ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 175 - (show annotations)
Mon Jun 11 13:38:38 2007 UTC (14 years, 4 months ago) by ph10
File size: 129372 byte(s)
Added the Perl 5.10 (?| "branch reset" feature.
1 ChangeLog for PCRE
2 ------------------
4 Version 7.2 05-June-07
5 ---------------------
7 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
8 which is apparently normally available under Windows.
10 2. Re-jig the pcregrep tests with different newline settings in an attempt
11 to make them independent of the local environment's newline setting.
13 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
15 4. Some of the "internals" tests were previously cut out when the link size
16 was not 2, because the output contained actual offsets. The recent new
17 "Z" feature of pcretest means that these can be cut out, making the tests
18 usable with all link sizes.
20 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
21 stack recursion. This gives a massive performance boost under BSD, but just
22 a small improvement under Linux. However, it saves one field in the frame
23 in all cases.
25 6. Added more features from the forthcoming Perl 5.10:
27 (a) (?-n) (where n is a string of digits) is a relative subroutine or
28 recursion call. It refers to the nth most recently opened parentheses.
30 (b) (?+n) is also a relative subroutine call; it refers to the nth next
31 to be opened parentheses.
33 (c) Conditions that refer to capturing parentheses can be specified
34 relatively, for example, (?(-2)... or (?(+3)...
36 (d) \K resets the start of the current match so that everything before
37 is not part of it.
39 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
41 (f) \g{name} is another synonym - part of Perl 5.10's unification of
42 reference syntax.
44 (g) (?| introduces a group in which the numbering of parentheses in each
45 alternative starts with the same number.
47 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
50 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
51 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
52 for detecting groups that can match an empty string.
54 9. A pattern with a very large number of alternatives (more than several
55 hundred) was running out of internal workspace during the pre-compile
56 phase, where pcre_compile() figures out how much memory will be needed. A
57 bit of new cunning has reduced the workspace needed for groups with
58 alternatives. The 1000-alternative test pattern now uses 12 bytes of
59 workspace instead of running out of the 4096 that are available.
62 Version 7.1 24-Apr-07
63 ---------------------
65 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
66 that is more "standard", making use of automake and other Autotools. There
67 is some re-arrangement of the files and adjustment of comments consequent
68 on this.
70 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
71 for recursive directory scanning broke on some systems because the files
72 are not scanned in any specific order and on different systems the order
73 was different. A call to "sort" has been inserted into RunGrepTest for the
74 approprate test as a short-term fix. In the longer term there may be an
75 alternative.
77 3. I had an email from Eric Raymond about problems translating some of PCRE's
78 man pages to HTML (despite the fact that I distribute HTML pages, some
79 people do their own conversions for various reasons). The problems
80 concerned the use of low-level troff macros .br and .in. I have therefore
81 removed all such uses from the man pages (some were redundant, some could
82 be replaced by .nf/.fi pairs). The 132html script that I use to generate
83 HTML has been updated to handle .nf/.fi and to complain if it encounters
84 .br or .in.
86 4. Updated comments in configure.ac that get placed in config.h.in and also
87 arranged for config.h to be included in the distribution, with the name
88 config.h.generic, for the benefit of those who have to compile without
89 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
91 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
92 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
93 makevp.bat for the current PCRE, using the additional files
94 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
96 6. A Windows user reported a minor discrepancy with test 2, which turned out
97 to be caused by a trailing space on an input line that had got lost in his
98 copy. The trailing space was an accident, so I've just removed it.
100 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
101 that is needed.
103 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
104 as "const" (a) because they are and (b) because it helps the PHP
105 maintainers who have recently made a script to detect big data structures
106 in the php code that should be moved to the .rodata section. I remembered
107 to update Builducptable as well, so it won't revert if ucptable.h is ever
108 re-created.
110 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
111 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
112 order to be able to cut out the UTF-8 tables in the latter when UTF-8
113 support is not required. This saves 1.5-2K of code, which is important in
114 some applications.
116 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
117 so as not to refer to the tables, even though these functions will never be
118 called when UTF-8 support is disabled. Otherwise there are problems with a
119 shared library.
121 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
123 (a) It was defining its arguments as char * instead of void *.
125 (b) It was assuming that all moves were upwards in memory; this was true
126 a long time ago when I wrote it, but is no longer the case.
128 The emulated memove() is provided for those environments that have neither
129 memmove() nor bcopy(). I didn't think anyone used it these days, but that
130 is clearly not the case, as these two bugs were recently reported.
132 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
133 and Detrail to create the HTML documentation, the .txt form of the man
134 pages, and it removes trailing spaces from listed files. It also creates
135 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
136 case, it wraps all the #defines with #ifndefs. This script should be run
137 before "make dist".
139 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
140 with Unicode property support.
142 (a) For a maximizing quantifier, if the two different cases of the
143 character were of different lengths in their UTF-8 codings (there are
144 some cases like this - I found 11), and the matching function had to
145 back up over a mixture of the two cases, it incorrectly assumed they
146 were both the same length.
148 (b) When PCRE was configured to use the heap rather than the stack for
149 recursion during matching, it was not correctly preserving the data for
150 the other case of a UTF-8 character when checking ahead for a match
151 while processing a minimizing repeat. If the check also involved
152 matching a wide character, but failed, corruption could cause an
153 erroneous result when trying to check for a repeat of the original
154 character.
156 13. Some tidying changes to the testing mechanism:
158 (a) The RunTest script now detects the internal link size and whether there
159 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
160 values substituted by "configure". (The RunGrepTest script already did
161 this for UTF-8.) The configure.ac script no longer substitutes the
162 relevant variables.
164 (b) The debugging options /B and /D in pcretest show the compiled bytecode
165 with length and offset values. This means that the output is different
166 for different internal link sizes. Test 2 is skipped for link sizes
167 other than 2 because of this, bypassing the problem. Unfortunately,
168 there was also a test in test 3 (the locale tests) that used /B and
169 failed for link sizes other than 2. Rather than cut the whole test out,
170 I have added a new /Z option to pcretest that replaces the length and
171 offset values with spaces. This is now used to make test 3 independent
172 of link size. (Test 2 will be tidied up later.)
174 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
175 segmentation fault instead of returning the appropriate error message.
177 15. In multiline mode when the newline sequence was set to "any", the pattern
178 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
179 This doesn't seem right; it now treats the CRLF combination as the line
180 ending, and so does not match in that case. It's only a pattern such as ^$
181 that would hit this one: something like ^ABC$ would have failed after \r
182 and then tried again after \r\n.
184 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
185 in an attempt to make files that differ only in their line terminators
186 compare equal. This works on Linux.
188 17. Under certain error circumstances pcregrep might try to free random memory
189 as it exited. This is now fixed, thanks to valgrind.
191 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
192 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
193 was because its rules for how to advance for /g after matching an empty
194 string at the end of a line did not allow for this case. They now check for
195 it specially.
197 20. pcretest is supposed to handle patterns and data of any length, by
198 extending its buffers when necessary. It was getting this wrong when the
199 buffer for a data line had to be extended.
201 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
202 CRLF as a newline sequence.
204 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
205 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
206 I have nevertheless tidied it up.
208 23. Added some casts to kill warnings from HP-UX ia64 compiler.
210 24. Added a man page for pcre-config.
213 Version 7.0 19-Dec-06
214 ---------------------
216 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
217 moving to gcc 4.1.1.
219 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
220 sys/time.h, which is documented as needed for this function. It doesn't
221 seem to matter on Linux, but it showed up on some releases of OS X.
223 3. It seems that there are systems where bytes whose values are greater than
224 127 match isprint() in the "C" locale. The "C" locale should be the
225 default when a C program starts up. In most systems, only ASCII printing
226 characters match isprint(). This difference caused the output from pcretest
227 to vary, making some of the tests fail. I have changed pcretest so that:
229 (a) When it is outputting text in the compiled version of a pattern, bytes
230 other than 32-126 are always shown as hex escapes.
232 (b) When it is outputting text that is a matched part of a subject string,
233 it does the same, unless a different locale has been set for the match
234 (using the /L modifier). In this case, it uses isprint() to decide.
236 4. Fixed a major bug that caused incorrect computation of the amount of memory
237 required for a compiled pattern when options that changed within the
238 pattern affected the logic of the preliminary scan that determines the
239 length. The relevant options are -x, and -i in UTF-8 mode. The result was
240 that the computed length was too small. The symptoms of this bug were
241 either the PCRE error "internal error: code overflow" from pcre_compile(),
242 or a glibc crash with a message such as "pcretest: free(): invalid next
243 size (fast)". Examples of patterns that provoked this bug (shown in
244 pcretest format) are:
246 /(?-x: )/x
247 /(?x)(?-x: \s*#\s*)/
248 /((?i)[\x{c0}])/8
249 /(?i:[\x{c0}])/8
251 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
252 is now done differently.
254 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
255 wrapper classes; (b) implement a new function in the C++ scanner that is
256 more efficient than the old way of doing things because it avoids levels of
257 recursion in the regex matching; (c) add a paragraph to the documentation
258 for the FullMatch() function.
260 6. The escape sequence \n was being treated as whatever was defined as
261 "newline". Not only was this contrary to the documentation, which states
262 that \n is character 10 (hex 0A), but it also went horribly wrong when
263 "newline" was defined as CRLF. This has been fixed.
265 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
266 was being set to -1 for the "end of line" case (supposedly a value that no
267 character can have). Though this value is never used (the check for end of
268 line is "zero bytes in current character"), it caused compiler complaints.
269 I've changed it to 0xffffffff.
271 8. In pcre_version.c, the version string was being built by a sequence of
272 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
273 string (as it is for production releases) called a macro with an empty
274 argument. The C standard says the result of this is undefined. The gcc
275 compiler treats it as an empty string (which was what was wanted) but it is
276 reported that Visual C gives an error. The source has been hacked around to
277 avoid this problem.
279 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
280 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
281 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
282 of them did).
284 10. Originally, pcretest opened its input and output without "b"; then I was
285 told that "b" was needed in some environments, so it was added for release
286 5.0 to both the input and output. (It makes no difference on Unix-like
287 systems.) Later I was told that it is wrong for the input on Windows. I've
288 now abstracted the modes into two macros, to make it easier to fiddle with
289 them, and removed "b" from the input mode under Windows.
291 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
293 12. Added -help and --help to pcretest as an official way of being reminded
294 of the options.
296 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
297 and pcrecpp.cc because they annoy compilers at high warning levels.
299 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
301 15. Fixed an occurrence of == in configure.ac that should have been = (shell
302 scripts are not C programs :-) and which was not noticed because it works
303 on Linux.
305 16. pcretest is supposed to handle any length of pattern and data line (as one
306 line or as a continued sequence of lines) by extending its input buffer if
307 necessary. This feature was broken for very long pattern lines, leading to
308 a string of junk being passed to pcre_compile() if the pattern was longer
309 than about 50K.
311 17. I have done a major re-factoring of the way pcre_compile() computes the
312 amount of memory needed for a compiled pattern. Previously, there was code
313 that made a preliminary scan of the pattern in order to do this. That was
314 OK when PCRE was new, but as the facilities have expanded, it has become
315 harder and harder to keep it in step with the real compile phase, and there
316 have been a number of bugs (see for example, 4 above). I have now found a
317 cunning way of running the real compile function in a "fake" mode that
318 enables it to compute how much memory it would need, while actually only
319 ever using a few hundred bytes of working memory and without too many
320 tests of the mode. This should make future maintenance and development
321 easier. A side effect of this work is that the limit of 200 on the nesting
322 depth of parentheses has been removed (though this was never a serious
323 limitation, I suspect). However, there is a downside: pcre_compile() now
324 runs more slowly than before (30% or more, depending on the pattern). I
325 hope this isn't a big issue. There is no effect on runtime performance.
327 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
328 newline (only possible for the last line of a file) and it was a
329 pattern that set a locale (followed by /Lsomething), pcretest crashed.
331 19. Added additional timing features to pcretest. (1) The -tm option now times
332 matching only, not compiling. (2) Both -t and -tm can be followed, as a
333 separate command line item, by a number that specifies the number of
334 repeats to use when timing. The default is 50000; this gives better
335 precision, but takes uncomfortably long for very large patterns.
337 20. Extended pcre_study() to be more clever in cases where a branch of a
338 subpattern has no definite first character. For example, (a*|b*)[cd] would
339 previously give no result from pcre_study(). Now it recognizes that the
340 first character must be a, b, c, or d.
342 21. There was an incorrect error "recursive call could loop indefinitely" if
343 a subpattern (or the entire pattern) that was being tested for matching an
344 empty string contained only one non-empty item after a nested subpattern.
345 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
346 incorrectly, because the \d was being skipped in the check.
348 22. The pcretest program now has a new pattern option /B and a command line
349 option -b, which is equivalent to adding /B to every pattern. This causes
350 it to show the compiled bytecode, without the additional information that
351 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
352 is the same as /B/I).
354 23. A new optimization is now able automatically to treat some sequences such
355 as a*b as a*+b. More specifically, if something simple (such as a character
356 or a simple class like \d) has an unlimited quantifier, and is followed by
357 something that cannot possibly match the quantified thing, the quantifier
358 is automatically "possessified".
360 24. A recursive reference to a subpattern whose number was greater than 39
361 went wrong under certain circumstances in UTF-8 mode. This bug could also
362 have affected the operation of pcre_study().
364 25. Realized that a little bit of performance could be had by replacing
365 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
367 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
369 27. Possessive quantifiers such as a++ were previously implemented by turning
370 them into atomic groups such as ($>a+). Now they have their own opcodes,
371 which improves performance. This includes the automatically created ones
372 from 23 above.
374 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
375 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
376 the first matched character to be a colon. This applied both to named and
377 numbered groups.
379 29. The ucpinternal.h header file was missing its idempotency #ifdef.
381 30. I was sent a "project" file called libpcre.a.dev which I understand makes
382 building PCRE on Windows easier, so I have included it in the distribution.
384 31. There is now a check in pcretest against a ridiculously large number being
385 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
386 loop, the loop is abandoned.
388 32. Forward references to subpatterns in conditions such as (?(2)...) where
389 subpattern 2 is defined later cause pcre_compile() to search forwards in
390 the pattern for the relevant set of parentheses. This search went wrong
391 when there were unescaped parentheses in a character class, parentheses
392 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
394 33. "Subroutine" calls and backreferences were previously restricted to
395 referencing subpatterns earlier in the regex. This restriction has now
396 been removed.
398 34. Added a number of extra features that are going to be in Perl 5.10. On the
399 whole, these are just syntactic alternatives for features that PCRE had
400 previously implemented using the Python syntax or my own invention. The
401 other formats are all retained for compatibility.
403 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
404 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
405 also .NET compatible.
407 (b) A recursion or subroutine call to a named group can now be defined as
408 (?&name) as well as (?P>name).
410 (c) A backreference to a named group can now be defined as \k<name> or
411 \k'name' as well as (?P=name). The new forms, as well as being in Perl
412 5.10, are also .NET compatible.
414 (d) A conditional reference to a named group can now use the syntax
415 (?(<name>) or (?('name') as well as (?(name).
417 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
418 groups (named and numbered) that are never evaluated inline, but can be
419 called as "subroutines" from elsewhere. In effect, the DEFINE condition
420 is always false. There may be only one alternative in such a group.
422 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
423 as the simple (?(R). The condition is true only if the most recent
424 recursion is that of the given number or name. It does not search out
425 through the entire recursion stack.
427 (g) The escape \gN or \g{N} has been added, where N is a positive or
428 negative number, specifying an absolute or relative reference.
430 35. Tidied to get rid of some further signed/unsigned compiler warnings and
431 some "unreachable code" warnings.
433 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
434 things, this adds five new scripts.
436 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
437 There were also incompatibilities regarding the handling of \Q..\E inside
438 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
439 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
441 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
442 matches an empty string, and forcibly breaks the loop. There were bugs in
443 this code in non-simple cases. For a pattern such as ^(a()*)* matched
444 against aaaa the result was just "a" rather than "aaaa", for example. Two
445 separate and independent bugs (that affected different cases) have been
446 fixed.
448 39. Refactored the code to abolish the use of different opcodes for small
449 capturing bracket numbers. This is a tidy that I avoided doing when I
450 removed the limit on the number of capturing brackets for 3.5 back in 2001.
451 The new approach is not only tidier, it makes it possible to reduce the
452 memory needed to fix the previous bug (38).
454 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
455 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
456 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
457 mode.
459 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
460 report.
462 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
463 copy construction and assignment in the C++ wrapper.
465 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
466 couple of bugs that could have given wrong results in the "--newline=crlf"
467 case.
469 44. Added a number of casts and did some reorganization of signed/unsigned int
470 variables following suggestions from Dair Grant. Also renamed the variable
471 "this" as "item" because it is a C++ keyword.
473 45. Arranged for dftables to add
475 #include "pcre_internal.h"
477 to pcre_chartables.c because without it, gcc 4.x may remove the array
478 definition from the final binary if PCRE is built into a static library and
479 dead code stripping is activated.
481 46. For an unanchored pattern, if a match attempt fails at the start of a
482 newline sequence, and the newline setting is CRLF or ANY, and the next two
483 characters are CRLF, advance by two characters instead of one.
486 Version 6.7 04-Jul-06
487 ---------------------
489 1. In order to handle tests when input lines are enormously long, pcretest has
490 been re-factored so that it automatically extends its buffers when
491 necessary. The code is crude, but this _is_ just a test program. The
492 default size has been increased from 32K to 50K.
494 2. The code in pcre_study() was using the value of the re argument before
495 testing it for NULL. (Of course, in any sensible call of the function, it
496 won't be NULL.)
498 3. The memmove() emulation function in pcre_internal.h, which is used on
499 systems that lack both memmove() and bcopy() - that is, hardly ever -
500 was missing a "static" storage class specifier.
502 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
503 containing an extended class (one that cannot be represented by a bitmap
504 because it contains high-valued characters or Unicode property items, e.g.
505 [\pZ]). Almost always one would set UTF-8 mode when processing such a
506 pattern, but PCRE should not loop if you do not (it no longer does).
507 [Detail: two cases were found: (a) a repeated subpattern containing an
508 extended class; (b) a recursive reference to a subpattern that followed a
509 previous extended class. It wasn't skipping over the extended class
510 correctly when UTF-8 mode was not set.]
512 5. A negated single-character class was not being recognized as fixed-length
513 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
514 compile error "lookbehind assertion is not fixed length".
516 6. The RunPerlTest auxiliary script was showing an unexpected difference
517 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
518 write a Perl script that can interpret lines of an input file either as
519 byte characters or as UTF-8, which is what "perltest" was being required to
520 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
521 can't do is switch easily at run time between having the "use utf8;" pragma
522 or not. In the end, I fudged it by using the RunPerlTest script to insert
523 "use utf8;" explicitly for the UTF-8 tests.
525 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
526 the end of the subject string, contrary to the documentation and to what
527 Perl does. This was true of both matching functions. Now it matches only at
528 the start of the subject and immediately after *internal* newlines.
530 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
531 a pointer to an int instead of a pointer to an unsigned long int. This
532 caused problems on 64-bit systems.
534 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
535 instance of the 'standard' template library not being so standard".
537 10. There was no check on the number of named subpatterns nor the maximum
538 length of a subpattern name. The product of these values is used to compute
539 the size of the memory block for a compiled pattern. By supplying a very
540 long subpattern name and a large number of named subpatterns, the size
541 computation could be caused to overflow. This is now prevented by limiting
542 the length of names to 32 characters, and the number of named subpatterns
543 to 10,000.
545 11. Subpatterns that are repeated with specific counts have to be replicated in
546 the compiled pattern. The size of memory for this was computed from the
547 length of the subpattern and the repeat count. The latter is limited to
548 65535, but there was no limit on the former, meaning that integer overflow
549 could in principle occur. The compiled length of a repeated subpattern is
550 now limited to 30,000 bytes in order to prevent this.
552 12. Added the optional facility to have named substrings with the same name.
554 13. Added the ability to use a named substring as a condition, using the
555 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
556 are numbers (not recommended). Forward references are permitted.
558 14. Added forward references in named backreferences (if you see what I mean).
560 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
561 pattern could run off the end of the subject. For example, the pattern
562 "(?s)(.{1,5})"8 did this with the subject "ab".
564 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
565 PCRE_CASELESS was set when matching characters that were quantified with ?
566 or *.
568 17. A character class other than a single negated character that had a minimum
569 but no maximum quantifier - for example [ab]{6,} - was not handled
570 correctly by pce_dfa_exec(). It would match only one character.
572 18. A valid (though odd) pattern that looked like a POSIX character
573 class but used an invalid character after [ (for example [[,abc,]]) caused
574 pcre_compile() to give the error "Failed: internal error: code overflow" or
575 in some cases to crash with a glibc free() error. This could even happen if
576 the pattern terminated after [[ but there just happened to be a sequence of
577 letters, a binary zero, and a closing ] in the memory that followed.
579 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
580 over the years. Originally (before any Unicode support), just the bottom 8
581 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
582 output from "man perlunicode" includes this:
584 The regular expression compiler produces polymorphic opcodes. That
585 is, the pattern adapts to the data and automatically switches to
586 the Unicode character scheme when presented with Unicode data--or
587 instead uses a traditional byte scheme when presented with byte
588 data.
590 Sadly, a wide octal escape does not cause a switch, and in a string with
591 no other multibyte characters, these octal escapes are treated as before.
592 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
593 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
594 Unicode string.
596 I have not perpetrated such confusion in PCRE. Up till now, it took just
597 the bottom 8 bits, as in old Perl. I have now made octal escapes with
598 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
599 translate to the appropriate multibyte character.
601 29. Applied some refactoring to reduce the number of warnings from Microsoft
602 and Borland compilers. This has included removing the fudge introduced
603 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
604 a warning about an unused variable.
606 21. PCRE has not included VT (character 0x0b) in the set of whitespace
607 characters since release 4.0, because Perl (from release 5.004) does not.
608 [Or at least, is documented not to: some releases seem to be in conflict
609 with the documentation.] However, when a pattern was studied with
610 pcre_study() and all its branches started with \s, PCRE still included VT
611 as a possible starting character. Of course, this did no harm; it just
612 caused an unnecessary match attempt.
614 22. Removed a now-redundant internal flag bit that recorded the fact that case
615 dependency changed within the pattern. This was once needed for "required
616 byte" processing, but is no longer used. This recovers a now-scarce options
617 bit. Also moved the least significant internal flag bit to the most-
618 significant bit of the word, which was not previously used (hangover from
619 the days when it was an int rather than a uint) to free up another bit for
620 the future.
622 23. Added support for CRLF line endings as well as CR and LF. As well as the
623 default being selectable at build time, it can now be changed at runtime
624 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
625 specify that it is scanning data with non-default line endings.
627 24. Changed the definition of CXXLINK to make it agree with the definition of
628 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
630 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
631 recursions. This makes a big different to stack usage for some patterns.
633 26. If a subpattern containing a named recursion or subroutine reference such
634 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
635 the space required for the compiled pattern went wrong and gave too small a
636 value. Depending on the environment, this could lead to "Failed: internal
637 error: code overflow at offset 49" or "glibc detected double free or
638 corruption" errors.
640 27. Applied patches from Google (a) to support the new newline modes and (b) to
641 advance over multibyte UTF-8 characters in GlobalReplace.
643 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
644 difference for some implementation of PCRE in some Windows version.
646 29. Added some extra testing facilities to pcretest:
648 \q<number> in a data line sets the "match limit" value
649 \Q<number> in a data line sets the "match recursion limt" value
650 -S <number> sets the stack size, where <number> is in megabytes
652 The -S option isn't available for Windows.
655 Version 6.6 06-Feb-06
656 ---------------------
658 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
659 in pcreposix.h. I have copied the definition from pcre.h.
661 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
662 because pcre.h is no longer a built file.
664 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
665 not normally included in the compiled code.
668 Version 6.5 01-Feb-06
669 ---------------------
671 1. When using the partial match feature with pcre_dfa_exec(), it was not
672 anchoring the second and subsequent partial matches at the new starting
673 point. This could lead to incorrect results. For example, with the pattern
674 /1234/, partially matching against "123" and then "a4" gave a match.
676 2. Changes to pcregrep:
678 (a) All non-match returns from pcre_exec() were being treated as failures
679 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
680 error message is output. Some extra information is given for the
682 probably the only errors that are likely to be caused by users (by
683 specifying a regex that has nested indefinite repeats, for instance).
684 If there are more than 20 of these errors, pcregrep is abandoned.
686 (b) A binary zero was treated as data while matching, but terminated the
687 output line if it was written out. This has been fixed: binary zeroes
688 are now no different to any other data bytes.
690 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
691 used to set a locale for matching. The --locale=xxxx long option has
692 been added (no short equivalent) to specify a locale explicitly on the
693 pcregrep command, overriding the environment variables.
695 (d) When -B was used with -n, some line numbers in the output were one less
696 than they should have been.
698 (e) Added the -o (--only-matching) option.
700 (f) If -A or -C was used with -c (count only), some lines of context were
701 accidentally printed for the final match.
703 (g) Added the -H (--with-filename) option.
705 (h) The combination of options -rh failed to suppress file names for files
706 that were found from directory arguments.
708 (i) Added the -D (--devices) and -d (--directories) options.
710 (j) Added the -F (--fixed-strings) option.
712 (k) Allow "-" to be used as a file name for -f as well as for a data file.
714 (l) Added the --colo(u)r option.
716 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
717 is not present by default.
719 3. A nasty bug was discovered in the handling of recursive patterns, that is,
720 items such as (?R) or (?1), when the recursion could match a number of
721 alternatives. If it matched one of the alternatives, but subsequently,
722 outside the recursion, there was a failure, the code tried to back up into
723 the recursion. However, because of the way PCRE is implemented, this is not
724 possible, and the result was an incorrect result from the match.
726 In order to prevent this happening, the specification of recursion has
727 been changed so that all such subpatterns are automatically treated as
728 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
730 4. I had overlooked the fact that, in some locales, there are characters for
731 which isalpha() is true but neither isupper() nor islower() are true. In
732 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
733 and ordfeminine) are like this. This affected the treatment of \w and \W
734 when they appeared in character classes, but not when they appeared outside
735 a character class. The bit map for "word" characters is now created
736 separately from the results of isalnum() instead of just taking it from the
737 upper, lower, and digit maps. (Plus the underscore character, of course.)
739 5. The above bug also affected the handling of POSIX character classes such as
740 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
741 permanent tables. Instead, the bit maps for such a class were previously
742 created as the appropriate unions of the upper, lower, and digit bitmaps.
743 Now they are created by subtraction from the [[:word:]] class, which has
744 its own bitmap.
746 6. The [[:blank:]] character class matches horizontal, but not vertical space.
747 It is created by subtracting the vertical space characters (\x09, \x0a,
748 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
749 subtraction was done in the overall bitmap for a character class, meaning
750 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
751 be recognized. This bug has been fixed.
753 7. Patches from the folks at Google:
755 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
756 real life, but is still worth protecting against".
758 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
759 regular expressions".
761 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
762 have it.
764 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
765 "configure" and the latter not, in order to fix a problem somebody had
766 with compiling the Arg class on HP-UX.
768 (e) Improve the error-handling of the C++ wrapper a little bit.
770 (f) New tests for checking recursion limiting.
772 8. The pcre_memmove() function, which is used only if the environment does not
773 have a standard memmove() function (and is therefore rarely compiled),
774 contained two bugs: (a) use of int instead of size_t, and (b) it was not
775 returning a result (though PCRE never actually uses the result).
777 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
778 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
779 returned instead of calling malloc() with an overflowing number that would
780 most likely cause subsequent chaos.
782 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
784 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
785 with this option is matched, the nmatch and pmatch options of regexec() are
786 ignored.
788 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
789 provided in case anyone wants to the the POSIX interface with UTF-8
790 strings.
792 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
793 C++ linking (needed for some HP-UX environments).
795 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
796 (unused parameter) and in the pcre_printint() function (omitted "default"
797 switch label when the default is to do nothing).
799 15. Added some code to make it possible, when PCRE is compiled as a C++
800 library, to replace subject pointers for pcre_exec() with a smart pointer
801 class, thus making it possible to process discontinuous strings.
803 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
804 much the same function. They were added by different people who were trying
805 to make PCRE easy to compile on non-Unix systems. It has been suggested
806 that PCRE_EXPORT be abolished now that there is more automatic apparatus
807 for compiling on Windows systems. I have therefore replaced it with
808 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
809 defaults to "extern" for C or "extern C" for C++, which works fine on
810 Unix-like systems. It is now possible to override the value of PCRE_DATA_
811 SCOPE with something explicit in config.h. In addition:
813 (a) pcreposix.h still had just "extern" instead of either of these macros;
814 I have replaced it with PCRE_DATA_SCOPE.
816 (b) Functions such as _pcre_xclass(), which are internal to the library,
817 but external in the C sense, all had PCRE_EXPORT in their definitions.
818 This is apparently wrong for the Windows case, so I have removed it.
819 (It makes no difference on Unix-like systems.)
821 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
822 of recursive calls to match(). This is different to MATCH_LIMIT because
823 that limits the total number of calls to match(), not all of which increase
824 the depth of recursion. Limiting the recursion depth limits the amount of
825 stack (or heap if NO_RECURSE is set) that is used. The default can be set
826 when PCRE is compiled, and changed at run time. A patch from Google adds
827 this functionality to the C++ interface.
829 18. Changes to the handling of Unicode character properties:
831 (a) Updated the table to Unicode 4.1.0.
833 (b) Recognize characters that are not in the table as "Cn" (undefined).
835 (c) I revised the way the table is implemented to a much improved format
836 which includes recognition of ranges. It now supports the ranges that
837 are defined in UnicodeData.txt, and it also amalgamates other
838 characters into ranges. This has reduced the number of entries in the
839 table from around 16,000 to around 3,000, thus reducing its size
840 considerably. I realized I did not need to use a tree structure after
841 all - a binary chop search is just as efficient. Having reduced the
842 number of entries, I extended their size from 6 bytes to 8 bytes to
843 allow for more data.
845 (d) Added support for Unicode script names via properties such as \p{Han}.
847 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
848 matching that character.
850 20. When matching a repeated Unicode property with a minimum greater than zero,
851 (for example \pL{2,}), PCRE could look past the end of the subject if it
852 reached it while seeking the minimum number of characters. This could
853 happen only if some of the characters were more than one byte long, because
854 there is a check for at least the minimum number of bytes.
856 21. Refactored the implementation of \p and \P so as to be more general, to
857 allow for more different types of property in future. This has changed the
858 compiled form incompatibly. Anybody with saved compiled patterns that use
859 \p or \P will have to recompile them.
861 22. Added "Any" and "L&" to the supported property types.
863 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
864 but give a compile time error if the value is greater than 0xff.
866 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
867 accidentally not being installed or uninstalled.
869 25. The pcre.h file was built from pcre.h.in, but the only changes that were
870 made were to insert the current release number. This seemed silly, because
871 it made things harder for people building PCRE on systems that don't run
872 "configure". I have turned pcre.h into a distributed file, no longer built
873 by "configure", with the version identification directly included. There is
874 no longer a pcre.h.in file.
876 However, this change necessitated a change to the pcre-config script as
877 well. It is built from pcre-config.in, and one of the substitutions was the
878 release number. I have updated configure.ac so that ./configure now finds
879 the release number by grepping pcre.h.
881 26. Added the ability to run the tests under valgrind.
884 Version 6.4 05-Sep-05
885 ---------------------
887 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
888 "--" to be printed when multiple files were scanned, even when none of the
889 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
890 consider it to be a bug, and have restored the previous behaviour.
892 2. A couple of code tidies to get rid of compiler warnings.
894 3. The pcretest program used to cheat by referring to symbols in the library
895 whose names begin with _pcre_. These are internal symbols that are not
896 really supposed to be visible externally, and in some environments it is
897 possible to suppress them. The cheating is now confined to including
898 certain files from the library's source, which is a bit cleaner.
900 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
901 file's purpose clearer.
903 5. Reorganized pcre_ucp_findchar().
906 Version 6.3 15-Aug-05
907 ---------------------
909 1. The file libpcre.pc.in did not have general read permission in the tarball.
911 2. There were some problems when building without C++ support:
913 (a) If C++ support was not built, "make install" and "make test" still
914 tried to test it.
916 (b) There were problems when the value of CXX was explicitly set. Some
917 changes have been made to try to fix these, and ...
919 (c) --disable-cpp can now be used to explicitly disable C++ support.
921 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
922 backslash in a target when C++ was disabled. This confuses some
923 versions of "make", apparently. Using an intermediate variable solves
924 this. (Same for CPP_LOBJ.)
926 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
927 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
928 necessary on certain architectures.
930 4. Added a setting of -export-symbols-regex to the link command to remove
931 those symbols that are exported in the C sense, but actually are local
932 within the library, and not documented. Their names all begin with
933 "_pcre_". This is not a perfect job, because (a) we have to except some
934 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
935 available (and never for static libraries). I have made a note to try to
936 find a way round (a) in the future.
939 Version 6.2 01-Aug-05
940 ---------------------
942 1. There was no test for integer overflow of quantifier values. A construction
943 such as {1111111111111111} would give undefined results. What is worse, if
944 a minimum quantifier for a parenthesized subpattern overflowed and became
945 negative, the calculation of the memory size went wrong. This could have
946 led to memory overwriting.
948 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
950 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
951 operating environments where this matters.
953 4. Applied Giuseppe Maxia's patch to add additional features for controlling
954 PCRE options from within the C++ wrapper.
956 5. Named capturing subpatterns were not being correctly counted when a pattern
957 was compiled. This caused two problems: (a) If there were more than 100
958 such subpatterns, the calculation of the memory needed for the whole
959 compiled pattern went wrong, leading to an overflow error. (b) Numerical
960 back references of the form \12, where the number was greater than 9, were
961 not recognized as back references, even though there were sufficient
962 previous subpatterns.
964 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
965 versions of gcc, e.g. 2.95.4.
968 Version 6.1 21-Jun-05
969 ---------------------
971 1. There was one reference to the variable "posix" in pcretest.c that was not
972 surrounded by "#if !defined NOPOSIX".
974 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
975 the cross-check on the old pcre_info() function, for the benefit of the
976 cut-down version of PCRE that is currently imported into Exim.
978 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
979 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
980 patterns that start (?i) but I don't think that matters. The use of (?i) is
981 just an example; this all applies to the other options as well.
983 4. Since libtool seems to echo the compile commands it is issuing, the output
984 from "make" can be reduced a bit by putting "@" in front of each libtool
985 compile command.
987 5. Patch from the folks at Google for configure.in to be a bit more thorough
988 in checking for a suitable C++ installation before trying to compile the
989 C++ stuff. This should fix a reported problem when a compiler was present,
990 but no suitable headers.
992 6. The man pages all had just "PCRE" as their title. I have changed them to
993 be the relevant file name. I have also arranged that these names are
994 retained in the file doc/pcre.txt, which is a concatenation in text format
995 of all the man pages except the little individual ones for each function.
997 7. The NON-UNIX-USE file had not been updated for the different set of source
998 files that come with release 6. I also added a few comments about the C++
999 wrapper.
1002 Version 6.0 07-Jun-05
1003 ---------------------
1005 1. Some minor internal re-organization to help with my DFA experiments.
1007 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1008 didn't matter for the library itself when fully configured, but did matter
1009 when compiling without UCP support, or within Exim, where the ucp files are
1010 not imported.
1012 3. Refactoring of the library code to split up the various functions into
1013 different source modules. The addition of the new DFA matching code (see
1014 below) to a single monolithic source would have made it really too
1015 unwieldy, quite apart from causing all the code to be include in a
1016 statically linked application, when only some functions are used. This is
1017 relevant even without the DFA addition now that patterns can be compiled in
1018 one application and matched in another.
1020 The downside of splitting up is that there have to be some external
1021 functions and data tables that are used internally in different modules of
1022 the library but which are not part of the API. These have all had their
1023 names changed to start with "_pcre_" so that they are unlikely to clash
1024 with other external names.
1026 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1027 a different (DFA) algorithm. Although it is slower than the original
1028 function, it does have some advantages for certain types of matching
1029 problem.
1031 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1032 including restarting after a partial match.
1034 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1035 defined when compiling for Windows was sent to me. I have put it into the
1036 code, though I have no means of testing or verifying it.
1038 7. Added the pcre_refcount() auxiliary function.
1040 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1041 match before or at the first newline in the subject string. In pcretest,
1042 the /f option on a pattern can be used to set this.
1044 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1045 would behave wrongly. This has been present in PCRE since release 4.0.
1047 10. A number of changes to the pcregrep command:
1049 (a) Refactored how -x works; insert ^(...)$ instead of setting
1050 PCRE_ANCHORED and checking the length, in preparation for adding
1051 something similar for -w.
1053 (b) Added the -w (match as a word) option.
1055 (c) Refactored the way lines are read and buffered so as to have more
1056 than one at a time available.
1058 (d) Implemented a pcregrep test script.
1060 (e) Added the -M (multiline match) option. This allows patterns to match
1061 over several lines of the subject. The buffering ensures that at least
1062 8K, or the rest of the document (whichever is the shorter) is available
1063 for matching (and similarly the previous 8K for lookbehind assertions).
1065 (f) Changed the --help output so that it now says
1067 -w, --word-regex(p)
1069 instead of two lines, one with "regex" and the other with "regexp"
1070 because that confused at least one person since the short forms are the
1071 same. (This required a bit of code, as the output is generated
1072 automatically from a table. It wasn't just a text change.)
1074 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1075 option but starts with a hyphen. Could be a pattern or a path name
1076 starting with a hyphen, for instance.
1078 (h) "-" can be given as a file name to represent stdin.
1080 (i) When file names are being printed, "(standard input)" is used for
1081 the standard input, for compatibility with GNU grep. Previously
1082 "<stdin>" was used.
1084 (j) The option --label=xxx can be used to supply a name to be used for
1085 stdin when file names are being printed. There is no short form.
1087 (k) Re-factored the options decoding logic because we are going to add
1088 two more options that take data. Such options can now be given in four
1089 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1091 (l) Added the -A, -B, and -C options for requesting that lines of context
1092 around matches be printed.
1094 (m) Added the -L option to print the names of files that do not contain
1095 any matching lines, that is, the complement of -l.
1097 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1098 continue to scan other files.
1100 (o) The -s option was incorrectly implemented. For compatibility with other
1101 greps, it now suppresses the error message for a non-existent or non-
1102 accessible file (but not the return code). There is a new option called
1103 -q that suppresses the output of matching lines, which was what -s was
1104 previously doing.
1106 (p) Added --include and --exclude options to specify files for inclusion
1107 and exclusion when recursing.
1109 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1110 Hopefully, it now does.
1112 12. Missing cast in pcre_study().
1114 13. Added an "uninstall" target to the makefile.
1116 14. Replaced "extern" in the function prototypes in Makefile.in with
1117 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1118 world, but is set differently for Windows.
1120 15. Added a second compiling function called pcre_compile2(). The only
1121 difference is that it has an extra argument, which is a pointer to an
1122 integer error code. When there is a compile-time failure, this is set
1123 non-zero, in addition to the error test pointer being set to point to an
1124 error message. The new argument may be NULL if no error number is required
1125 (but then you may as well call pcre_compile(), which is now just a
1126 wrapper). This facility is provided because some applications need a
1127 numeric error indication, but it has also enabled me to tidy up the way
1128 compile-time errors are handled in the POSIX wrapper.
1130 16. Added VPATH=.libs to the makefile; this should help when building with one
1131 prefix path and installing with another. (Or so I'm told by someone who
1132 knows more about this stuff than I do.)
1134 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1135 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1136 match everything, including newlines. This is not POSIX-compatible, but
1137 somebody wanted the feature. From pcretest it can be activated by using
1138 both the P and the s flags.
1140 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1142 19. libpcre.pc was being incorrectly installed as executable.
1144 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1145 it now also looks for '\r' so that it will work unmodified on Windows.
1147 21. Added Google's contributed C++ wrapper to the distribution.
1149 22. Added some untidy missing memory free() calls in pcretest, to keep
1150 Electric Fence happy when testing.
1154 Version 5.0 13-Sep-04
1155 ---------------------
1157 1. Internal change: literal characters are no longer packed up into items
1158 containing multiple characters in a single byte-string. Each character
1159 is now matched using a separate opcode. However, there may be more than one
1160 byte in the character in UTF-8 mode.
1162 2. The pcre_callout_block structure has two new fields: pattern_position and
1163 next_item_length. These contain the offset in the pattern to the next match
1164 item, and its length, respectively.
1166 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1167 insertion of callouts before each pattern item. Added the /C option to
1168 pcretest to make use of this.
1170 4. On the advice of a Windows user, the lines
1172 #if defined(_WIN32) || defined(WIN32)
1173 _setmode( _fileno( stdout ), 0x8000 );
1174 #endif /* defined(_WIN32) || defined(WIN32) */
1176 have been added to the source of pcretest. This apparently does useful
1177 magic in relation to line terminators.
1179 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1180 for the benefit of those environments where the "b" makes a difference.
1182 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1183 to know about it. I have put a hack into configure.in that adds in code
1184 to set GCC=yes if CC=icc. This seems to end up at a point in the
1185 generated configure script that is early enough to affect the setting of
1186 compiler options, which is what is needed, but I have no means of testing
1187 whether it really works. (The user who reported this had patched the
1188 generated configure script, which of course I cannot do.)
1190 LATER: After change 22 below (new libtool files), the configure script
1191 seems to know about icc (and also ecc). Therefore, I have commented out
1192 this hack in configure.in.
1194 7. Added support for pkg-config (2 patches were sent in).
1196 8. Negated POSIX character classes that used a combination of internal tables
1197 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1198 [[:^ascii]]. Typically, they would match almost any characters. The other
1199 POSIX classes were not broken in this way.
1201 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1202 to find the match, as PCRE was deluded into thinking that the match had to
1203 start at the start point or following a newline. The same bug applied to
1204 patterns with negative forward assertions or any backward assertions
1205 preceding ".*" at the start, unless the pattern required a fixed first
1206 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1208 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1209 starting at the last subject character, bytes beyond the end of the subject
1210 string were read.
1212 11. Renamed the variable "class" as "classbits" to make life easier for C++
1213 users. (Previously there was a macro definition, but it apparently wasn't
1214 enough.)
1216 12. Added the new field "tables" to the extra data so that tables can be passed
1217 in at exec time, or the internal tables can be re-selected. This allows
1218 a compiled regex to be saved and re-used at a later time by a different
1219 program that might have everything at different addresses.
1221 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1222 -R library as well as a -L library.
1224 14. The debugging options of pcretest (-d on the command line or D on a
1225 pattern) showed incorrect output for anything following an extended class
1226 that contained multibyte characters and which was followed by a quantifier.
1228 15. Added optional support for general category Unicode character properties
1229 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1230 support. It adds about 90K to the size of the library. The meanings of the
1231 inbuilt class escapes such as \d and \s have NOT been changed.
1233 16. Updated pcredemo.c to include calls to free() to release the memory for the
1234 compiled pattern.
1236 17. The generated file chartables.c was being created in the source directory
1237 instead of in the building directory. This caused the build to fail if the
1238 source directory was different from the building directory, and was
1239 read-only.
1241 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1242 file. No doubt somebody will tell me if they don't make sense... Also added
1243 Dan Mooney's comments about building on OpenVMS.
1245 19. Added support for partial matching via the PCRE_PARTIAL option for
1246 pcre_exec() and the \P data escape in pcretest.
1248 20. Extended pcretest with 3 new pattern features:
1250 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1251 write the compiled pattern to the file whose name is "rest-of-line".
1252 This is a straight binary dump of the data, with the saved pointer to
1253 the character tables forced to be NULL. The study data, if any, is
1254 written too. After writing, pcretest reads a new pattern.
1256 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1257 compiled pattern from the given file. There must not be any
1258 occurrences of "<" in the file name (pretty unlikely); if there are,
1259 pcretest will instead treat the initial "<" as a pattern delimiter.
1260 After reading in the pattern, pcretest goes on to read data lines as
1261 usual.
1263 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1264 and 16-bit fields in a compiled pattern, to simulate a pattern that
1265 was compiled on a host of opposite endianness.
1267 21. The pcre-exec() function can now cope with patterns that were compiled on
1268 hosts of opposite endianness, with this restriction:
1270 As for any compiled expression that is saved and used later, the tables
1271 pointer field cannot be preserved; the extra_data field in the arguments
1272 to pcre_exec() should be used to pass in a tables address if a value
1273 other than the default internal tables were used at compile time.
1275 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1276 now diagnosed as an error. Previously, most of the time, a negative number
1277 would have been treated as zero, but if in addition "ovector" was passed as
1278 NULL, a crash could occur.
1280 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1281 new versions from the libtool 1.5 distribution (the last one is a copy of
1282 a file called libtool.m4). This seems to have fixed the need to patch
1283 "configure" to support Darwin 1.3 (which I used to do). However, I still
1284 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1285 workstation).
1287 24. Changed the PCRE licence to be the more standard "BSD" licence.
1290 Version 4.5 01-Dec-03
1291 ---------------------
1293 1. There has been some re-arrangement of the code for the match() function so
1294 that it can be compiled in a version that does not call itself recursively.
1295 Instead, it keeps those local variables that need separate instances for
1296 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1297 needs to "recurse". Keeping track of where control must go is done by means
1298 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1299 hide most of the details from the main code, and operates only if
1300 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1301 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1302 operating.
1304 To make it easier for callers to provide specially tailored get/free
1305 functions for this usage, two new functions, pcre_stack_malloc, and
1306 pcre_stack_free, are used. They are always called in strict stacking order,
1307 and the size of block requested is always the same.
1309 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1310 PCRE has been compiled to use the stack or the heap for recursion. The
1311 -C option of pcretest uses this to show which version is compiled.
1313 A new data escape \S, is added to pcretest; it causes the amounts of store
1314 obtained and freed by both kinds of malloc/free at match time to be added
1315 to the output.
1317 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1318 what's available on my current Linux desktop machine.
1320 3. When matching a UTF-8 string, the test for a valid string at the start has
1321 been extended. If start_offset is not zero, PCRE now checks that it points
1322 to a byte that is the start of a UTF-8 character. If not, it returns
1323 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1324 this is necessary because there may be backward assertions in the pattern.
1325 When matching the same subject several times, it may save resources to use
1326 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1328 4. The code for checking the validity of UTF-8 strings has been tightened so
1329 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1330 containing "overlong sequences".
1332 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1333 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1334 should have been "&", but it just so happened that all the cases this let
1335 through by mistake were picked up later in the function.
1337 6. I had used a variable called "isblank" - this is a C99 function, causing
1338 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1340 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1341 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1342 so the version is shown; (c) stop "make test" echoing "./RunTest".
1344 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1346 9. The prototype for memmove() for systems that don't have it was using
1347 size_t, but the inclusion of the header that defines size_t was later. I've
1348 moved the #includes for the C headers earlier to avoid this.
1350 10. Added some adjustments to the code to make it easier to compiler on certain
1351 special systems:
1353 (a) Some "const" qualifiers were missing.
1354 (b) Added the macro EXPORT before all exported functions; by default this
1355 is defined to be empty.
1356 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1357 that it reads its output file name as an argument instead of writing
1358 to the standard output and assuming this can be redirected.
1360 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1361 class containing characters with values greater than 255, PCRE compilation
1362 went into a loop.
1364 12. A recursive reference to a subpattern that was within another subpattern
1365 that had a minimum quantifier of zero caused PCRE to crash. For example,
1366 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1367 recursion. If the recursively-called subpattern itself had a zero repeat,
1368 that was OK.
1370 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1371 buffer into which it was copied (for escape processing) was still set at
1372 1024, so long lines caused crashes.
1374 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1375 "internal error: code overflow...". This applied to any character class
1376 that was followed by a possessive quantifier.
1378 15. Modified the Makefile to add libpcre.la as a prerequisite for
1379 libpcreposix.la because I was told this is needed for a parallel build to
1380 work.
1382 16. If a pattern that contained .* following optional items at the start was
1383 studied, the wrong optimizing data was generated, leading to matching
1384 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1385 matching string must start with a or b or c. The correct conclusion for
1386 this pattern is that a match can start with any character.
1389 Version 4.4 13-Aug-03
1390 ---------------------
1392 1. In UTF-8 mode, a character class containing characters with values between
1393 127 and 255 was not handled correctly if the compiled pattern was studied.
1394 In fixing this, I have also improved the studying algorithm for such
1395 classes (slightly).
1397 2. Three internal functions had redundant arguments passed to them. Removal
1398 might give a very teeny performance improvement.
1400 3. Documentation bug: the value of the capture_top field in a callout is *one
1401 more than* the number of the hightest numbered captured substring.
1403 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1404 in incorrectly linking with a previously installed version. They now link
1405 explicitly with libpcre.la.
1407 5. configure.in no longer needs to recognize Cygwin specially.
1409 6. A problem in pcre.in for Windows platforms is fixed.
1411 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1412 pcretest, it used to include the size of the study block as part of its
1413 output. Unfortunately, the structure contains a field that has a different
1414 size on different hardware architectures. This meant that the tests that
1415 showed this size failed. As the block is currently always of a fixed size,
1416 this information isn't actually particularly useful in pcretest output, so
1417 I have just removed it.
1419 8. Three pre-processor statements accidentally did not start in column 1.
1420 Sadly, there are *still* compilers around that complain, even though
1421 standard C has not required this for well over a decade. Sigh.
1423 9. In pcretest, the code for checking callouts passed small integers in the
1424 callout_data field, which is a void * field. However, some picky compilers
1425 complained about the casts involved for this on 64-bit systems. Now
1426 pcretest passes the address of the small integer instead, which should get
1427 rid of the warnings.
1429 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1430 both compile and run time, and gives an error if an invalid UTF-8 sequence
1431 is found. There is a option for disabling this check in cases where the
1432 string is known to be correct and/or the maximum performance is wanted.
1434 11. In response to a bug report, I changed one line in Makefile.in from
1436 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1437 to
1438 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1440 to look similar to other lines, but I have no way of telling whether this
1441 is the right thing to do, as I do not use Windows. No doubt I'll get told
1442 if it's wrong...
1445 Version 4.3 21-May-03
1446 ---------------------
1448 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1449 Makefile.
1451 2. Some refactoring to improve the quality of the code:
1453 (i) The utf8_table... variables are now declared "const".
1455 (ii) The code for \cx, which used the "case flipping" table to upper case
1456 lower case letters, now just substracts 32. This is ASCII-specific,
1457 but the whole concept of \cx is ASCII-specific, so it seems
1458 reasonable.
1460 (iii) PCRE was using its character types table to recognize decimal and
1461 hexadecimal digits in the pattern. This is silly, because it handles
1462 only 0-9, a-f, and A-F, but the character types table is locale-
1463 specific, which means strange things might happen. A private
1464 table is now used for this - though it costs 256 bytes, a table is
1465 much faster than multiple explicit tests. Of course, the standard
1466 character types table is still used for matching digits in subject
1467 strings against \d.
1469 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1470 ending in _t are). So I've renamed it as ESC_tee.
1472 3. The first argument for regexec() in the POSIX wrapper should have been
1473 defined as "const".
1475 4. Changed pcretest to use malloc() for its buffers so that they can be
1476 Electric Fenced for debugging.
1478 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1479 to read one or more bytes before the start of the subject string. Often this
1480 had no effect on PCRE's behaviour, but in some circumstances it could
1481 provoke a segmentation fault.
1483 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1484 to try to read one or more bytes before the start of the subject string.
1486 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1487 UTF-8 support could misbehave in various ways if the subject string
1488 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1489 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1490 back over UTF-8 characters.)
1493 Version 4.2 14-Apr-03
1494 ---------------------
1496 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1498 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1499 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1500 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1501 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1502 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1504 Note: automatic setting of the BUILD variables is not yet working
1505 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1506 compile-time but not at link-time
1507 [LINK]: use for linking executables only
1508 make different versions for Windows and non-Windows
1509 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1510 libraries
1511 [LINK_FOR_BUILD]: new variable
1512 [OBJEXT]: use throughout
1513 [EXEEXT]: use throughout
1514 <winshared>: new target
1515 <wininstall>: new target
1516 <dftables.o>: use native compiler
1517 <dftables>: use native linker
1518 <install>: handle Windows platform correctly
1519 <clean>: ditto
1520 <check>: ditto
1521 copy DLL to top builddir before testing
1523 As part of these changes, -no-undefined was removed again. This was reported
1524 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1525 in any case.
1527 3. Some tidies to get rid of compiler warnings:
1529 . In the match_data structure, match_limit was an unsigned long int, whereas
1530 match_call_count was an int. I've made them both unsigned long ints.
1532 . In pcretest the fact that a const uschar * doesn't automatically cast to
1533 a void * provoked a warning.
1535 . Turning on some more compiler warnings threw up some "shadow" variables
1536 and a few more missing casts.
1538 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1539 option, a class that contained a single character with a value between 128
1540 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1542 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1543 option, a class that contained several characters, but with at least one
1544 whose value was between 128 and 255 caused PCRE to crash.
1547 Version 4.1 12-Mar-03
1548 ---------------------
1550 1. Compiling with gcc -pedantic found a couple of places where casts were
1551 needed, and a string in dftables.c that was longer than standard compilers are
1552 required to support.
1554 2. Compiling with Sun's compiler found a few more places where the code could
1555 be tidied up in order to avoid warnings.
1557 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1558 first of these names is deprecated in the latest Autoconf in favour of the name
1559 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1560 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1561 analogy I have changed it to CFLAGS_FOR_BUILD.
1563 4. Added -no-undefined to the linking command in the Makefile, because this is
1564 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1565 linking step for the pcreposix library.
1567 5. PCRE was failing to diagnose the case of two named groups with the same
1568 name.
1570 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1571 literal character that is needed in the subject for a match, and scans along to
1572 ensure that it is present before embarking on the full matching process. This
1573 saves time in cases of nested unlimited repeats that are never going to match.
1574 Problem: the scan can take a lot of time if the subject is very long (e.g.
1575 megabytes), thus penalizing straightforward matches. It is now done only if the
1576 amount of subject to be scanned is less than 1000 bytes.
1578 7. A lesser problem with the same optimization is that it was recording the
1579 first character of an anchored pattern as "needed", thus provoking a search
1580 right along the subject, even when the first match of the pattern was going to
1581 fail. The "needed" character is now not set for anchored patterns, unless it
1582 follows something in the pattern that is of non-fixed length. Thus, it still
1583 fulfils its original purpose of finding quick non-matches in cases of nested
1584 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1587 Version 4.0 17-Feb-03
1588 ---------------------
1590 1. If a comment in an extended regex that started immediately after a meta-item
1591 extended to the end of string, PCRE compiled incorrect data. This could lead to
1592 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1594 2. Moved to autoconf 2.53 and libtool 1.4.2.
1596 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1597 the special perltest8 script is no longer needed - all the tests can be run
1598 from a single perltest script.
1600 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1601 by \s. It has now been removed in PCRE. This means it isn't recognized as
1602 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1603 class [:space:] *does* include VT, thereby creating a mess.
1605 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1606 space and tab.
1608 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1609 its new features into the main test script, reducing the number of scripts.
1611 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1612 were backward compatible, and made the (?i) apply to the whole pattern, as if
1613 /i were given. Now it behaves more logically, and applies the option setting
1614 only to what follows. PCRE has been changed to follow suit. However, if it
1615 finds options settings right at the start of the pattern, it extracts them into
1616 the global options, as before. Thus, they show up in the info data.
1618 8. Added support for the \Q...\E escape sequence. Characters in between are
1619 treated as literals. This is slightly different from Perl in that $ and @ are
1620 also handled as literals inside the quotes. In Perl, they will cause variable
1621 interpolation. Note the following examples:
1623 Pattern PCRE matches Perl matches
1625 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1626 \Qabc\$xyz\E abc\$xyz abc\$xyz
1627 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1629 For compatibility with Perl, \Q...\E sequences are recognized inside character
1630 classes as well as outside them.
1632 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1633 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1634 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1635 signed/unsigned warnings.
1637 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1638 option for pcretest, so I've replaced it by a simple function that does just
1639 that job.
1641 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1642 "pcregrep -".
1644 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1645 Java package. This provides some syntactic sugar for simple cases of what my
1646 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1647 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1648 item, you can use this simplified notation. Note that only makes sense with
1649 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1650 greediness, whatever the setting of the PCRE_UNGREEDY option.
1652 13. A change of greediness default within a pattern was not taking effect at
1653 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1654 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1655 was abstracted outside.
1657 14. PCRE now supports the \G assertion. It is true when the current matching
1658 position is at the start point of the match. This differs from \A when the
1659 starting offset is non-zero. Used with the /g option of pcretest (or similar
1660 code), it works in the same way as it does for Perl's /g option. If all
1661 alternatives of a regex begin with \G, the expression is anchored to the start
1662 match position, and the "anchored" flag is set in the compiled expression.
1664 15. Some bugs concerning the handling of certain option changes within patterns
1665 have been fixed. These applied to options other than (?ims). For example,
1666 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1667 been the other way round. Some of this was related to change 7 above.
1669 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1670 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1671 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1672 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1674 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1675 mode. Unlike ".", it always matches newline, whatever the setting of
1676 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1677 assertions. Perl allows it, but it doesn't (in general) work because it can't
1678 calculate the length of the lookbehind. At least, that's the case for Perl
1679 5.8.0 - I've been told they are going to document that it doesn't work in
1680 future.
1682 18. Added an error diagnosis for escapes that PCRE does not support: these are
1683 \L, \l, \N, \P, \p, \U, \u, and \X.
1685 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1686 reading past the end of the pattern in cases such as /[abcd/.
1688 20. PCRE was getting more memory than necessary for patterns with classes that
1689 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1691 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1692 compiling PCRE for use with Virtual Pascal.
1694 22. Small fix to the Makefile to make it work properly if the build is done
1695 outside the source tree.
1697 23. Added a new extension: a condition to go with recursion. If a conditional
1698 subpattern starts with (?(R) the "true" branch is used if recursion has
1699 happened, whereas the "false" branch is used only at the top level.
1701 24. When there was a very long string of literal characters (over 255 bytes
1702 without UTF support, over 250 bytes with UTF support), the computation of how
1703 much memory was required could be incorrect, leading to segfaults or other
1704 strange effects.
1706 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1707 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1708 there was a subsequent back reference to those brackets. This meant that, for
1709 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1710 possible to check for precisely this case. All we can do is abandon the
1711 optimization if .* occurs inside capturing brackets when there are any back
1712 references whatsoever. (See below for a better fix that came later.)
1714 26. The handling of the optimization for finding the first character of a
1715 non-anchored pattern, and for finding a character that is required later in the
1716 match were failing in some cases. This didn't break the matching; it just
1717 failed to optimize when it could. The way this is done has been re-implemented.
1719 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1721 28. Added a new feature that provides some of the functionality that Perl
1722 provides with (?{...}). The facility is termed a "callout". The way it is done
1723 in PCRE is for the caller to provide an optional function, by setting
1724 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1725 global variable. By default it is unset, which disables all calling out. To get
1726 the function called, the regex must include (?C) at appropriate points. This
1727 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1728 This provides a means of identifying different callout points. When PCRE
1729 reaches such a point in the regex, if pcre_callout has been set, the external
1730 function is called. It is provided with data in a structure called
1731 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1732 matching continues; if it returns a non-zero value, the match at the current
1733 point fails. However, backtracking will occur if possible. [This was changed
1734 later and other features added - see item 49 below.]
1736 29. pcretest is upgraded to test the callout functionality. It provides a
1737 callout function that displays information. By default, it shows the start of
1738 the match and the current position in the text. There are some new data escapes
1739 to vary what happens:
1741 \C+ in addition, show current contents of captured substrings
1742 \C- do not supply a callout function
1743 \C!n return 1 when callout number n is reached
1744 \C!n!m return 1 when callout number n is reached for the mth time
1746 30. If pcregrep was called with the -l option and just a single file name, it
1747 output "<stdin>" if a match was found, instead of the file name.
1749 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1750 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1751 pcre_exec(). This saves a malloc/free per call. The default value of
1752 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1753 when configuring.
1755 32. The default maximum size of a compiled pattern is 64K. There have been a
1756 few cases of people hitting this limit. The code now uses macros to handle the
1757 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1758 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1759 configuring. Tests 2 and 5 work only with 2-byte links because they output
1760 debugging information about compiled patterns.
1762 33. Internal code re-arrangements:
1764 (a) Moved the debugging function for printing out a compiled regex into
1765 its own source file (printint.c) and used #include to pull it into
1766 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1767 separate copies.
1769 (b) Defined the list of op-code names for debugging as a macro in
1770 internal.h so that it is next to the definition of the opcodes.
1772 (c) Defined a table of op-code lengths for simpler skipping along compiled
1773 code. This is again a macro in internal.h so that it is next to the
1774 definition of the opcodes.
1776 34. Added support for recursive calls to individual subpatterns, along the
1777 lines of Robin Houston's patch (but implemented somewhat differently).
1779 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1780 allow it to read and process whole directories in Win32. This code was
1781 contributed by Lionel Fourquaux; it has not been tested by me.
1783 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1784 used to name a group. Names consist of alphanumerics and underscores, and must
1785 be unique. Back references use the syntax (?P=name) and recursive calls use
1786 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1787 numbers. The function pcre_fullinfo() can be used after compilation to extract
1788 a name/number map. There are three relevant calls:
1790 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1791 PCRE_INFO_NAMECOUNT yields the number of entries
1792 PCRE_INFO_NAMETABLE yields a pointer to the map.
1794 The map is a vector of fixed-size entries. The size of each entry depends on
1795 the length of the longest name used. The first two bytes of each entry are the
1796 group number, most significant byte first. There follows the corresponding
1797 name, zero terminated. The names are in alphabetical order.
1799 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1800 case instead of 255. Making it the same both with and without UTF-8 support
1801 means that the same test output works with both.
1803 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1804 calling malloc() with a zero argument.
1806 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1807 optimization. I've improved things by keeping a bitmap of backreferences with
1808 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1809 fact referenced, the optimization can be applied. It is unlikely that a
1810 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1811 the match to follow \n) will appear inside brackets with a number greater than
1812 31, but if it does, any back reference > 31 suppresses the optimization.
1814 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1815 of disabling numbered capturing parentheses. Any opening parenthesis that is
1816 not followed by ? behaves as if it were followed by ?: but named parentheses
1817 can still be used for capturing (and they will acquire numbers in the usual
1818 way).
1820 41. Redesigned the return codes from the match() function into yes/no/error so
1821 that errors can be passed back from deep inside the nested calls. A malloc
1822 failure while inside a recursive subpattern call now causes the
1823 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1825 42. It is now possible to set a limit on the number of times the match()
1826 function is called in a call to pcre_exec(). This facility makes it possible to
1827 limit the amount of recursion and backtracking, though not in a directly
1828 obvious way, because the match() function is used in a number of different
1829 circumstances. The count starts from zero for each position in the subject
1830 string (for non-anchored patterns). The default limit is, for compatibility, a
1831 large number, namely 10 000 000. You can change this in two ways:
1833 (a) When configuring PCRE before making, you can use --with-match-limit=n
1834 to set a default value for the compiled library.
1836 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1837 a different value is set. See 45 below.
1839 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1841 43. Added a new function pcre_config(int, void *) to enable run-time extraction
1842 of things that can be changed at compile time. The first argument specifies
1843 what is wanted and the second points to where the information is to be placed.
1844 The current list of available information is:
1848 The output is an integer that is set to one if UTF-8 support is available;
1849 otherwise it is set to zero.
1853 The output is an integer that it set to the value of the code that is used for
1854 newline. It is either LF (10) or CR (13).
1858 The output is an integer that contains the number of bytes used for internal
1859 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1863 The output is an integer that contains the threshold above which the POSIX
1864 interface uses malloc() for output vectors. See item 31 above.
1868 The output is an unsigned integer that contains the default limit of the number
1869 of match() calls in a pcre_exec() execution. See 42 above.
1871 44. pcretest has been upgraded by the addition of the -C option. This causes it
1872 to extract all the available output from the new pcre_config() function, and to
1873 output it. The program then exits immediately.
1875 45. A need has arisen to pass over additional data with calls to pcre_exec() in
1876 order to support additional features. One way would have been to define
1877 pcre_exec2() (for example) with extra arguments, but this would not have been
1878 extensible, and would also have required all calls to the original function to
1879 be mapped to the new one. Instead, I have chosen to extend the mechanism that
1880 is used for passing in "extra" data from pcre_study().
1882 The pcre_extra structure is now exposed and defined in pcre.h. It currently
1883 contains the following fields:
1885 flags a bitmap indicating which of the following fields are set
1886 study_data opaque data from pcre_study()
1887 match_limit a way of specifying a limit on match() calls for a specific
1888 call to pcre_exec()
1889 callout_data data for callouts (see 49 below)
1891 The flag bits are also defined in pcre.h, and are
1897 The pcre_study() function now returns one of these new pcre_extra blocks, with
1898 the actual study data pointed to by the study_data field, and the
1899 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1900 before. That is, this change is entirely upwards-compatible and requires no
1901 change to existing code.
1903 If you want to pass in additional data to pcre_exec(), you can either place it
1904 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1905 block.
1907 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1908 data string contains the escape sequence \M, pcretest calls pcre_exec() several
1909 times with different match limits, until it finds the minimum value needed for
1910 pcre_exec() to complete. The value is then output. This can be instructive; for
1911 most simple matches the number is quite small, but for pathological cases it
1912 gets very large very quickly.
1914 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1915 returns the size of the data block pointed to by the study_data field in a
1916 pcre_extra block, that is, the value that was passed as the argument to
1917 pcre_malloc() when PCRE was getting memory in which to place the information
1918 created by pcre_study(). The fourth argument should point to a size_t variable.
1919 pcretest has been extended so that this information is shown after a successful
1920 pcre_study() call when information about the compiled regex is being displayed.
1922 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1923 because what follows is always an absolute path. (Later: it turns out that this
1924 is more than cosmetic for MinGW, because it doesn't like empty path
1925 components.)
1927 49. Some changes have been made to the callout feature (see 28 above):
1929 (i) A callout function now has three choices for what it returns:
1931 0 => success, carry on matching
1932 > 0 => failure at this point, but backtrack if possible
1933 < 0 => serious error, return this value from pcre_exec()
1935 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1936 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1937 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1938 use by callout functions. It will never be used by PCRE itself.
1940 (ii) The pcre_extra structure (see 45 above) has a void * field called
1941 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1942 pcre_callout_block structure has a field of the same name. The contents of
1943 the field passed in the pcre_extra structure are passed to the callout
1944 function in the corresponding field in the callout block. This makes it
1945 easier to use the same callout-containing regex from multiple threads. For
1946 testing, the pcretest program has a new data escape
1948 \C*n pass the number n (may be negative) as callout_data
1950 If the callout function in pcretest receives a non-zero value as
1951 callout_data, it returns that value.
1953 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1954 there were some redundant $(CFLAGS) in commands that are now specified as
1955 $(LINK), which already includes $(CFLAGS).
1957 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1958 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1959 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1960 one-byte characters throughout. Note that case-insensitive matching applies
1961 only to characters whose values are less than 256. PCRE doesn't support the
1962 notion of cases for higher-valued characters.
1964 (i) A character class whose characters are all within 0-255 is handled as
1965 a bit map, and the map is inverted for negative classes. Previously, a
1966 character > 255 always failed to match such a class; however it should
1967 match if the class was a negative one (e.g. [^ab]). This has been fixed.
1969 (ii) A negated character class with a single character < 255 is coded as
1970 "not this character" (OP_NOT). This wasn't working properly when the test
1971 character was multibyte, either singly or repeated.
1973 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1974 mode, for example: \x{100}{2,3}.
1976 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1977 singly or repeated) now correctly test multibyte characters. However,
1978 PCRE doesn't recognize any characters with values greater than 255 as
1979 digits, spaces, or word characters. Such characters always match \D, \S,
1980 and \W, and never match \d, \s, or \w.
1982 (v) Classes may now contain characters and character ranges with values
1983 greater than 255. For example: [ab\x{100}-\x{400}].
1985 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1986 PCRE in UTF-8 mode.
1988 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1989 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1990 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1991 value.)
1993 53. The single man page has become too large. I have therefore split it up into
1994 a number of separate man pages. These also give rise to individual HTML pages;
1995 these are now put in a separate directory, and there is an index.html page that
1996 lists them all. Some hyperlinking between the pages has been installed.
1998 54. Added convenience functions for handling named capturing parentheses.
2000 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2001 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2002 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2003 are faulted.
2005 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2006 calling configure. These values are used when compiling the dftables.c program
2007 which is run to generate the source of the default character tables. They
2008 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2009 you will need to set these values.
2011 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2014 Version 3.9 02-Jan-02
2015 ---------------------
2017 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2019 2. If --disable-static was given, the building process failed when trying to
2020 build pcretest and pcregrep. (For some reason it was using libtool to compile
2021 them, which is not right, as they aren't part of the library.)
2024 Version 3.8 18-Dec-01
2025 ---------------------
2027 1. The experimental UTF-8 code was completely screwed up. It was packing the
2028 bytes in the wrong order. How dumb can you get?
2031 Version 3.7 29-Oct-01
2032 ---------------------
2034 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2035 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2036 this didn't happen under Solaris 8, where I normally test things.
2038 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2039 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2040 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2041 this fix an hour or so after the initial 3.7 release.)
2044 Version 3.6 23-Oct-01
2045 ---------------------
2047 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2048 offsets passed as NULL with zero offset count.
2050 2. The config.guess and config.sub files had not been updated when I moved to
2051 the latest autoconf.
2054 Version 3.5 15-Aug-01
2055 ---------------------
2057 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2058 had been forgotten.
2060 2. By using declared but undefined structures, we can avoid using "void"
2061 definitions in pcre.h while keeping the internal definitions of the structures
2062 private.
2064 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2065 user point of view, this means that both static and shared libraries are built
2066 by default, but this can be individually controlled. More of the work of
2067 handling this static/shared cases is now inside libtool instead of PCRE's make
2068 file.
2070 4. The pcretest utility is now installed along with pcregrep because it is
2071 useful for users (to test regexs) and by doing this, it automatically gets
2072 relinked by libtool. The documentation has been turned into a man page, so
2073 there are now .1, .txt, and .html versions in /doc.
2075 5. Upgrades to pcregrep:
2076 (i) Added long-form option names like gnu grep.
2077 (ii) Added --help to list all options with an explanatory phrase.
2078 (iii) Added -r, --recursive to recurse into sub-directories.
2079 (iv) Added -f, --file to read patterns from a file.
2081 6. pcre_exec() was referring to its "code" argument before testing that
2082 argument for NULL (and giving an error if it was NULL).
2084 7. Upgraded Makefile.in to allow for compiling in a different directory from
2085 the source directory.
2087 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2088 options bits, the pointer it was passed was to an int instead of to an unsigned
2089 long int. This mattered only on 64-bit systems.
2091 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2092 generated) instead of pcre.in, which it its source. Also made the same change
2093 in several of the .c files.
2095 10. A new release of gcc defines printf() as a macro, which broke pcretest
2096 because it had an ifdef in the middle of a string argument for printf(). Fixed
2097 by using separate calls to printf().
2099 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2100 script, to force use of CR or LF instead of \n in the source. On non-Unix
2101 systems, the value can be set in config.h.
2103 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2104 absolute limit. Changed the text of the error message to make this clear, and
2105 likewise updated the man page.
2107 13. The limit of 99 on the number of capturing subpatterns has been removed.
2108 The new limit is 65535, which I hope will not be a "real" limit.
2111 Version 3.4 22-Aug-00
2112 ---------------------
2114 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2116 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2119 Version 3.3 01-Aug-00
2120 ---------------------
2122 1. If an octal character was given, but the value was greater than \377, it
2123 was not getting masked to the least significant bits, as documented. This could
2124 lead to crashes in some systems.
2126 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2127 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2129 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2130 These just pass their arguments on to (pcre_free)(), but they are provided
2131 because some uses of PCRE bind it to non-C systems that can call its functions,
2132 but cannot call free() or pcre_free() directly.
2134 4. Add "make test" as a synonym for "make check". Corrected some comments in
2135 the Makefile.
2137 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2138 Makefile.
2140 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2141 command called pgrep for grepping around the active processes.
2143 7. Added the beginnings of support for UTF-8 character strings.
2145 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2146 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2147 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2148 out for the ar command.)
2151 Version 3.2 12-May-00
2152 ---------------------
2154 This is purely a bug fixing release.
2156 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2157 of ZA. This was just one example of several cases that could provoke this bug,
2158 which was introduced by change 9 of version 2.00. The code for breaking
2159 infinite loops after an iteration that matches an empty string was't working
2160 correctly.
2162 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2163 when matched against abbab (for example). After matching an empty string, it
2164 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2165 caused it to match further down the string than it should.
2167 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2168 was there because it doesn't seem to be needed, and it causes trouble on some
2169 systems, as it is not a Standard C header. It has been removed.
2171 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2172 were reported on the Macintosh. The changes were from
2174 while ((c = *(++ptr)) != 0 && c != '\n');
2175 to
2176 while ((c = *(++ptr)) != 0 && c != '\n') ;
2178 Totally extraordinary, but if that's what it takes...
2180 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2181 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2182 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2183 assumes the way PCRE uses memmove() (always moving upwards).
2185 6. PCRE is being used in one environment where strchr() is not available. There
2186 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2187 faster code anyway.
2190 Version 3.1 09-Feb-00
2191 ---------------------
2193 The only change in this release is the fixing of some bugs in Makefile.in for
2194 the "install" target:
2196 (1) It was failing to install pcreposix.h.
2198 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2201 Version 3.0 01-Feb-00
2202 ---------------------
2204 1. Add support for the /+ modifier to perltest (to output $` like it does in
2205 pcretest).
2207 2. Add support for the /g modifier to perltest.
2209 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2210 matches null strings.
2212 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2213 pattern. Perl treats empty patterns specially - it reuses the most recent
2214 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2215 effect.
2217 5. The POSIX interface was broken in that it was just handing over the POSIX
2218 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2219 required a bigger vector, with some working space on the end. This means that
2220 the POSIX wrapper now has to get and free some memory, and copy the results.
2222 6. Added some simple autoconf support, placing the test data and the
2223 documentation in separate directories, re-organizing some of the
2224 information files, and making it build pcre-config (a GNU standard). Also added
2225 libtool support for building PCRE as a shared library, which is now the
2226 default.
2228 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2229 09 are not valid octal constants. Single digits will be used for minor values
2230 less than 10.
2232 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2233 existing programs that set these in the POSIX interface can use PCRE without
2234 modification.
2236 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2237 return all that pcre_info() returns, plus additional data. The pcre_info()
2238 function is retained for compatibility, but is considered to be obsolete.
2240 10. Added experimental recursion feature (?R) to handle one common case that
2241 Perl 5.6 will be able to do with (?p{...}).
2243 11. Added support for POSIX character classes like [:alpha:], which Perl is
2244 adopting.
2247 Version 2.08 31-Aug-99
2248 ----------------------
2250 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2251 trying to match at the startoffset position, but instead was moving forward to
2252 the next newline as if a previous match had failed.
2254 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2255 and could get into a loop if a null string was matched other than at the start
2256 of the subject.
2258 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2259 be distinguished at compile time, and for completeness also added PCRE_DATE.
2261 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2262 in GnuWin32 environments.
2265 Version 2.07 29-Jul-99
2266 ----------------------
2268 1. The documentation is now supplied in plain text form and HTML as well as in
2269 the form of man page sources.
2271 2. C++ compilers don't like assigning (void *) values to other pointer types.
2272 In particular this affects malloc(). Although there is no problem in Standard
2273 C, I've put in casts to keep C++ compilers happy.
2275 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2276 should be (const char *).
2278 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2279 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2280 However, I haven't made this a standard facility. The documentation doesn't
2281 mention it, and the Makefile doesn't support it.
2283 5. The Makefile now contains an "install" target, with editable destinations at
2284 the top of the file. The pcretest program is not installed.
2286 6. pgrep -V now gives the PCRE version number and date.
2288 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2289 causing the entire string to be ignored, instead of just the last character.
2291 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2292 non-matching string, it can take a very, very long time, even for strings of
2293 quite modest length, because of the nested recursion. PCRE now does better in
2294 some of these cases. It does this by remembering the last required literal
2295 character in the pattern, and pre-searching the subject to ensure it is present
2296 before running the real match. In other words, it applies a heuristic to detect
2297 some types of certain failure quickly, and in the above example, if presented
2298 with a string that has no trailing " it gives "no match" very quickly.
2300 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2301 other alternatives are tried instead.
2304 Version 2.06 09-Jun-99
2305 ----------------------
2307 1. Change pcretest's output for amount of store used to show just the code
2308 space, because the remainder (the data block) varies in size between 32-bit and
2309 64-bit systems.
2311 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2312 start matching at. This allows lookbehinds to work when searching for multiple
2313 occurrences in a string.
2315 3. Added additional options to pcretest for testing multiple occurrences:
2317 /+ outputs the rest of the string that follows a match
2318 /g loops for multiple occurrences, using the new startoffset argument
2319 /G loops for multiple occurrences by passing an incremented pointer
2321 4. PCRE wasn't doing the "first character" optimization for patterns starting
2322 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2323 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2324 the letter 'x'. On long subject strings, this gives a significant speed-up.
2327 Version 2.05 21-Apr-99
2328 ----------------------
2330 1. Changed the type of magic_number from int to long int so that it works
2331 properly on 16-bit systems.
2333 2. Fixed a bug which caused patterns starting with .* not to work correctly
2334 when the subject string contained newline characters. PCRE was assuming
2335 anchoring for such patterns in all cases, which is not correct because .* will
2336 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2337 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2338 must be retried after every newline in the subject.
2341 Version 2.04 18-Feb-99
2342 ----------------------
2344 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2345 computation of the store needed to hold the pattern was incorrect (too large).
2346 If such patterns were nested a few deep, this could multiply and become a real
2347 problem.
2349 2. Added /M option to pcretest to show the memory requirement of a specific
2350 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2352 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2353 compiled in such a way that the backtracking after subsequent failure was
2354 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2355 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2358 Version 2.03 02-Feb-99
2359 ----------------------
2361 1. Fixed typo and small mistake in man page.
2363 2. Added 4th condition (GPL supersedes if conflict) and created separate
2364 LICENCE file containing the conditions.
2366 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2367 Perl, that is the internal \ allows the delimiter to be included in the
2368 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2369 the final delimiter, add \ to the end of the pattern (to test the error).
2371 4. Added the convenience functions for extracting substrings after a successful
2372 match. Updated pcretest to make it able to test these functions.
2375 Version 2.02 14-Jan-99
2376 ----------------------
2378 1. Initialized the working variables associated with each extraction so that
2379 their saving and restoring doesn't refer to uninitialized store.
2381 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2382 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2383 fix the problem.
2385 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2386 calls, and wasn't printing the correct value for compiling calls. Increased the
2387 default value of LOOPREPEAT, and the number of significant figures in the
2388 times.
2390 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2392 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2393 a building problem on Windows NT with a FAT file system.
2396 Version 2.01 21-Oct-98
2397 ----------------------
2399 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2400 to character tables built by pcre_maketables() in the current locale. If NULL
2401 is passed, the default tables are used.
2404 Version 2.00 24-Sep-98
2405 ----------------------
2407 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2408 it any more.
2410 2. Allow quantification of (?>) groups, and make it work correctly.
2412 3. The first character computation wasn't working for (?>) groups.
2414 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2415 end of the subject) and add 5.005's \z, which really does match only at the
2416 very end of the subject.
2418 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2420 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2421 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2422 localized options. All options to pcre_study() were also removed.
2424 7. Add other new features from 5.005:
2426 $(?<= positive lookbehind
2427 $(?<! negative lookbehind
2428 (?imsx-imsx) added the unsetting capability
2429 such a setting is global if at outer level; local otherwise
2430 (?imsx-imsx:) non-capturing groups with option setting
2431 (?(cond)re|re) conditional pattern matching
2433 A backreference to itself in a repeated group matches the previous
2434 captured string.
2436 8. General tidying up of studying (both automatic and via "study")
2437 consequential on the addition of new assertions.
2439 9. As in 5.005, unlimited repeated groups that could match an empty substring
2440 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2441 runtime if any iteration does actually match an empty substring.
2443 10. Include the RunTest script in the distribution.
2445 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2446 discrepancies, some of which were old and were also with respect to 5.004. They
2447 have now been fixed.
2450 Version 1.09 28-Apr-98
2451 ----------------------
2453 1. A negated single character class followed by a quantifier with a minimum
2454 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2455 program crashes, or just wrong answers. This did not apply to negated classes
2456 containing more than one character, or to minima other than one.
2459 Version 1.08 27-Mar-98
2460 ----------------------
2462 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2464 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2465 latter must appear before anything that relies on it in the pattern.
2468 Version 1.07 16-Feb-98
2469 ----------------------
2471 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2472 repeat of a potentially empty string).
2475 Version 1.06 23-Jan-98
2476 ----------------------
2478 1. Added Markus Oberhumer's little patches for C++.
2480 2. Literal strings longer than 255 characters were broken.
2483 Version 1.05 23-Dec-97
2484 ----------------------
2486 1. Negated character classes containing more than one character were failing if
2487 PCRE_CASELESS was set at run time.
2490 Version 1.04 19-Dec-97
2491 ----------------------
2493 1. Corrected the man page, where some "const" qualifiers had been omitted.
2495 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2496 input syntax.
2498 3. Fixed memory leak which occurred when a regex with back references was
2499 matched with an offsets vector that wasn't big enough. The temporary memory
2500 that is used in this case wasn't being freed if the match failed.
2502 4. Tidied pcretest to ensure it frees memory that it gets.
2504 5. Temporary memory was being obtained in the case where the passed offsets
2505 vector was exactly big enough.
2507 6. Corrected definition of offsetof() from change 5 below.
2509 7. I had screwed up change 6 below and broken the rules for the use of
2510 setjmp(). Now fixed.
2513 Version 1.03 18-Dec-97
2514 ----------------------
2516 1. A erroneous regex with a missing opening parenthesis was correctly
2517 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2518 on some systems.
2520 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2521 it was reported that one broken compiler failed on the former because "code" is
2522 also an independent variable.
2524 3. The erroneous regex a[]b caused an array overrun reference.
2526 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2527 fail on data ending with that character. (It was going on too far, and checking
2528 the next character, typically a binary zero.) This was specific to the
2529 optimized code for single-character negative classes.
2531 5. Added a contributed patch from the TIN world which does the following:
2533 + Add an undef for memmove, in case the the system defines a macro for it.
2535 + Add a definition of offsetof(), in case there isn't one. (I don't know
2536 the reason behind this - offsetof() is part of the ANSI standard - but
2537 it does no harm).
2539 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2540 most of the places where whitespace preceded '#'. I have given up and
2541 allowed the remaining 2 cases to be at the margin.
2543 + Rename some variables in pcre to eliminate shadowing. This seems very
2544 pedantic, but does no harm, of course.
2546 6. Moved the call to setjmp() into its own function, to get rid of warnings
2547 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2549 7. Constructs such as \d{8,} were compiling into the equivalent of
2550 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2551 outcome, but in this particular case used more store than had been allocated,
2552 which caused the bug to be discovered because it threw up an internal error.
2554 8. The debugging code in both pcre and pcretest for outputting the compiled
2555 form of a regex was going wrong in the case of back references followed by
2556 curly-bracketed repeats.
2559 Version 1.02 12-Dec-97
2560 ----------------------
2562 1. Typos in pcre.3 and comments in the source fixed.
2564 2. Applied a contributed patch to get rid of places where it used to remove
2565 'const' from variables, and fixed some signed/unsigned and uninitialized
2566 variable warnings.
2568 3. Added the "runtest" target to Makefile.
2570 4. Set default compiler flag to -O2 rather than just -O.
2573 Version 1.01 19-Nov-97
2574 ----------------------
2576 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2577 like /([ab]*)*/, that is, for classes with more than one character in them.
2579 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2580 as /((?>a*))*/ (a PCRE_EXTRA facility).
2583 Version 1.00 18-Nov-97
2584 ----------------------
2586 1. Added compile-time macros to support systems such as SunOS4 which don't have
2587 memmove() or strerror() but have other things that can be used instead.
2589 2. Arranged that "make clean" removes the executables.
2592 Version 0.99 27-Oct-97
2593 ----------------------
2595 1. Fixed bug in code for optimizing classes with only one character. It was
2596 initializing a 32-byte map regardless, which could cause it to run off the end
2597 of the memory it had got.
2599 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2602 Version 0.98 22-Oct-97
2603 ----------------------
2605 1. Fixed bug in code for handling temporary memory usage when there are more
2606 back references than supplied space in the ovector. This could cause segfaults.
2609 Version 0.97 21-Oct-97
2610 ----------------------
2612 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2614 2. Optimized negated single characters not to use a bit map.
2616 3. Brought error texts together as macro definitions; clarified some of them;
2617 fixed one that was wrong - it said "range out of order" when it meant "invalid
2618 escape sequence".
2620 4. Changed some char * arguments to const char *.
2622 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2624 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2625 pcretest.
2628 Version 0.96 16-Oct-97
2629 ----------------------
2631 1. Added a simple "pgrep" utility to the distribution.
2633 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2634 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2635 where "ddd" means "one or more decimal digits".
2637 3. Fixed serious bug. If a pattern had a back reference, but the call to
2638 pcre_exec() didn't supply a large enough ovector to record the related
2639 identifying subpattern, the match always failed. PCRE now remembers the number
2640 of the largest back reference, and gets some temporary memory in which to save
2641 the offsets during matching if necessary, in order to ensure that
2642 backreferences always work.
2644 4. Increased the compatibility with Perl in a number of ways:
2646 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2647 to request this handling. The option can be set at compile or exec time.
2649 (b) $ matches before a terminating newline by default; an option
2650 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2651 mode). The option can be set at compile or exec time.
2653 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2654 the same as Perl's. If the decimal number it represents is less than 10
2655 or there aren't that many previous left capturing parentheses, an octal
2656 escape is read. Inside a character class, it's always an octal escape,
2657 even if it is a single digit.
2659 (d) An escaped but undefined alphabetic character is taken as a literal,
2660 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2661 escapes.
2663 (e) {0} is now permitted. (The previous item is removed from the compiled
2664 pattern).
2666 5. Changed all the names of code files so that the basic parts are no longer
2667 than 10 characters, and abolished the teeny "globals.c" file.
2669 6. Changed the handling of character classes; they are now done with a 32-byte
2670 bit map always.
2672 7. Added the -d and /D options to pcretest to make it possible to look at the
2673 internals of compilation without having to recompile pcre.
2676 Version 0.95 23-Sep-97
2677 ----------------------
2679 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2680 \x20 at the start of a run of normal characters. These were being treated as
2681 real characters, instead of the source characters being re-checked.
2684 Version 0.94 18-Sep-97
2685 ----------------------
2687 1. The functions are now thread-safe, with the caveat that the global variables
2688 containing pointers to malloc() and free() or alternative functions are the
2689 same for all threads.
2691 2. Get pcre_study() to generate a bitmap of initial characters for non-
2692 anchored patterns when this is possible, and use it if passed to pcre_exec().
2695 Version 0.93 15-Sep-97
2696 ----------------------
2698 1. /(b)|(:+)/ was computing an incorrect first character.
2700 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2701 but not actually doing anything yet.
2703 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2704 as Perl does (e.g. [-az] or [az-]).
2706 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2707 all possible positions.
2709 5. Split up into different modules to avoid including unneeded functions in a
2710 compiled binary. However, compile and exec are still in one module. The "study"
2711 function is split off.
2713 6. The character tables are now in a separate module whose source is generated
2714 by an auxiliary program - but can then be edited by hand if required. There are
2715 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2716 toupper() in the code.
2718 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2719 make them global. Abolish the function for setting them, as the caller can now
2720 set them directly.
2723 Version 0.92 11-Sep-97
2724 ----------------------
2726 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2727 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2729 2. Caseless matching was not working in character classes if the characters in
2730 the pattern were in upper case.
2732 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2734 4. Make PCRE_ANCHORED public and accept as a compile option.
2736 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2737 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2738 pass them.
2740 6. Give an error if bad option bits passed at compile or run time.
2742 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2743 pcretest to cause it to pass that flag.
2745 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2746 options, and the first character, if set.
2748 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2751 Version 0.91 10-Sep-97
2752 ----------------------
2754 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2755 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2757 2. PCRE was looping on encountering an indefinitely repeated back reference to
2758 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2759 Perl does - treats the match as successful.
2761 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5