursive subpattern call now causes the
1870 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1872 42. It is now possible to set a limit on the number of times the match()
1873 function is called in a call to pcre_exec(). This facility makes it possible to
1874 limit the amount of recursion and backtracking, though not in a directly
1875 obvious way, because the match() function is used in a number of different
1876 circumstances. The count starts from zero for each position in the subject
1877 string (for non-anchored patterns). The default limit is, for compatibility, a
1878 large number, namely 10 000 000. You can change this in two ways:
1880 (a) When configuring PCRE before making, you can use --with-match-limit=n
1881 to set a default value for the compiled library.
1883 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1884 a different value is set. See 45 below.
1886 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1888 43. Added a new function pcre_config(int, void *) to enable run-time extraction
1889 of things that can be changed at compile time. The first argument specifies
1890 what is wanted and the second points to where the information is to be placed.
1891 The current list of available information is:
1895 The output is an integer that is set to one if UTF-8 support is available;
1896 otherwise it is set to zero.
1900 The output is an integer that it set to the value of the code that is used for
1901 newline. It is either LF (10) or CR (13).
1905 The output is an integer that contains the number of bytes used for internal
1906 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1910 The output is an integer that contains the threshold above which the POSIX
1911 interface uses malloc() for output vectors. See item 31 above.
1915 The output is an unsigned integer that contains the default limit of the number
1916 of match() calls in a pcre_exec() execution. See 42 above.
1918 44. pcretest has been upgraded by the addition of the -C option. This causes it
1919 to extract all the available output from the new pcre_config() function, and to
1920 output it. The program then exits immediately.
1922 45. A need has arisen to pass over additional data with calls to pcre_exec() in
1923 order to support additional features. One way would have been to define
1924 pcre_exec2() (for example) with extra arguments, but this would not have been
1925 extensible, and would also have required all calls to the original function to
1926 be mapped to the new one. Instead, I have chosen to extend the mechanism that
1927 is used for passing in "extra" data from pcre_study().
1929 The pcre_extra structure is now exposed and defined in pcre.h. It currently
1930 contains the following fields:
1932 flags a bitmap indicating which of the following fields are set
1933 study_data opaque data from pcre_study()
1934 match_limit a way of specifying a limit on match() calls for a specific
1935 call to pcre_exec()
1936 callout_data data for callouts (see 49 below)
1938 The flag bits are also defined in pcre.h, and are
1944 The pcre_study() function now returns one of these new pcre_extra blocks, with
1945 the actual study data pointed to by the study_data field, and the
1946 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1947 before. That is, this change is entirely upwards-compatible and requires no
1948 change to existing code.
1950 If you want to pass in additional data to pcre_exec(), you can either place it
1951 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1952 block.
1954 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1955 data string contains the escape sequence \M, pcretest calls pcre_exec() several
1956 times with different match limits, until it finds the minimum value needed for
1957 pcre_exec() to complete. The value is then output. This can be instructive; for
1958 most simple matches the number is quite small, but for pathological cases it
1959 gets very large very quickly.
1961 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1962 returns the size of the data block pointed to by the study_data field in a
1963 pcre_extra block, that is, the value that was passed as the argument to
1964 pcre_malloc() when PCRE was getting memory in which to place the information
1965 created by pcre_study(). The fourth argument should point to a size_t variable.
1966 pcretest has been extended so that this information is shown after a successful
1967 pcre_study() call when information about the compiled regex is being displayed.
1969 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1970 because what follows is always an absolute path. (Later: it turns out that this
1971 is more than cosmetic for MinGW, because it doesn't like empty path
1972 components.)
1974 49. Some changes have been made to the callout feature (see 28 above):
1976 (i) A callout function now has three choices for what it returns:
1978 0 => success, carry on matching
1979 > 0 => failure at this point, but backtrack if possible
1980 < 0 => serious error, return this value from pcre_exec()
1982 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1983 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1984 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1985 use by callout functions. It will never be used by PCRE itself.
1987 (ii) The pcre_extra structure (see 45 above) has a void * field called
1988 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1989 pcre_callout_block structure has a field of the same name. The contents of
1990 the field passed in the pcre_extra structure are passed to the callout
1991 function in the corresponding field in the callout block. This makes it
1992 easier to use the same callout-containing regex from multiple threads. For
1993 testing, the pcretest program has a new data escape
1995 \C*n pass the number n (may be negative) as callout_data
1997 If the callout function in pcretest receives a non-zero value as
1998 callout_data, it returns that value.
2000 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2001 there were some redundant $(CFLAGS) in commands that are now specified as
2002 $(LINK), which already includes $(CFLAGS).
2004 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2005 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2006 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2007 one-byte characters throughout. Note that case-insensitive matching applies
2008 only to characters whose values are less than 256. PCRE doesn't support the
2009 notion of cases for higher-valued characters.
2011 (i) A character class whose characters are all within 0-255 is handled as
2012 a bit map, and the map is inverted for negative classes. Previously, a
2013 character > 255 always failed to match such a class; however it should
2014 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2016 (ii) A negated character class with a single character < 255 is coded as
2017 "not this character" (OP_NOT). This wasn't working properly when the test
2018 character was multibyte, either singly or repeated.
2020 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2021 mode, for example: \x{100}{2,3}.
2023 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2024 singly or repeated) now correctly test multibyte characters. However,
2025 PCRE doesn't recognize any characters with values greater than 255 as
2026 digits, spaces, or word characters. Such characters always match \D, \S,
2027 and \W, and never match \d, \s, or \w.
2029 (v) Classes may now contain characters and character ranges with values
2030 greater than 255. For example: [ab\x{100}-\x{400}].
2032 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2033 PCRE in UTF-8 mode.
2035 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2036 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2037 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2038 value.)
2040 53. The single man page has become too large. I have therefore split it up into
2041 a number of separate man pages. These also give rise to individual HTML pages;
2042 these are now put in a separate directory, and there is an index.html page that
2043 lists them all. Some hyperlinking between the pages has been installed.
2045 54. Added convenience functions for handling named capturing parentheses.
2047 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2048 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2049 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2050 are faulted.
2052 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2053 calling configure. These values are used when compiling the dftables.c program
2054 which is run to generate the source of the default character tables. They
2055 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2056 you will need to set these values.
2058 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2061 Version 3.9 02-Jan-02
2062 ---------------------
2064 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2066 2. If --disable-static was given, the building process failed when trying to
2067 build pcretest and pcregrep. (For some reason it was using libtool to compile
2068 them, which is not right, as they aren't part of the library.)
2071 Version 3.8 18-Dec-01
2072 ---------------------
2074 1. The experimental UTF-8 code was completely screwed up. It was packing the
2075 bytes in the wrong order. How dumb can you get?
2078 Version 3.7 29-Oct-01
2079 ---------------------
2081 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2082 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2083 this didn't happen under Solaris 8, where I normally test things.
2085 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2086 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2087 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2088 this fix an hour or so after the initial 3.7 release.)
2091 Version 3.6 23-Oct-01
2092 ---------------------
2094 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2095 offsets passed as NULL with zero offset count.
2097 2. The config.guess and config.sub files had not been updated when I moved to
2098 the latest autoconf.
2101 Version 3.5 15-Aug-01
2102 ---------------------
2104 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2105 had been forgotten.
2107 2. By using declared but undefined structures, we can avoid using "void"
2108 definitions in pcre.h while keeping the internal definitions of the structures
2109 private.
2111 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2112 user point of view, this means that both static and shared libraries are built
2113 by default, but this can be individually controlled. More of the work of
2114 handling this static/shared cases is now inside libtool instead of PCRE's make
2115 file.
2117 4. The pcretest utility is now installed along with pcregrep because it is
2118 useful for users (to test regexs) and by doing this, it automatically gets
2119 relinked by libtool. The documentation has been turned into a man page, so
2120 there are now .1, .txt, and .html versions in /doc.
2122 5. Upgrades to pcregrep:
2123 (i) Added long-form option names like gnu grep.
2124 (ii) Added --help to list all options with an explanatory phrase.
2125 (iii) Added -r, --recursive to recurse into sub-directories.
2126 (iv) Added -f, --file to read patterns from a file.
2128 6. pcre_exec() was referring to its "code" argument before testing that
2129 argument for NULL (and giving an error if it was NULL).
2131 7. Upgraded Makefile.in to allow for compiling in a different directory from
2132 the source directory.
2134 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2135 options bits, the pointer it was passed was to an int instead of to an unsigned
2136 long int. This mattered only on 64-bit systems.
2138 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2139 generated) instead of pcre.in, which it its source. Also made the same change
2140 in several of the .c files.
2142 10. A new release of gcc defines printf() as a macro, which broke pcretest
2143 because it had an ifdef in the middle of a string argument for printf(). Fixed
2144 by using separate calls to printf().
2146 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2147 script, to force use of CR or LF instead of \n in the source. On non-Unix
2148 systems, the value can be set in config.h.
2150 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2151 absolute limit. Changed the text of the error message to make this clear, and
2152 likewise updated the man page.
2154 13. The limit of 99 on the number of capturing subpatterns has been removed.
2155 The new limit is 65535, which I hope will not be a "real" limit.
2158 Version 3.4 22-Aug-00
2159 ---------------------
2161 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2163 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2166 Version 3.3 01-Aug-00
2167 ---------------------
2169 1. If an octal character was given, but the value was greater than \377, it
2170 was not getting masked to the least significant bits, as documented. This could
2171 lead to crashes in some systems.
2173 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2174 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2176 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2177 These just pass their arguments on to (pcre_free)(), but they are provided
2178 because some uses of PCRE bind it to non-C systems that can call its functions,
2179 but cannot call free() or pcre_free() directly.
2181 4. Add "make test" as a synonym for "make check". Corrected some comments in
2182 the Makefile.
2184 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2185 Makefile.
2187 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2188 command called pgrep for grepping around the active processes.
2190 7. Added the beginnings of support for UTF-8 character strings.
2192 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2193 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2194 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2195 out for the ar command.)
2198 Version 3.2 12-May-00
2199 ---------------------
2201 This is purely a bug fixing release.
2203 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2204 of ZA. This was just one example of several cases that could provoke this bug,
2205 which was introduced by change 9 of version 2.00. The code for breaking
2206 infinite loops after an iteration that matches an empty string was't working
2207 correctly.
2209 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2210 when matched against abbab (for example). After matching an empty string, it
2211 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2212 caused it to match further down the string than it should.
2214 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2215 was there because it doesn't seem to be needed, and it causes trouble on some
2216 systems, as it is not a Standard C header. It has been removed.
2218 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2219 were reported on the Macintosh. The changes were from
2221 while ((c = *(++ptr)) != 0 && c != '\n');
2222 to
2223 while ((c = *(++ptr)) != 0 && c != '\n') ;
2225 Totally extraordinary, but if that's what it takes...
2227 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2228 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2229 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2230 assumes the way PCRE uses memmove() (always moving upwards).
2232 6. PCRE is being used in one environment where strchr() is not available. There
2233 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2234 faster code anyway.
2237 Version 3.1 09-Feb-00
2238 ---------------------
2240 The only change in this release is the fixing of some bugs in Makefile.in for
2241 the "install" target:
2243 (1) It was failing to install pcreposix.h.
2245 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2248 Version 3.0 01-Feb-00
2249 ---------------------
2251 1. Add support for the /+ modifier to perltest (to output $` like it does in
2252 pcretest).
2254 2. Add support for the /g modifier to perltest.
2256 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2257 matches null strings.
2259 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2260 pattern. Perl treats empty patterns specially - it reuses the most recent
2261 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2262 effect.
2264 5. The POSIX interface was broken in that it was just handing over the POSIX
2265 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2266 required a bigger vector, with some working space on the end. This means that
2267 the POSIX wrapper now has to get and free some memory, and copy the results.
2269 6. Added some simple autoconf support, placing the test data and the
2270 documentation in separate directories, re-organizing some of the
2271 information files, and making it build pcre-config (a GNU standard). Also added
2272 libtool support for building PCRE as a shared library, which is now the
2273 default.
2275 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2276 09 are not valid octal constants. Single digits will be used for minor values
2277 less than 10.
2279 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2280 existing programs that set these in the POSIX interface can use PCRE without
2281 modification.
2283 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2284 return all that pcre_info() returns, plus additional data. The pcre_info()
2285 function is retained for compatibility, but is considered to be obsolete.
2287 10. Added experimental recursion feature (?R) to handle one common case that
2288 Perl 5.6 will be able to do with (?p{...}).
2290 11. Added support for POSIX character classes like [:alpha:], which Perl is
2291 adopting.
2294 Version 2.08 31-Aug-99
2295 ----------------------
2297 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2298 trying to match at the startoffset position, but instead was moving forward to
2299 the next newline as if a previous match had failed.
2301 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2302 and could get into a loop if a null string was matched other than at the start
2303 of the subject.
2305 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2306 be distinguished at compile time, and for completeness also added PCRE_DATE.
2308 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2309 in GnuWin32 environments.
2312 Version 2.07 29-Jul-99
2313 ----------------------
2315 1. The documentation is now supplied in plain text form and HTML as well as in
2316 the form of man page sources.
2318 2. C++ compilers don't like assigning (void *) values to other pointer types.
2319 In particular this affects malloc(). Although there is no problem in Standard
2320 C, I've put in casts to keep C++ compilers happy.
2322 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2323 should be (const char *).
2325 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2326 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2327 However, I haven't made this a standard facility. The documentation doesn't
2328 mention it, and the Makefile doesn't support it.
2330 5. The Makefile now contains an "install" target, with editable destinations at
2331 the top of the file. The pcretest program is not installed.
2333 6. pgrep -V now gives the PCRE version number and date.
2335 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2336 causing the entire string to be ignored, instead of just the last character.
2338 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2339 non-matching string, it can take a very, very long time, even for strings of
2340 quite modest length, because of the nested recursion. PCRE now does better in
2341 some of these cases. It does this by remembering the last required literal
2342 character in the pattern, and pre-searching the subject to ensure it is present
2343 before running the real match. In other words, it applies a heuristic to detect
2344 some types of certain failure quickly, and in the above example, if presented
2345 with a string that has no trailing " it gives "no match" very quickly.
2347 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2348 other alternatives are tried instead.
2351 Version 2.06 09-Jun-99
2352 ----------------------
2354 1. Change pcretest's output for amount of store used to show just the code
2355 space, because the remainder (the data block) varies in size between 32-bit and
2356 64-bit systems.
2358 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2359 start matching at. This allows lookbehinds to work when searching for multiple
2360 occurrences in a string.
2362 3. Added additional options to pcretest for testing multiple occurrences:
2364 /+ outputs the rest of the string that follows a match
2365 /g loops for multiple occurrences, using the new startoffset argument
2366 /G loops for multiple occurrences by passing an incremented pointer
2368 4. PCRE wasn't doing the "first character" optimization for patterns starting
2369 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2370 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2371 the letter 'x'. On long subject strings, this gives a significant speed-up.
2374 Version 2.05 21-Apr-99
2375 ----------------------
2377 1. Changed the type of magic_number from int to long int so that it works
2378 properly on 16-bit systems.
2380 2. Fixed a bug which caused patterns starting with .* not to work correctly
2381 when the subject string contained newline characters. PCRE was assuming
2382 anchoring for such patterns in all cases, which is not correct because .* will
2383 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2384 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2385 must be retried after every newline in the subject.
2388 Version 2.04 18-Feb-99
2389 ----------------------
2391 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2392 computation of the store needed to hold the pattern was incorrect (too large).
2393 If such patterns were nested a few deep, this could multiply and become a real
2394 problem.
2396 2. Added /M option to pcretest to show the memory requirement of a specific
2397 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2399 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2400 compiled in such a way that the backtracking after subsequent failure was
2401 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2402 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2405 Version 2.03 02-Feb-99
2406 ----------------------
2408 1. Fixed typo and small mistake in man page.
2410 2. Added 4th condition (GPL supersedes if conflict) and created separate
2411 LICENCE file containing the conditions.
2413 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2414 Perl, that is the internal \ allows the delimiter to be included in the
2415 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2416 the final delimiter, add \ to the end of the pattern (to test the error).
2418 4. Added the convenience functions for extracting substrings after a successful
2419 match. Updated pcretest to make it able to test these functions.
2422 Version 2.02 14-Jan-99
2423 ----------------------
2425 1. Initialized the working variables associated with each extraction so that
2426 their saving and restoring doesn't refer to uninitialized store.
2428 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2429 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2430 fix the problem.
2432 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2433 calls, and wasn't printing the correct value for compiling calls. Increased the
2434 default value of LOOPREPEAT, and the number of significant figures in the
2435 times.
2437 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2439 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2440 a building problem on Windows NT with a FAT file system.
2443 Version 2.01 21-Oct-98
2444 ----------------------
2446 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2447 to character tables built by pcre_maketables() in the current locale. If NULL
2448 is passed, the default tables are used.
2451 Version 2.00 24-Sep-98
2452 ----------------------
2454 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2455 it any more.
2457 2. Allow quantification of (?>) groups, and make it work correctly.
2459 3. The first character computation wasn't working for (?>) groups.
2461 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2462 end of the subject) and add 5.005's \z, which really does match only at the
2463 very end of the subject.
2465 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2467 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2468 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2469 localized options. All options to pcre_study() were also removed.
2471 7. Add other new features from 5.005:
2473 $(?<= positive lookbehind
2474 $(?<! negative lookbehind
2475 (?imsx-imsx) added the unsetting capability
2476 such a setting is global if at outer level; local otherwise
2477 (?imsx-imsx:) non-capturing groups with option setting
2478 (?(cond)re|re) conditional pattern matching
2480 A backreference to itself in a repeated group matches the previous
2481 captured string.
2483 8. General tidying up of studying (both automatic and via "study")
2484 consequential on the addition of new assertions.
2486 9. As in 5.005, unlimited repeated groups that could match an empty substring
2487 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2488 runtime if any iteration does actually match an empty substring.
2490 10. Include the RunTest script in the distribution.
2492 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2493 discrepancies, some of which were old and were also with respect to 5.004. They
2494 have now been fixed.
2497 Version 1.09 28-Apr-98
2498 ----------------------
2500 1. A negated single character class followed by a quantifier with a minimum
2501 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2502 program crashes, or just wrong answers. This did not apply to negated classes
2503 containing more than one character, or to minima other than one.
2506 Version 1.08 27-Mar-98
2507 ----------------------
2509 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2511 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2512 latter must appear before anything that relies on it in the pattern.
2515 Version 1.07 16-Feb-98
2516 ----------------------
2518 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2519 repeat of a potentially empty string).
2522 Version 1.06 23-Jan-98
2523 ----------------------
2525 1. Added Markus Oberhumer's little patches for C++.
2527 2. Literal strings longer than 255 characters were broken.
2530 Version 1.05 23-Dec-97
2531 ----------------------
2533 1. Negated character classes containing more than one character were failing if
2534 PCRE_CASELESS was set at run time.
2537 Version 1.04 19-Dec-97
2538 ----------------------
2540 1. Corrected the man page, where some "const" qualifiers had been omitted.
2542 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2543 input syntax.
2545 3. Fixed memory leak which occurred when a regex with back references was
2546 matched with an offsets vector that wasn't big enough. The temporary memory
2547 that is used in this case wasn't being freed if the match failed.
2549 4. Tidied pcretest to ensure it frees memory that it gets.
2551 5. Temporary memory was being obtained in the case where the passed offsets
2552 vector was exactly big enough.
2554 6. Corrected definition of offsetof() from change 5 below.
2556 7. I had screwed up change 6 below and broken the rules for the use of
2557 setjmp(). Now fixed.
2560 Version 1.03 18-Dec-97
2561 ----------------------
2563 1. A erroneous regex with a missing opening parenthesis was correctly
2564 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2565 on some systems.
2567 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2568 it was reported that one broken compiler failed on the former because "code" is
2569 also an independent variable.
2571 3. The erroneous regex a[]b caused an array overrun reference.
2573 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2574 fail on data ending with that character. (It was going on too far, and checking
2575 the next character, typically a binary zero.) This was specific to the
2576 optimized code for single-character negative classes.
2578 5. Added a contributed patch from the TIN world which does the following:
2580 + Add an undef for memmove, in case the the system defines a macro for it.
2582 + Add a definition of offsetof(), in case there isn't one. (I don't know
2583 the reason behind this - offsetof() is part of the ANSI standard - but
2584 it does no harm).
2586 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2587 most of the places where whitespace preceded '#'. I have given up and
2588 allowed the remaining 2 cases to be at the margin.
2590 + Rename some variables in pcre to eliminate shadowing. This seems very
2591 pedantic, but does no harm, of course.
2593 6. Moved the call to setjmp() into its own function, to get rid of warnings
2594 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2596 7. Constructs such as \d{8,} were compiling into the equivalent of
2597 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2598 outcome, but in this particular case used more store than had been allocated,
2599 which caused the bug to be discovered because it threw up an internal error.
2601 8. The debugging code in both pcre and pcretest for outputting the compiled
2602 form of a regex was going wrong in the case of back references followed by
2603 curly-bracketed repeats.
2606 Version 1.02 12-Dec-97
2607 ----------------------
2609 1. Typos in pcre.3 and comments in the source fixed.
2611 2. Applied a contributed patch to get rid of places where it used to remove
2612 'const' from variables, and fixed some signed/unsigned and uninitialized
2613 variable warnings.
2615 3. Added the "runtest" target to Makefile.
2617 4. Set default compiler flag to -O2 rather than just -O.
2620 Version 1.01 19-Nov-97
2621 ----------------------
2623 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2624 like /([ab]*)*/, that is, for classes with more than one character in them.
2626 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2627 as /((?>a*))*/ (a PCRE_EXTRA facility).
2630 Version 1.00 18-Nov-97
2631 ----------------------
2633 1. Added compile-time macros to support systems such as SunOS4 which don't have
2634 memmove() or strerror() but have other things that can be used instead.
2636 2. Arranged that "make clean" removes the executables.
2639 Version 0.99 27-Oct-97
2640 ----------------------
2642 1. Fixed bug in code for optimizing classes with only one character. It was
2643 initializing a 32-byte map regardless, which could cause it to run off the end
2644 of the memory it had got.
2646 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2649 Version 0.98 22-Oct-97
2650 ----------------------
2652 1. Fixed bug in code for handling temporary memory usage when there are more
2653 back references than supplied space in the ovector. This could cause segfaults.
2656 Version 0.97 21-Oct-97
2657 ----------------------
2659 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2661 2. Optimized negated single characters not to use a bit map.
2663 3. Brought error texts together as macro definitions; clarified some of them;
2664 fixed one that was wrong - it said "range out of order" when it meant "invalid
2665 escape sequence".
2667 4. Changed some char * arguments to const char *.
2669 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2671 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2672 pcretest.
2675 Version 0.96 16-Oct-97
2676 ----------------------
2678 1. Added a simple "pgrep" utility to the distribution.
2680 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2681 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2682 where "ddd" means "one or more decimal digits".
2684 3. Fixed serious bug. If a pattern had a back reference, but the call to
2685 pcre_exec() didn't supply a large enough ovector to record the related
2686 identifying subpattern, the match always failed. PCRE now remembers the number
2687 of the largest back reference, and gets some temporary memory in which to save
2688 the offsets during matching if necessary, in order to ensure that
2689 backreferences always work.
2691 4. Increased the compatibility with Perl in a number of ways:
2693 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2694 to request this handling. The option can be set at compile or exec time.
2696 (b) $ matches before a terminating newline by default; an option
2697 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2698 mode). The option can be set at compile or exec time.
2700 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2701 the same as Perl's. If the decimal number it represents is less than 10
2702 or there aren't that many previous left capturing parentheses, an octal
2703 escape is read. Inside a character class, it's always an octal escape,
2704 even if it is a single digit.
2706 (d) An escaped but undefined alphabetic character is taken as a literal,
2707 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2708 escapes.
2710 (e) {0} is now permitted. (The previous item is removed from the compiled
2711 pattern).
2713 5. Changed all the names of code files so that the basic parts are no longer
2714 than 10 characters, and abolished the teeny "globals.c" file.
2716 6. Changed the handling of character classes; they are now done with a 32-byte
2717 bit map always.
2719 7. Added the -d and /D options to pcretest to make it possible to look at the
2720 internals of compilation without having to recompile pcre.
2723 Version 0.95 23-Sep-97
2724 ----------------------
2726 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2727 \x20 at the start of a run of normal characters. These were being treated as
2728 real characters, instead of the source characters being re-checked.
2731 Version 0.94 18-Sep-97
2732 ----------------------
2734 1. The functions are now thread-safe, with the caveat that the global variables
2735 containing pointers to malloc() and free() or alternative functions are the
2736 same for all threads.
2738 2. Get pcre_study() to generate a bitmap of initial characters for non-
2739 anchored patterns when this is possible, and use it if passed to pcre_exec().
2742 Version 0.93 15-Sep-97
2743 ----------------------
2745 1. /(b)|(:+)/ was computing an incorrect first character.
2747 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2748 but not actually doing anything yet.
2750 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2751 as Perl does (e.g. [-az] or [az-]).
2753 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2754 all possible positions.
2756 5. Split up into different modules to avoid including unneeded functions in a
2757 compiled binary. However, compile and exec are still in one module. The "study"
2758 function is split off.
2760 6. The character tables are now in a separate module whose source is generated
2761 by an auxiliary program - but can then be edited by hand if required. There are
2762 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2763 toupper() in the code.
2765 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2766 make them global. Abolish the function for setting them, as the caller can now
2767 set them directly.
2770 Version 0.92 11-Sep-97
2771 ----------------------
2773 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2774 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2776 2. Caseless matching was not working in character classes if the characters in
2777 the pattern were in upper case.
2779 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2781 4. Make PCRE_ANCHORED public and accept as a compile option.
2783 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2784 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2785 pass them.
2787 6. Give an error if bad option bits passed at compile or run time.
2789 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2790 pcretest to cause it to pass that flag.
2792 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2793 options, and the first character, if set.
2795 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2798 Version 0.91 10-Sep-97
2799 ----------------------
2801 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2802 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2804 2. PCRE was looping on encountering an indefinitely repeated back reference to
2805 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2806 Perl does - treats the match as successful.
2808 ****


Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

  ViewVC Help
Powered by ViewVC 1.1.5