ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 192 - (show annotations)
Mon Jul 30 11:23:21 2007 UTC (8 years, 3 months ago) by ph10
File size: 131260 byte(s)
Error occurred while calculating annotation data.
Adjust opcode names in debugging output.
1 ChangeLog for PCRE
2 ------------------
4 Version 7.3 05-Jul-07
5 ---------------------
7 1. In the rejigging of the build system that eventually resulted in 7.1, the
8 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
9 brackets there is not right, since it causes compilers to look for an
10 installed pcre.h, not the version that is in the source that is being
11 compiled (which of course may be different). I have changed it back to:
13 #include "pcre.h"
15 I have a vague recollection that the change was concerned with compiling in
16 different directories, but in the new build system, that is taken care of
17 by the VPATH setting the Makefile.
19 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
20 when the subject happened to end in the byte 0x85 (e.g. if the last
21 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
22 characters but of course it shouldn't be taken as a newline when it is part
23 of another character. The bug was that, for an unlimited repeat of . in
24 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
25 characters when looking for a newline.
27 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
29 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
30 in debug output.
34 Version 7.2 19-Jun-07
35 ---------------------
37 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
38 which is apparently normally available under Windows.
40 2. Re-jig the pcregrep tests with different newline settings in an attempt
41 to make them independent of the local environment's newline setting.
43 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
45 4. Some of the "internals" tests were previously cut out when the link size
46 was not 2, because the output contained actual offsets. The recent new
47 "Z" feature of pcretest means that these can be cut out, making the tests
48 usable with all link sizes.
50 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
51 stack recursion. This gives a massive performance boost under BSD, but just
52 a small improvement under Linux. However, it saves one field in the frame
53 in all cases.
55 6. Added more features from the forthcoming Perl 5.10:
57 (a) (?-n) (where n is a string of digits) is a relative subroutine or
58 recursion call. It refers to the nth most recently opened parentheses.
60 (b) (?+n) is also a relative subroutine call; it refers to the nth next
61 to be opened parentheses.
63 (c) Conditions that refer to capturing parentheses can be specified
64 relatively, for example, (?(-2)... or (?(+3)...
66 (d) \K resets the start of the current match so that everything before
67 is not part of it.
69 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
71 (f) \g{name} is another synonym - part of Perl 5.10's unification of
72 reference syntax.
74 (g) (?| introduces a group in which the numbering of parentheses in each
75 alternative starts with the same number.
77 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
79 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
82 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
83 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
84 for detecting groups that can match an empty string.
86 9. A pattern with a very large number of alternatives (more than several
87 hundred) was running out of internal workspace during the pre-compile
88 phase, where pcre_compile() figures out how much memory will be needed. A
89 bit of new cunning has reduced the workspace needed for groups with
90 alternatives. The 1000-alternative test pattern now uses 12 bytes of
91 workspace instead of running out of the 4096 that are available.
93 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
95 11. Applied patch from Google to remove an optimization that didn't quite work.
96 The report of the bug said:
98 pcrecpp::RE("a*").FullMatch("aaa") matches, while
99 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
100 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
102 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
103 it matched the wrong number of bytes.
106 Version 7.1 24-Apr-07
107 ---------------------
109 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
110 that is more "standard", making use of automake and other Autotools. There
111 is some re-arrangement of the files and adjustment of comments consequent
112 on this.
114 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
115 for recursive directory scanning broke on some systems because the files
116 are not scanned in any specific order and on different systems the order
117 was different. A call to "sort" has been inserted into RunGrepTest for the
118 approprate test as a short-term fix. In the longer term there may be an
119 alternative.
121 3. I had an email from Eric Raymond about problems translating some of PCRE's
122 man pages to HTML (despite the fact that I distribute HTML pages, some
123 people do their own conversions for various reasons). The problems
124 concerned the use of low-level troff macros .br and .in. I have therefore
125 removed all such uses from the man pages (some were redundant, some could
126 be replaced by .nf/.fi pairs). The 132html script that I use to generate
127 HTML has been updated to handle .nf/.fi and to complain if it encounters
128 .br or .in.
130 4. Updated comments in configure.ac that get placed in config.h.in and also
131 arranged for config.h to be included in the distribution, with the name
132 config.h.generic, for the benefit of those who have to compile without
133 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
135 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
136 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
137 makevp.bat for the current PCRE, using the additional files
138 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
140 6. A Windows user reported a minor discrepancy with test 2, which turned out
141 to be caused by a trailing space on an input line that had got lost in his
142 copy. The trailing space was an accident, so I've just removed it.
144 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
145 that is needed.
147 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
148 as "const" (a) because they are and (b) because it helps the PHP
149 maintainers who have recently made a script to detect big data structures
150 in the php code that should be moved to the .rodata section. I remembered
151 to update Builducptable as well, so it won't revert if ucptable.h is ever
152 re-created.
154 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
155 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
156 order to be able to cut out the UTF-8 tables in the latter when UTF-8
157 support is not required. This saves 1.5-2K of code, which is important in
158 some applications.
160 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
161 so as not to refer to the tables, even though these functions will never be
162 called when UTF-8 support is disabled. Otherwise there are problems with a
163 shared library.
165 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
167 (a) It was defining its arguments as char * instead of void *.
169 (b) It was assuming that all moves were upwards in memory; this was true
170 a long time ago when I wrote it, but is no longer the case.
172 The emulated memove() is provided for those environments that have neither
173 memmove() nor bcopy(). I didn't think anyone used it these days, but that
174 is clearly not the case, as these two bugs were recently reported.
176 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
177 and Detrail to create the HTML documentation, the .txt form of the man
178 pages, and it removes trailing spaces from listed files. It also creates
179 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
180 case, it wraps all the #defines with #ifndefs. This script should be run
181 before "make dist".
183 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
184 with Unicode property support.
186 (a) For a maximizing quantifier, if the two different cases of the
187 character were of different lengths in their UTF-8 codings (there are
188 some cases like this - I found 11), and the matching function had to
189 back up over a mixture of the two cases, it incorrectly assumed they
190 were both the same length.
192 (b) When PCRE was configured to use the heap rather than the stack for
193 recursion during matching, it was not correctly preserving the data for
194 the other case of a UTF-8 character when checking ahead for a match
195 while processing a minimizing repeat. If the check also involved
196 matching a wide character, but failed, corruption could cause an
197 erroneous result when trying to check for a repeat of the original
198 character.
200 13. Some tidying changes to the testing mechanism:
202 (a) The RunTest script now detects the internal link size and whether there
203 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
204 values substituted by "configure". (The RunGrepTest script already did
205 this for UTF-8.) The configure.ac script no longer substitutes the
206 relevant variables.
208 (b) The debugging options /B and /D in pcretest show the compiled bytecode
209 with length and offset values. This means that the output is different
210 for different internal link sizes. Test 2 is skipped for link sizes
211 other than 2 because of this, bypassing the problem. Unfortunately,
212 there was also a test in test 3 (the locale tests) that used /B and
213 failed for link sizes other than 2. Rather than cut the whole test out,
214 I have added a new /Z option to pcretest that replaces the length and
215 offset values with spaces. This is now used to make test 3 independent
216 of link size. (Test 2 will be tidied up later.)
218 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
219 segmentation fault instead of returning the appropriate error message.
221 15. In multiline mode when the newline sequence was set to "any", the pattern
222 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
223 This doesn't seem right; it now treats the CRLF combination as the line
224 ending, and so does not match in that case. It's only a pattern such as ^$
225 that would hit this one: something like ^ABC$ would have failed after \r
226 and then tried again after \r\n.
228 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
229 in an attempt to make files that differ only in their line terminators
230 compare equal. This works on Linux.
232 17. Under certain error circumstances pcregrep might try to free random memory
233 as it exited. This is now fixed, thanks to valgrind.
235 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
236 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
237 was because its rules for how to advance for /g after matching an empty
238 string at the end of a line did not allow for this case. They now check for
239 it specially.
241 20. pcretest is supposed to handle patterns and data of any length, by
242 extending its buffers when necessary. It was getting this wrong when the
243 buffer for a data line had to be extended.
245 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
246 CRLF as a newline sequence.
248 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
249 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
250 I have nevertheless tidied it up.
252 23. Added some casts to kill warnings from HP-UX ia64 compiler.
254 24. Added a man page for pcre-config.
257 Version 7.0 19-Dec-06
258 ---------------------
260 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
261 moving to gcc 4.1.1.
263 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
264 sys/time.h, which is documented as needed for this function. It doesn't
265 seem to matter on Linux, but it showed up on some releases of OS X.
267 3. It seems that there are systems where bytes whose values are greater than
268 127 match isprint() in the "C" locale. The "C" locale should be the
269 default when a C program starts up. In most systems, only ASCII printing
270 characters match isprint(). This difference caused the output from pcretest
271 to vary, making some of the tests fail. I have changed pcretest so that:
273 (a) When it is outputting text in the compiled version of a pattern, bytes
274 other than 32-126 are always shown as hex escapes.
276 (b) When it is outputting text that is a matched part of a subject string,
277 it does the same, unless a different locale has been set for the match
278 (using the /L modifier). In this case, it uses isprint() to decide.
280 4. Fixed a major bug that caused incorrect computation of the amount of memory
281 required for a compiled pattern when options that changed within the
282 pattern affected the logic of the preliminary scan that determines the
283 length. The relevant options are -x, and -i in UTF-8 mode. The result was
284 that the computed length was too small. The symptoms of this bug were
285 either the PCRE error "internal error: code overflow" from pcre_compile(),
286 or a glibc crash with a message such as "pcretest: free(): invalid next
287 size (fast)". Examples of patterns that provoked this bug (shown in
288 pcretest format) are:
290 /(?-x: )/x
291 /(?x)(?-x: \s*#\s*)/
292 /((?i)[\x{c0}])/8
293 /(?i:[\x{c0}])/8
295 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
296 is now done differently.
298 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
299 wrapper classes; (b) implement a new function in the C++ scanner that is
300 more efficient than the old way of doing things because it avoids levels of
301 recursion in the regex matching; (c) add a paragraph to the documentation
302 for the FullMatch() function.
304 6. The escape sequence \n was being treated as whatever was defined as
305 "newline". Not only was this contrary to the documentation, which states
306 that \n is character 10 (hex 0A), but it also went horribly wrong when
307 "newline" was defined as CRLF. This has been fixed.
309 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
310 was being set to -1 for the "end of line" case (supposedly a value that no
311 character can have). Though this value is never used (the check for end of
312 line is "zero bytes in current character"), it caused compiler complaints.
313 I've changed it to 0xffffffff.
315 8. In pcre_version.c, the version string was being built by a sequence of
316 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
317 string (as it is for production releases) called a macro with an empty
318 argument. The C standard says the result of this is undefined. The gcc
319 compiler treats it as an empty string (which was what was wanted) but it is
320 reported that Visual C gives an error. The source has been hacked around to
321 avoid this problem.
323 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
324 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
325 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
326 of them did).
328 10. Originally, pcretest opened its input and output without "b"; then I was
329 told that "b" was needed in some environments, so it was added for release
330 5.0 to both the input and output. (It makes no difference on Unix-like
331 systems.) Later I was told that it is wrong for the input on Windows. I've
332 now abstracted the modes into two macros, to make it easier to fiddle with
333 them, and removed "b" from the input mode under Windows.
335 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
337 12. Added -help and --help to pcretest as an official way of being reminded
338 of the options.
340 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
341 and pcrecpp.cc because they annoy compilers at high warning levels.
343 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
345 15. Fixed an occurrence of == in configure.ac that should have been = (shell
346 scripts are not C programs :-) and which was not noticed because it works
347 on Linux.
349 16. pcretest is supposed to handle any length of pattern and data line (as one
350 line or as a continued sequence of lines) by extending its input buffer if
351 necessary. This feature was broken for very long pattern lines, leading to
352 a string of junk being passed to pcre_compile() if the pattern was longer
353 than about 50K.
355 17. I have done a major re-factoring of the way pcre_compile() computes the
356 amount of memory needed for a compiled pattern. Previously, there was code
357 that made a preliminary scan of the pattern in order to do this. That was
358 OK when PCRE was new, but as the facilities have expanded, it has become
359 harder and harder to keep it in step with the real compile phase, and there
360 have been a number of bugs (see for example, 4 above). I have now found a
361 cunning way of running the real compile function in a "fake" mode that
362 enables it to compute how much memory it would need, while actually only
363 ever using a few hundred bytes of working memory and without too many
364 tests of the mode. This should make future maintenance and development
365 easier. A side effect of this work is that the limit of 200 on the nesting
366 depth of parentheses has been removed (though this was never a serious
367 limitation, I suspect). However, there is a downside: pcre_compile() now
368 runs more slowly than before (30% or more, depending on the pattern). I
369 hope this isn't a big issue. There is no effect on runtime performance.
371 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
372 newline (only possible for the last line of a file) and it was a
373 pattern that set a locale (followed by /Lsomething), pcretest crashed.
375 19. Added additional timing features to pcretest. (1) The -tm option now times
376 matching only, not compiling. (2) Both -t and -tm can be followed, as a
377 separate command line item, by a number that specifies the number of
378 repeats to use when timing. The default is 50000; this gives better
379 precision, but takes uncomfortably long for very large patterns.
381 20. Extended pcre_study() to be more clever in cases where a branch of a
382 subpattern has no definite first character. For example, (a*|b*)[cd] would
383 previously give no result from pcre_study(). Now it recognizes that the
384 first character must be a, b, c, or d.
386 21. There was an incorrect error "recursive call could loop indefinitely" if
387 a subpattern (or the entire pattern) that was being tested for matching an
388 empty string contained only one non-empty item after a nested subpattern.
389 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
390 incorrectly, because the \d was being skipped in the check.
392 22. The pcretest program now has a new pattern option /B and a command line
393 option -b, which is equivalent to adding /B to every pattern. This causes
394 it to show the compiled bytecode, without the additional information that
395 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
396 is the same as /B/I).
398 23. A new optimization is now able automatically to treat some sequences such
399 as a*b as a*+b. More specifically, if something simple (such as a character
400 or a simple class like \d) has an unlimited quantifier, and is followed by
401 something that cannot possibly match the quantified thing, the quantifier
402 is automatically "possessified".
404 24. A recursive reference to a subpattern whose number was greater than 39
405 went wrong under certain circumstances in UTF-8 mode. This bug could also
406 have affected the operation of pcre_study().
408 25. Realized that a little bit of performance could be had by replacing
409 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
411 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
413 27. Possessive quantifiers such as a++ were previously implemented by turning
414 them into atomic groups such as ($>a+). Now they have their own opcodes,
415 which improves performance. This includes the automatically created ones
416 from 23 above.
418 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
419 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
420 the first matched character to be a colon. This applied both to named and
421 numbered groups.
423 29. The ucpinternal.h header file was missing its idempotency #ifdef.
425 30. I was sent a "project" file called libpcre.a.dev which I understand makes
426 building PCRE on Windows easier, so I have included it in the distribution.
428 31. There is now a check in pcretest against a ridiculously large number being
429 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
430 loop, the loop is abandoned.
432 32. Forward references to subpatterns in conditions such as (?(2)...) where
433 subpattern 2 is defined later cause pcre_compile() to search forwards in
434 the pattern for the relevant set of parentheses. This search went wrong
435 when there were unescaped parentheses in a character class, parentheses
436 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
438 33. "Subroutine" calls and backreferences were previously restricted to
439 referencing subpatterns earlier in the regex. This restriction has now
440 been removed.
442 34. Added a number of extra features that are going to be in Perl 5.10. On the
443 whole, these are just syntactic alternatives for features that PCRE had
444 previously implemented using the Python syntax or my own invention. The
445 other formats are all retained for compatibility.
447 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
448 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
449 also .NET compatible.
451 (b) A recursion or subroutine call to a named group can now be defined as
452 (?&name) as well as (?P>name).
454 (c) A backreference to a named group can now be defined as \k<name> or
455 \k'name' as well as (?P=name). The new forms, as well as being in Perl
456 5.10, are also .NET compatible.
458 (d) A conditional reference to a named group can now use the syntax
459 (?(<name>) or (?('name') as well as (?(name).
461 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
462 groups (named and numbered) that are never evaluated inline, but can be
463 called as "subroutines" from elsewhere. In effect, the DEFINE condition
464 is always false. There may be only one alternative in such a group.
466 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
467 as the simple (?(R). The condition is true only if the most recent
468 recursion is that of the given number or name. It does not search out
469 through the entire recursion stack.
471 (g) The escape \gN or \g{N} has been added, where N is a positive or
472 negative number, specifying an absolute or relative reference.
474 35. Tidied to get rid of some further signed/unsigned compiler warnings and
475 some "unreachable code" warnings.
477 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
478 things, this adds five new scripts.
480 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
481 There were also incompatibilities regarding the handling of \Q..\E inside
482 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
483 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
485 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
486 matches an empty string, and forcibly breaks the loop. There were bugs in
487 this code in non-simple cases. For a pattern such as ^(a()*)* matched
488 against aaaa the result was just "a" rather than "aaaa", for example. Two
489 separate and independent bugs (that affected different cases) have been
490 fixed.
492 39. Refactored the code to abolish the use of different opcodes for small
493 capturing bracket numbers. This is a tidy that I avoided doing when I
494 removed the limit on the number of capturing brackets for 3.5 back in 2001.
495 The new approach is not only tidier, it makes it possible to reduce the
496 memory needed to fix the previous bug (38).
498 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
499 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
500 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
501 mode.
503 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
504 report.
506 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
507 copy construction and assignment in the C++ wrapper.
509 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
510 couple of bugs that could have given wrong results in the "--newline=crlf"
511 case.
513 44. Added a number of casts and did some reorganization of signed/unsigned int
514 variables following suggestions from Dair Grant. Also renamed the variable
515 "this" as "item" because it is a C++ keyword.
517 45. Arranged for dftables to add
519 #include "pcre_internal.h"
521 to pcre_chartables.c because without it, gcc 4.x may remove the array
522 definition from the final binary if PCRE is built into a static library and
523 dead code stripping is activated.
525 46. For an unanchored pattern, if a match attempt fails at the start of a
526 newline sequence, and the newline setting is CRLF or ANY, and the next two
527 characters are CRLF, advance by two characters instead of one.
530 Version 6.7 04-Jul-06
531 ---------------------
533 1. In order to handle tests when input lines are enormously long, pcretest has
534 been re-factored so that it automatically extends its buffers when
535 necessary. The code is crude, but this _is_ just a test program. The
536 default size has been increased from 32K to 50K.
538 2. The code in pcre_study() was using the value of the re argument before
539 testing it for NULL. (Of course, in any sensible call of the function, it
540 won't be NULL.)
542 3. The memmove() emulation function in pcre_internal.h, which is used on
543 systems that lack both memmove() and bcopy() - that is, hardly ever -
544 was missing a "static" storage class specifier.
546 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
547 containing an extended class (one that cannot be represented by a bitmap
548 because it contains high-valued characters or Unicode property items, e.g.
549 [\pZ]). Almost always one would set UTF-8 mode when processing such a
550 pattern, but PCRE should not loop if you do not (it no longer does).
551 [Detail: two cases were found: (a) a repeated subpattern containing an
552 extended class; (b) a recursive reference to a subpattern that followed a
553 previous extended class. It wasn't skipping over the extended class
554 correctly when UTF-8 mode was not set.]
556 5. A negated single-character class was not being recognized as fixed-length
557 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
558 compile error "lookbehind assertion is not fixed length".
560 6. The RunPerlTest auxiliary script was showing an unexpected difference
561 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
562 write a Perl script that can interpret lines of an input file either as
563 byte characters or as UTF-8, which is what "perltest" was being required to
564 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
565 can't do is switch easily at run time between having the "use utf8;" pragma
566 or not. In the end, I fudged it by using the RunPerlTest script to insert
567 "use utf8;" explicitly for the UTF-8 tests.
569 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
570 the end of the subject string, contrary to the documentation and to what
571 Perl does. This was true of both matching functions. Now it matches only at
572 the start of the subject and immediately after *internal* newlines.
574 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
575 a pointer to an int instead of a pointer to an unsigned long int. This
576 caused problems on 64-bit systems.
578 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
579 instance of the 'standard' template library not being so standard".
581 10. There was no check on the number of named subpatterns nor the maximum
582 length of a subpattern name. The product of these values is used to compute
583 the size of the memory block for a compiled pattern. By supplying a very
584 long subpattern name and a large number of named subpatterns, the size
585 computation could be caused to overflow. This is now prevented by limiting
586 the length of names to 32 characters, and the number of named subpatterns
587 to 10,000.
589 11. Subpatterns that are repeated with specific counts have to be replicated in
590 the compiled pattern. The size of memory for this was computed from the
591 length of the subpattern and the repeat count. The latter is limited to
592 65535, but there was no limit on the former, meaning that integer overflow
593 could in principle occur. The compiled length of a repeated subpattern is
594 now limited to 30,000 bytes in order to prevent this.
596 12. Added the optional facility to have named substrings with the same name.
598 13. Added the ability to use a named substring as a condition, using the
599 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
600 are numbers (not recommended). Forward references are permitted.
602 14. Added forward references in named backreferences (if you see what I mean).
604 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
605 pattern could run off the end of the subject. For example, the pattern
606 "(?s)(.{1,5})"8 did this with the subject "ab".
608 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
609 PCRE_CASELESS was set when matching characters that were quantified with ?
610 or *.
612 17. A character class other than a single negated character that had a minimum
613 but no maximum quantifier - for example [ab]{6,} - was not handled
614 correctly by pce_dfa_exec(). It would match only one character.
616 18. A valid (though odd) pattern that looked like a POSIX character
617 class but used an invalid character after [ (for example [[,abc,]]) caused
618 pcre_compile() to give the error "Failed: internal error: code overflow" or
619 in some cases to crash with a glibc free() error. This could even happen if
620 the pattern terminated after [[ but there just happened to be a sequence of
621 letters, a binary zero, and a closing ] in the memory that followed.
623 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
624 over the years. Originally (before any Unicode support), just the bottom 8
625 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
626 output from "man perlunicode" includes this:
628 The regular expression compiler produces polymorphic opcodes. That
629 is, the pattern adapts to the data and automatically switches to
630 the Unicode character scheme when presented with Unicode data--or
631 instead uses a traditional byte scheme when presented with byte
632 data.
634 Sadly, a wide octal escape does not cause a switch, and in a string with
635 no other multibyte characters, these octal escapes are treated as before.
636 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
637 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
638 Unicode string.
640 I have not perpetrated such confusion in PCRE. Up till now, it took just
641 the bottom 8 bits, as in old Perl. I have now made octal escapes with
642 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
643 translate to the appropriate multibyte character.
645 29. Applied some refactoring to reduce the number of warnings from Microsoft
646 and Borland compilers. This has included removing the fudge introduced
647 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
648 a warning about an unused variable.
650 21. PCRE has not included VT (character 0x0b) in the set of whitespace
651 characters since release 4.0, because Perl (from release 5.004) does not.
652 [Or at least, is documented not to: some releases seem to be in conflict
653 with the documentation.] However, when a pattern was studied with
654 pcre_study() and all its branches started with \s, PCRE still included VT
655 as a possible starting character. Of course, this did no harm; it just
656 caused an unnecessary match attempt.
658 22. Removed a now-redundant internal flag bit that recorded the fact that case
659 dependency changed within the pattern. This was once needed for "required
660 byte" processing, but is no longer used. This recovers a now-scarce options
661 bit. Also moved the least significant internal flag bit to the most-
662 significant bit of the word, which was not previously used (hangover from
663 the days when it was an int rather than a uint) to free up another bit for
664 the future.
666 23. Added support for CRLF line endings as well as CR and LF. As well as the
667 default being selectable at build time, it can now be changed at runtime
668 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
669 specify that it is scanning data with non-default line endings.
671 24. Changed the definition of CXXLINK to make it agree with the definition of
672 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
674 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
675 recursions. This makes a big different to stack usage for some patterns.
677 26. If a subpattern containing a named recursion or subroutine reference such
678 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
679 the space required for the compiled pattern went wrong and gave too small a
680 value. Depending on the environment, this could lead to "Failed: internal
681 error: code overflow at offset 49" or "glibc detected double free or
682 corruption" errors.
684 27. Applied patches from Google (a) to support the new newline modes and (b) to
685 advance over multibyte UTF-8 characters in GlobalReplace.
687 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
688 difference for some implementation of PCRE in some Windows version.
690 29. Added some extra testing facilities to pcretest:
692 \q<number> in a data line sets the "match limit" value
693 \Q<number> in a data line sets the "match recursion limt" value
694 -S <number> sets the stack size, where <number> is in megabytes
696 The -S option isn't available for Windows.
699 Version 6.6 06-Feb-06
700 ---------------------
702 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
703 in pcreposix.h. I have copied the definition from pcre.h.
705 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
706 because pcre.h is no longer a built file.
708 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
709 not normally included in the compiled code.
712 Version 6.5 01-Feb-06
713 ---------------------
715 1. When using the partial match feature with pcre_dfa_exec(), it was not
716 anchoring the second and subsequent partial matches at the new starting
717 point. This could lead to incorrect results. For example, with the pattern
718 /1234/, partially matching against "123" and then "a4" gave a match.
720 2. Changes to pcregrep:
722 (a) All non-match returns from pcre_exec() were being treated as failures
723 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
724 error message is output. Some extra information is given for the
726 probably the only errors that are likely to be caused by users (by
727 specifying a regex that has nested indefinite repeats, for instance).
728 If there are more than 20 of these errors, pcregrep is abandoned.
730 (b) A binary zero was treated as data while matching, but terminated the
731 output line if it was written out. This has been fixed: binary zeroes
732 are now no different to any other data bytes.
734 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
735 used to set a locale for matching. The --locale=xxxx long option has
736 been added (no short equivalent) to specify a locale explicitly on the
737 pcregrep command, overriding the environment variables.
739 (d) When -B was used with -n, some line numbers in the output were one less
740 than they should have been.
742 (e) Added the -o (--only-matching) option.
744 (f) If -A or -C was used with -c (count only), some lines of context were
745 accidentally printed for the final match.
747 (g) Added the -H (--with-filename) option.
749 (h) The combination of options -rh failed to suppress file names for files
750 that were found from directory arguments.
752 (i) Added the -D (--devices) and -d (--directories) options.
754 (j) Added the -F (--fixed-strings) option.
756 (k) Allow "-" to be used as a file name for -f as well as for a data file.
758 (l) Added the --colo(u)r option.
760 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
761 is not present by default.
763 3. A nasty bug was discovered in the handling of recursive patterns, that is,
764 items such as (?R) or (?1), when the recursion could match a number of
765 alternatives. If it matched one of the alternatives, but subsequently,
766 outside the recursion, there was a failure, the code tried to back up into
767 the recursion. However, because of the way PCRE is implemented, this is not
768 possible, and the result was an incorrect result from the match.
770 In order to prevent this happening, the specification of recursion has
771 been changed so that all such subpatterns are automatically treated as
772 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
774 4. I had overlooked the fact that, in some locales, there are characters for
775 which isalpha() is true but neither isupper() nor islower() are true. In
776 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
777 and ordfeminine) are like this. This affected the treatment of \w and \W
778 when they appeared in character classes, but not when they appeared outside
779 a character class. The bit map for "word" characters is now created
780 separately from the results of isalnum() instead of just taking it from the
781 upper, lower, and digit maps. (Plus the underscore character, of course.)
783 5. The above bug also affected the handling of POSIX character classes such as
784 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
785 permanent tables. Instead, the bit maps for such a class were previously
786 created as the appropriate unions of the upper, lower, and digit bitmaps.
787 Now they are created by subtraction from the [[:word:]] class, which has
788 its own bitmap.
790 6. The [[:blank:]] character class matches horizontal, but not vertical space.
791 It is created by subtracting the vertical space characters (\x09, \x0a,
792 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
793 subtraction was done in the overall bitmap for a character class, meaning
794 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
795 be recognized. This bug has been fixed.
797 7. Patches from the folks at Google:
799 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
800 real life, but is still worth protecting against".
802 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
803 regular expressions".
805 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
806 have it.
808 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
809 "configure" and the latter not, in order to fix a problem somebody had
810 with compiling the Arg class on HP-UX.
812 (e) Improve the error-handling of the C++ wrapper a little bit.
814 (f) New tests for checking recursion limiting.
816 8. The pcre_memmove() function, which is used only if the environment does not
817 have a standard memmove() function (and is therefore rarely compiled),
818 contained two bugs: (a) use of int instead of size_t, and (b) it was not
819 returning a result (though PCRE never actually uses the result).
821 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
822 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
823 returned instead of calling malloc() with an overflowing number that would
824 most likely cause subsequent chaos.
826 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
828 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
829 with this option is matched, the nmatch and pmatch options of regexec() are
830 ignored.
832 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
833 provided in case anyone wants to the the POSIX interface with UTF-8
834 strings.
836 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
837 C++ linking (needed for some HP-UX environments).
839 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
840 (unused parameter) and in the pcre_printint() function (omitted "default"
841 switch label when the default is to do nothing).
843 15. Added some code to make it possible, when PCRE is compiled as a C++
844 library, to replace subject pointers for pcre_exec() with a smart pointer
845 class, thus making it possible to process discontinuous strings.
847 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
848 much the same function. They were added by different people who were trying
849 to make PCRE easy to compile on non-Unix systems. It has been suggested
850 that PCRE_EXPORT be abolished now that there is more automatic apparatus
851 for compiling on Windows systems. I have therefore replaced it with
852 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
853 defaults to "extern" for C or "extern C" for C++, which works fine on
854 Unix-like systems. It is now possible to override the value of PCRE_DATA_
855 SCOPE with something explicit in config.h. In addition:
857 (a) pcreposix.h still had just "extern" instead of either of these macros;
858 I have replaced it with PCRE_DATA_SCOPE.
860 (b) Functions such as _pcre_xclass(), which are internal to the library,
861 but external in the C sense, all had PCRE_EXPORT in their definitions.
862 This is apparently wrong for the Windows case, so I have removed it.
863 (It makes no difference on Unix-like systems.)
865 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
866 of recursive calls to match(). This is different to MATCH_LIMIT because
867 that limits the total number of calls to match(), not all of which increase
868 the depth of recursion. Limiting the recursion depth limits the amount of
869 stack (or heap if NO_RECURSE is set) that is used. The default can be set
870 when PCRE is compiled, and changed at run time. A patch from Google adds
871 this functionality to the C++ interface.
873 18. Changes to the handling of Unicode character properties:
875 (a) Updated the table to Unicode 4.1.0.
877 (b) Recognize characters that are not in the table as "Cn" (undefined).
879 (c) I revised the way the table is implemented to a much improved format
880 which includes recognition of ranges. It now supports the ranges that
881 are defined in UnicodeData.txt, and it also amalgamates other
882 characters into ranges. This has reduced the number of entries in the
883 table from around 16,000 to around 3,000, thus reducing its size
884 considerably. I realized I did not need to use a tree structure after
885 all - a binary chop search is just as efficient. Having reduced the
886 number of entries, I extended their size from 6 bytes to 8 bytes to
887 allow for more data.
889 (d) Added support for Unicode script names via properties such as \p{Han}.
891 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
892 matching that character.
894 20. When matching a repeated Unicode property with a minimum greater than zero,
895 (for example \pL{2,}), PCRE could look past the end of the subject if it
896 reached it while seeking the minimum number of characters. This could
897 happen only if some of the characters were more than one byte long, because
898 there is a check for at least the minimum number of bytes.
900 21. Refactored the implementation of \p and \P so as to be more general, to
901 allow for more different types of property in future. This has changed the
902 compiled form incompatibly. Anybody with saved compiled patterns that use
903 \p or \P will have to recompile them.
905 22. Added "Any" and "L&" to the supported property types.
907 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
908 but give a compile time error if the value is greater than 0xff.
910 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
911 accidentally not being installed or uninstalled.
913 25. The pcre.h file was built from pcre.h.in, but the only changes that were
914 made were to insert the current release number. This seemed silly, because
915 it made things harder for people building PCRE on systems that don't run
916 "configure". I have turned pcre.h into a distributed file, no longer built
917 by "configure", with the version identification directly included. There is
918 no longer a pcre.h.in file.
920 However, this change necessitated a change to the pcre-config script as
921 well. It is built from pcre-config.in, and one of the substitutions was the
922 release number. I have updated configure.ac so that ./configure now finds
923 the release number by grepping pcre.h.
925 26. Added the ability to run the tests under valgrind.
928 Version 6.4 05-Sep-05
929 ---------------------
931 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
932 "--" to be printed when multiple files were scanned, even when none of the
933 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
934 consider it to be a bug, and have restored the previous behaviour.
936 2. A couple of code tidies to get rid of compiler warnings.
938 3. The pcretest program used to cheat by referring to symbols in the library
939 whose names begin with _pcre_. These are internal symbols that are not
940 really supposed to be visible externally, and in some environments it is
941 possible to suppress them. The cheating is now confined to including
942 certain files from the library's source, which is a bit cleaner.
944 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
945 file's purpose clearer.
947 5. Reorganized pcre_ucp_findchar().
950 Version 6.3 15-Aug-05
951 ---------------------
953 1. The file libpcre.pc.in did not have general read permission in the tarball.
955 2. There were some problems when building without C++ support:
957 (a) If C++ support was not built, "make install" and "make test" still
958 tried to test it.
960 (b) There were problems when the value of CXX was explicitly set. Some
961 changes have been made to try to fix these, and ...
963 (c) --disable-cpp can now be used to explicitly disable C++ support.
965 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
966 backslash in a target when C++ was disabled. This confuses some
967 versions of "make", apparently. Using an intermediate variable solves
968 this. (Same for CPP_LOBJ.)
970 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
971 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
972 necessary on certain architectures.
974 4. Added a setting of -export-symbols-regex to the link command to remove
975 those symbols that are exported in the C sense, but actually are local
976 within the library, and not documented. Their names all begin with
977 "_pcre_". This is not a perfect job, because (a) we have to except some
978 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
979 available (and never for static libraries). I have made a note to try to
980 find a way round (a) in the future.
983 Version 6.2 01-Aug-05
984 ---------------------
986 1. There was no test for integer overflow of quantifier values. A construction
987 such as {1111111111111111} would give undefined results. What is worse, if
988 a minimum quantifier for a parenthesized subpattern overflowed and became
989 negative, the calculation of the memory size went wrong. This could have
990 led to memory overwriting.
992 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
994 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
995 operating environments where this matters.
997 4. Applied Giuseppe Maxia's patch to add additional features for controlling
998 PCRE options from within the C++ wrapper.
1000 5. Named capturing subpatterns were not being correctly counted when a pattern
1001 was compiled. This caused two problems: (a) If there were more than 100
1002 such subpatterns, the calculation of the memory needed for the whole
1003 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1004 back references of the form \12, where the number was greater than 9, were
1005 not recognized as back references, even though there were sufficient
1006 previous subpatterns.
1008 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1009 versions of gcc, e.g. 2.95.4.
1012 Version 6.1 21-Jun-05
1013 ---------------------
1015 1. There was one reference to the variable "posix" in pcretest.c that was not
1016 surrounded by "#if !defined NOPOSIX".
1018 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1019 the cross-check on the old pcre_info() function, for the benefit of the
1020 cut-down version of PCRE that is currently imported into Exim.
1022 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1023 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1024 patterns that start (?i) but I don't think that matters. The use of (?i) is
1025 just an example; this all applies to the other options as well.
1027 4. Since libtool seems to echo the compile commands it is issuing, the output
1028 from "make" can be reduced a bit by putting "@" in front of each libtool
1029 compile command.
1031 5. Patch from the folks at Google for configure.in to be a bit more thorough
1032 in checking for a suitable C++ installation before trying to compile the
1033 C++ stuff. This should fix a reported problem when a compiler was present,
1034 but no suitable headers.
1036 6. The man pages all had just "PCRE" as their title. I have changed them to
1037 be the relevant file name. I have also arranged that these names are
1038 retained in the file doc/pcre.txt, which is a concatenation in text format
1039 of all the man pages except the little individual ones for each function.
1041 7. The NON-UNIX-USE file had not been updated for the different set of source
1042 files that come with release 6. I also added a few comments about the C++
1043 wrapper.
1046 Version 6.0 07-Jun-05
1047 ---------------------
1049 1. Some minor internal re-organization to help with my DFA experiments.
1051 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1052 didn't matter for the library itself when fully configured, but did matter
1053 when compiling without UCP support, or within Exim, where the ucp files are
1054 not imported.
1056 3. Refactoring of the library code to split up the various functions into
1057 different source modules. The addition of the new DFA matching code (see
1058 below) to a single monolithic source would have made it really too
1059 unwieldy, quite apart from causing all the code to be include in a
1060 statically linked application, when only some functions are used. This is
1061 relevant even without the DFA addition now that patterns can be compiled in
1062 one application and matched in another.
1064 The downside of splitting up is that there have to be some external
1065 functions and data tables that are used internally in different modules of
1066 the library but which are not part of the API. These have all had their
1067 names changed to start with "_pcre_" so that they are unlikely to clash
1068 with other external names.
1070 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1071 a different (DFA) algorithm. Although it is slower than the original
1072 function, it does have some advantages for certain types of matching
1073 problem.
1075 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1076 including restarting after a partial match.
1078 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1079 defined when compiling for Windows was sent to me. I have put it into the
1080 code, though I have no means of testing or verifying it.
1082 7. Added the pcre_refcount() auxiliary function.
1084 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1085 match before or at the first newline in the subject string. In pcretest,
1086 the /f option on a pattern can be used to set this.
1088 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1089 would behave wrongly. This has been present in PCRE since release 4.0.
1091 10. A number of changes to the pcregrep command:
1093 (a) Refactored how -x works; insert ^(...)$ instead of setting
1094 PCRE_ANCHORED and checking the length, in preparation for adding
1095 something similar for -w.
1097 (b) Added the -w (match as a word) option.
1099 (c) Refactored the way lines are read and buffered so as to have more
1100 than one at a time available.
1102 (d) Implemented a pcregrep test script.
1104 (e) Added the -M (multiline match) option. This allows patterns to match
1105 over several lines of the subject. The buffering ensures that at least
1106 8K, or the rest of the document (whichever is the shorter) is available
1107 for matching (and similarly the previous 8K for lookbehind assertions).
1109 (f) Changed the --help output so that it now says
1111 -w, --word-regex(p)
1113 instead of two lines, one with "regex" and the other with "regexp"
1114 because that confused at least one person since the short forms are the
1115 same. (This required a bit of code, as the output is generated
1116 automatically from a table. It wasn't just a text change.)
1118 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1119 option but starts with a hyphen. Could be a pattern or a path name
1120 starting with a hyphen, for instance.
1122 (h) "-" can be given as a file name to represent stdin.
1124 (i) When file names are being printed, "(standard input)" is used for
1125 the standard input, for compatibility with GNU grep. Previously
1126 "<stdin>" was used.
1128 (j) The option --label=xxx can be used to supply a name to be used for
1129 stdin when file names are being printed. There is no short form.
1131 (k) Re-factored the options decoding logic because we are going to add
1132 two more options that take data. Such options can now be given in four
1133 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1135 (l) Added the -A, -B, and -C options for requesting that lines of context
1136 around matches be printed.
1138 (m) Added the -L option to print the names of files that do not contain
1139 any matching lines, that is, the complement of -l.
1141 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1142 continue to scan other files.
1144 (o) The -s option was incorrectly implemented. For compatibility with other
1145 greps, it now suppresses the error message for a non-existent or non-
1146 accessible file (but not the return code). There is a new option called
1147 -q that suppresses the output of matching lines, which was what -s was
1148 previously doing.
1150 (p) Added --include and --exclude options to specify files for inclusion
1151 and exclusion when recursing.
1153 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1154 Hopefully, it now does.
1156 12. Missing cast in pcre_study().
1158 13. Added an "uninstall" target to the makefile.
1160 14. Replaced "extern" in the function prototypes in Makefile.in with
1161 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1162 world, but is set differently for Windows.
1164 15. Added a second compiling function called pcre_compile2(). The only
1165 difference is that it has an extra argument, which is a pointer to an
1166 integer error code. When there is a compile-time failure, this is set
1167 non-zero, in addition to the error test pointer being set to point to an
1168 error message. The new argument may be NULL if no error number is required
1169 (but then you may as well call pcre_compile(), which is now just a
1170 wrapper). This facility is provided because some applications need a
1171 numeric error indication, but it has also enabled me to tidy up the way
1172 compile-time errors are handled in the POSIX wrapper.
1174 16. Added VPATH=.libs to the makefile; this should help when building with one
1175 prefix path and installing with another. (Or so I'm told by someone who
1176 knows more about this stuff than I do.)
1178 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1179 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1180 match everything, including newlines. This is not POSIX-compatible, but
1181 somebody wanted the feature. From pcretest it can be activated by using
1182 both the P and the s flags.
1184 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1186 19. libpcre.pc was being incorrectly installed as executable.
1188 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1189 it now also looks for '\r' so that it will work unmodified on Windows.
1191 21. Added Google's contributed C++ wrapper to the distribution.
1193 22. Added some untidy missing memory free() calls in pcretest, to keep
1194 Electric Fence happy when testing.
1198 Version 5.0 13-Sep-04
1199 ---------------------
1201 1. Internal change: literal characters are no longer packed up into items
1202 containing multiple characters in a single byte-string. Each character
1203 is now matched using a separate opcode. However, there may be more than one
1204 byte in the character in UTF-8 mode.
1206 2. The pcre_callout_block structure has two new fields: pattern_position and
1207 next_item_length. These contain the offset in the pattern to the next match
1208 item, and its length, respectively.
1210 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1211 insertion of callouts before each pattern item. Added the /C option to
1212 pcretest to make use of this.
1214 4. On the advice of a Windows user, the lines
1216 #if defined(_WIN32) || defined(WIN32)
1217 _setmode( _fileno( stdout ), 0x8000 );
1218 #endif /* defined(_WIN32) || defined(WIN32) */
1220 have been added to the source of pcretest. This apparently does useful
1221 magic in relation to line terminators.
1223 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1224 for the benefit of those environments where the "b" makes a difference.
1226 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1227 to know about it. I have put a hack into configure.in that adds in code
1228 to set GCC=yes if CC=icc. This seems to end up at a point in the
1229 generated configure script that is early enough to affect the setting of
1230 compiler options, which is what is needed, but I have no means of testing
1231 whether it really works. (The user who reported this had patched the
1232 generated configure script, which of course I cannot do.)
1234 LATER: After change 22 below (new libtool files), the configure script
1235 seems to know about icc (and also ecc). Therefore, I have commented out
1236 this hack in configure.in.
1238 7. Added support for pkg-config (2 patches were sent in).
1240 8. Negated POSIX character classes that used a combination of internal tables
1241 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1242 [[:^ascii]]. Typically, they would match almost any characters. The other
1243 POSIX classes were not broken in this way.
1245 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1246 to find the match, as PCRE was deluded into thinking that the match had to
1247 start at the start point or following a newline. The same bug applied to
1248 patterns with negative forward assertions or any backward assertions
1249 preceding ".*" at the start, unless the pattern required a fixed first
1250 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1252 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1253 starting at the last subject character, bytes beyond the end of the subject
1254 string were read.
1256 11. Renamed the variable "class" as "classbits" to make life easier for C++
1257 users. (Previously there was a macro definition, but it apparently wasn't
1258 enough.)
1260 12. Added the new field "tables" to the extra data so that tables can be passed
1261 in at exec time, or the internal tables can be re-selected. This allows
1262 a compiled regex to be saved and re-used at a later time by a different
1263 program that might have everything at different addresses.
1265 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1266 -R library as well as a -L library.
1268 14. The debugging options of pcretest (-d on the command line or D on a
1269 pattern) showed incorrect output for anything following an extended class
1270 that contained multibyte characters and which was followed by a quantifier.
1272 15. Added optional support for general category Unicode character properties
1273 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1274 support. It adds about 90K to the size of the library. The meanings of the
1275 inbuilt class escapes such as \d and \s have NOT been changed.
1277 16. Updated pcredemo.c to include calls to free() to release the memory for the
1278 compiled pattern.
1280 17. The generated file chartables.c was being created in the source directory
1281 instead of in the building directory. This caused the build to fail if the
1282 source directory was different from the building directory, and was
1283 read-only.
1285 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1286 file. No doubt somebody will tell me if they don't make sense... Also added
1287 Dan Mooney's comments about building on OpenVMS.
1289 19. Added support for partial matching via the PCRE_PARTIAL option for
1290 pcre_exec() and the \P data escape in pcretest.
1292 20. Extended pcretest with 3 new pattern features:
1294 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1295 write the compiled pattern to the file whose name is "rest-of-line".
1296 This is a straight binary dump of the data, with the saved pointer to
1297 the character tables forced to be NULL. The study data, if any, is
1298 written too. After writing, pcretest reads a new pattern.
1300 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1301 compiled pattern from the given file. There must not be any
1302 occurrences of "<" in the file name (pretty unlikely); if there are,
1303 pcretest will instead treat the initial "<" as a pattern delimiter.
1304 After reading in the pattern, pcretest goes on to read data lines as
1305 usual.
1307 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1308 and 16-bit fields in a compiled pattern, to simulate a pattern that
1309 was compiled on a host of opposite endianness.
1311 21. The pcre-exec() function can now cope with patterns that were compiled on
1312 hosts of opposite endianness, with this restriction:
1314 As for any compiled expression that is saved and used later, the tables
1315 pointer field cannot be preserved; the extra_data field in the arguments
1316 to pcre_exec() should be used to pass in a tables address if a value
1317 other than the default internal tables were used at compile time.
1319 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1320 now diagnosed as an error. Previously, most of the time, a negative number
1321 would have been treated as zero, but if in addition "ovector" was passed as
1322 NULL, a crash could occur.
1324 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1325 new versions from the libtool 1.5 distribution (the last one is a copy of
1326 a file called libtool.m4). This seems to have fixed the need to patch
1327 "configure" to support Darwin 1.3 (which I used to do). However, I still
1328 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1329 workstation).
1331 24. Changed the PCRE licence to be the more standard "BSD" licence.
1334 Version 4.5 01-Dec-03
1335 ---------------------
1337 1. There has been some re-arrangement of the code for the match() function so
1338 that it can be compiled in a version that does not call itself recursively.
1339 Instead, it keeps those local variables that need separate instances for
1340 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1341 needs to "recurse". Keeping track of where control must go is done by means
1342 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1343 hide most of the details from the main code, and operates only if
1344 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1345 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1346 operating.
1348 To make it easier for callers to provide specially tailored get/free
1349 functions for this usage, two new functions, pcre_stack_malloc, and
1350 pcre_stack_free, are used. They are always called in strict stacking order,
1351 and the size of block requested is always the same.
1353 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1354 PCRE has been compiled to use the stack or the heap for recursion. The
1355 -C option of pcretest uses this to show which version is compiled.
1357 A new data escape \S, is added to pcretest; it causes the amounts of store
1358 obtained and freed by both kinds of malloc/free at match time to be added
1359 to the output.
1361 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1362 what's available on my current Linux desktop machine.
1364 3. When matching a UTF-8 string, the test for a valid string at the start has
1365 been extended. If start_offset is not zero, PCRE now checks that it points
1366 to a byte that is the start of a UTF-8 character. If not, it returns
1367 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1368 this is necessary because there may be backward assertions in the pattern.
1369 When matching the same subject several times, it may save resources to use
1370 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1372 4. The code for checking the validity of UTF-8 strings has been tightened so
1373 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1374 containing "overlong sequences".
1376 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1377 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1378 should have been "&", but it just so happened that all the cases this let
1379 through by mistake were picked up later in the function.
1381 6. I had used a variable called "isblank" - this is a C99 function, causing
1382 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1384 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1385 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1386 so the version is shown; (c) stop "make test" echoing "./RunTest".
1388 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1390 9. The prototype for memmove() for systems that don't have it was using
1391 size_t, but the inclusion of the header that defines size_t was later. I've
1392 moved the #includes for the C headers earlier to avoid this.
1394 10. Added some adjustments to the code to make it easier to compiler on certain
1395 special systems:
1397 (a) Some "const" qualifiers were missing.
1398 (b) Added the macro EXPORT before all exported functions; by default this
1399 is defined to be empty.
1400 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1401 that it reads its output file name as an argument instead of writing
1402 to the standard output and assuming this can be redirected.
1404 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1405 class containing characters with values greater than 255, PCRE compilation
1406 went into a loop.
1408 12. A recursive reference to a subpattern that was within another subpattern
1409 that had a minimum quantifier of zero caused PCRE to crash. For example,
1410 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1411 recursion. If the recursively-called subpattern itself had a zero repeat,
1412 that was OK.
1414 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1415 buffer into which it was copied (for escape processing) was still set at
1416 1024, so long lines caused crashes.
1418 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1419 "internal error: code overflow...". This applied to any character class
1420 that was followed by a possessive quantifier.
1422 15. Modified the Makefile to add libpcre.la as a prerequisite for
1423 libpcreposix.la because I was told this is needed for a parallel build to
1424 work.
1426 16. If a pattern that contained .* following optional items at the start was
1427 studied, the wrong optimizing data was generated, leading to matching
1428 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1429 matching string must start with a or b or c. The correct conclusion for
1430 this pattern is that a match can start with any character.
1433 Version 4.4 13-Aug-03
1434 ---------------------
1436 1. In UTF-8 mode, a character class containing characters with values between
1437 127 and 255 was not handled correctly if the compiled pattern was studied.
1438 In fixing this, I have also improved the studying algorithm for such
1439 classes (slightly).
1441 2. Three internal functions had redundant arguments passed to them. Removal
1442 might give a very teeny performance improvement.
1444 3. Documentation bug: the value of the capture_top field in a callout is *one
1445 more than* the number of the hightest numbered captured substring.
1447 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1448 in incorrectly linking with a previously installed version. They now link
1449 explicitly with libpcre.la.
1451 5. configure.in no longer needs to recognize Cygwin specially.
1453 6. A problem in pcre.in for Windows platforms is fixed.
1455 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1456 pcretest, it used to include the size of the study block as part of its
1457 output. Unfortunately, the structure contains a field that has a different
1458 size on different hardware architectures. This meant that the tests that
1459 showed this size failed. As the block is currently always of a fixed size,
1460 this information isn't actually particularly useful in pcretest output, so
1461 I have just removed it.
1463 8. Three pre-processor statements accidentally did not start in column 1.
1464 Sadly, there are *still* compilers around that complain, even though
1465 standard C has not required this for well over a decade. Sigh.
1467 9. In pcretest, the code for checking callouts passed small integers in the
1468 callout_data field, which is a void * field. However, some picky compilers
1469 complained about the casts involved for this on 64-bit systems. Now
1470 pcretest passes the address of the small integer instead, which should get
1471 rid of the warnings.
1473 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1474 both compile and run time, and gives an error if an invalid UTF-8 sequence
1475 is found. There is a option for disabling this check in cases where the
1476 string is known to be correct and/or the maximum performance is wanted.
1478 11. In response to a bug report, I changed one line in Makefile.in from
1480 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1481 to
1482 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1484 to look similar to other lines, but I have no way of telling whether this
1485 is the right thing to do, as I do not use Windows. No doubt I'll get told
1486 if it's wrong...
1489 Version 4.3 21-May-03
1490 ---------------------
1492 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1493 Makefile.
1495 2. Some refactoring to improve the quality of the code:
1497 (i) The utf8_table... variables are now declared "const".
1499 (ii) The code for \cx, which used the "case flipping" table to upper case
1500 lower case letters, now just substracts 32. This is ASCII-specific,
1501 but the whole concept of \cx is ASCII-specific, so it seems
1502 reasonable.
1504 (iii) PCRE was using its character types table to recognize decimal and
1505 hexadecimal digits in the pattern. This is silly, because it handles
1506 only 0-9, a-f, and A-F, but the character types table is locale-
1507 specific, which means strange things might happen. A private
1508 table is now used for this - though it costs 256 bytes, a table is
1509 much faster than multiple explicit tests. Of course, the standard
1510 character types table is still used for matching digits in subject
1511 strings against \d.
1513 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1514 ending in _t are). So I've renamed it as ESC_tee.
1516 3. The first argument for regexec() in the POSIX wrapper should have been
1517 defined as "const".
1519 4. Changed pcretest to use malloc() for its buffers so that they can be
1520 Electric Fenced for debugging.
1522 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1523 to read one or more bytes before the start of the subject string. Often this
1524 had no effect on PCRE's behaviour, but in some circumstances it could
1525 provoke a segmentation fault.
1527 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1528 to try to read one or more bytes before the start of the subject string.
1530 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1531 UTF-8 support could misbehave in various ways if the subject string
1532 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1533 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1534 back over UTF-8 characters.)
1537 Version 4.2 14-Apr-03
1538 ---------------------
1540 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1542 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1543 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1544 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1545 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1546 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1548 Note: automatic setting of the BUILD variables is not yet working
1549 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1550 compile-time but not at link-time
1551 [LINK]: use for linking executables only
1552 make different versions for Windows and non-Windows
1553 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1554 libraries
1555 [LINK_FOR_BUILD]: new variable
1556 [OBJEXT]: use throughout
1557 [EXEEXT]: use throughout
1558 <winshared>: new target
1559 <wininstall>: new target
1560 <dftables.o>: use native compiler
1561 <dftables>: use native linker
1562 <install>: handle Windows platform correctly
1563 <clean>: ditto
1564 <check>: ditto
1565 copy DLL to top builddir before testing
1567 As part of these changes, -no-undefined was removed again. This was reported
1568 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1569 in any case.
1571 3. Some tidies to get rid of compiler warnings:
1573 . In the match_data structure, match_limit was an unsigned long int, whereas
1574 match_call_count was an int. I've made them both unsigned long ints.
1576 . In pcretest the fact that a const uschar * doesn't automatically cast to
1577 a void * provoked a warning.
1579 . Turning on some more compiler warnings threw up some "shadow" variables
1580 and a few more missing casts.
1582 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1583 option, a class that contained a single character with a value between 128
1584 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1586 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1587 option, a class that contained several characters, but with at least one
1588 whose value was between 128 and 255 caused PCRE to crash.
1591 Version 4.1 12-Mar-03
1592 ---------------------
1594 1. Compiling with gcc -pedantic found a couple of places where casts were
1595 needed, and a string in dftables.c that was longer than standard compilers are
1596 required to support.
1598 2. Compiling with Sun's compiler found a few more places where the code could
1599 be tidied up in order to avoid warnings.
1601 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1602 first of these names is deprecated in the latest Autoconf in favour of the name
1603 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1604 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1605 analogy I have changed it to CFLAGS_FOR_BUILD.
1607 4. Added -no-undefined to the linking command in the Makefile, because this is
1608 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1609 linking step for the pcreposix library.
1611 5. PCRE was failing to diagnose the case of two named groups with the same
1612 name.
1614 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1615 literal character that is needed in the subject for a match, and scans along to
1616 ensure that it is present before embarking on the full matching process. This
1617 saves time in cases of nested unlimited repeats that are never going to match.
1618 Problem: the scan can take a lot of time if the subject is very long (e.g.
1619 megabytes), thus penalizing straightforward matches. It is now done only if the
1620 amount of subject to be scanned is less than 1000 bytes.
1622 7. A lesser problem with the same optimization is that it was recording the
1623 first character of an anchored pattern as "needed", thus provoking a search
1624 right along the subject, even when the first match of the pattern was going to
1625 fail. The "needed" character is now not set for anchored patterns, unless it
1626 follows something in the pattern that is of non-fixed length. Thus, it still
1627 fulfils its original purpose of finding quick non-matches in cases of nested
1628 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1631 Version 4.0 17-Feb-03
1632 ---------------------
1634 1. If a comment in an extended regex that started immediately after a meta-item
1635 extended to the end of string, PCRE compiled incorrect data. This could lead to
1636 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1638 2. Moved to autoconf 2.53 and libtool 1.4.2.
1640 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1641 the special perltest8 script is no longer needed - all the tests can be run
1642 from a single perltest script.
1644 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1645 by \s. It has now been removed in PCRE. This means it isn't recognized as
1646 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1647 class [:space:] *does* include VT, thereby creating a mess.
1649 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1650 space and tab.
1652 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1653 its new features into the main test script, reducing the number of scripts.
1655 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1656 were backward compatible, and made the (?i) apply to the whole pattern, as if
1657 /i were given. Now it behaves more logically, and applies the option setting
1658 only to what follows. PCRE has been changed to follow suit. However, if it
1659 finds options settings right at the start of the pattern, it extracts them into
1660 the global options, as before. Thus, they show up in the info data.
1662 8. Added support for the \Q...\E escape sequence. Characters in between are
1663 treated as literals. This is slightly different from Perl in that $ and @ are
1664 also handled as literals inside the quotes. In Perl, they will cause variable
1665 interpolation. Note the following examples:
1667 Pattern PCRE matches Perl matches
1669 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1670 \Qabc\$xyz\E abc\$xyz abc\$xyz
1671 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1673 For compatibility with Perl, \Q...\E sequences are recognized inside character
1674 classes as well as outside them.
1676 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1677 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1678 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1679 signed/unsigned warnings.
1681 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1682 option for pcretest, so I've replaced it by a simple function that does just
1683 that job.
1685 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1686 "pcregrep -".
1688 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1689 Java package. This provides some syntactic sugar for simple cases of what my
1690 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1691 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1692 item, you can use this simplified notation. Note that only makes sense with
1693 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1694 greediness, whatever the setting of the PCRE_UNGREEDY option.
1696 13. A change of greediness default within a pattern was not taking effect at
1697 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1698 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1699 was abstracted outside.
1701 14. PCRE now supports the \G assertion. It is true when the current matching
1702 position is at the start point of the match. This differs from \A when the
1703 starting offset is non-zero. Used with the /g option of pcretest (or similar
1704 code), it works in the same way as it does for Perl's /g option. If all
1705 alternatives of a regex begin with \G, the expression is anchored to the start
1706 match position, and the "anchored" flag is set in the compiled expression.
1708 15. Some bugs concerning the handling of certain option changes within patterns
1709 have been fixed. These applied to options other than (?ims). For example,
1710 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1711 been the other way round. Some of this was related to change 7 above.
1713 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1714 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1715 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1716 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1718 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1719 mode. Unlike ".", it always matches newline, whatever the setting of
1720 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1721 assertions. Perl allows it, but it doesn't (in general) work because it can't
1722 calculate the length of the lookbehind. At least, that's the case for Perl
1723 5.8.0 - I've been told they are going to document that it doesn't work in
1724 future.
1726 18. Added an error diagnosis for escapes that PCRE does not support: these are
1727 \L, \l, \N, \P, \p, \U, \u, and \X.
1729 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1730 reading past the end of the pattern in cases such as /[abcd/.
1732 20. PCRE was getting more memory than necessary for patterns with classes that
1733 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1735 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1736 compiling PCRE for use with Virtual Pascal.
1738 22. Small fix to the Makefile to make it work properly if the build is done
1739 outside the source tree.
1741 23. Added a new extension: a condition to go with recursion. If a conditional
1742 subpattern starts with (?(R) the "true" branch is used if recursion has
1743 happened, whereas the "false" branch is used only at the top level.
1745 24. When there was a very long string of literal characters (over 255 bytes
1746 without UTF support, over 250 bytes with UTF support), the computation of how
1747 much memory was required could be incorrect, leading to segfaults or other
1748 strange effects.
1750 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1751 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1752 there was a subsequent back reference to those brackets. This meant that, for
1753 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1754 possible to check for precisely this case. All we can do is abandon the
1755 optimization if .* occurs inside capturing brackets when there are any back
1756 references whatsoever. (See below for a better fix that came later.)
1758 26. The handling of the optimization for finding the first character of a
1759 non-anchored pattern, and for finding a character that is required later in the
1760 match were failing in some cases. This didn't break the matching; it just
1761 failed to optimize when it could. The way this is done has been re-implemented.
1763 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1765 28. Added a new feature that provides some of the functionality that Perl
1766 provides with (?{...}). The facility is termed a "callout". The way it is done
1767 in PCRE is for the caller to provide an optional function, by setting
1768 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1769 global variable. By default it is unset, which disables all calling out. To get
1770 the function called, the regex must include (?C) at appropriate points. This
1771 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1772 This provides a means of identifying different callout points. When PCRE
1773 reaches such a point in the regex, if pcre_callout has been set, the external
1774 function is called. It is provided with data in a structure called
1775 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1776 matching continues; if it returns a non-zero value, the match at the current
1777 point fails. However, backtracking will occur if possible. [This was changed
1778 later and other features added - see item 49 below.]
1780 29. pcretest is upgraded to test the callout functionality. It provides a
1781 callout function that displays information. By default, it shows the start of
1782 the match and the current position in the text. There are some new data escapes
1783 to vary what happens:
1785 \C+ in addition, show current contents of captured substrings
1786 \C- do not supply a callout function
1787 \C!n return 1 when callout number n is reached
1788 \C!n!m return 1 when callout number n is reached for the mth time
1790 30. If pcregrep was called with the -l option and just a single file name, it
1791 output "<stdin>" if a match was found, instead of the file name.
1793 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1794 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1795 pcre_exec(). This saves a malloc/free per call. The default value of
1796 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1797 when configuring.
1799 32. The default maximum size of a compiled pattern is 64K. There have been a
1800 few cases of people hitting this limit. The code now uses macros to handle the
1801 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1802 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1803 configuring. Tests 2 and 5 work only with 2-byte links because they output
1804 debugging information about compiled patterns.
1806 33. Internal code re-arrangements:
1808 (a) Moved the debugging function for printing out a compiled regex into
1809 its own source file (printint.c) and used #include to pull it into
1810 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1811 separate copies.
1813 (b) Defined the list of op-code names for debugging as a macro in
1814 internal.h so that it is next to the definition of the opcodes.
1816 (c) Defined a table of op-code lengths for simpler skipping along compiled
1817 code. This is again a macro in internal.h so that it is next to the
1818 definition of the opcodes.
1820 34. Added support for recursive calls to individual subpatterns, along the
1821 lines of Robin Houston's patch (but implemented somewhat differently).
1823 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1824 allow it to read and process whole directories in Win32. This code was
1825 contributed by Lionel Fourquaux; it has not been tested by me.
1827 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1828 used to name a group. Names consist of alphanumerics and underscores, and must
1829 be unique. Back references use the syntax (?P=name) and recursive calls use
1830 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1831 numbers. The function pcre_fullinfo() can be used after compilation to extract
1832 a name/number map. There are three relevant calls:
1834 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1835 PCRE_INFO_NAMECOUNT yields the number of entries
1836 PCRE_INFO_NAMETABLE yields a pointer to the map.
1838 The map is a vector of fixed-size entries. The size of each entry depends on
1839 the length of the longest name used. The first two bytes of each entry are the
1840 group number, most significant byte first. There follows the corresponding
1841 name, zero terminated. The names are in alphabetical order.
1843 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1844 case instead of 255. Making it the same both with and without UTF-8 support
1845 means that the same test output works with both.
1847 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1848 calling malloc() with a zero argument.
1850 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1851 optimization. I've improved things by keeping a bitmap of backreferences with
1852 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1853 fact referenced, the optimization can be applied. It is unlikely that a
1854 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1855 the match to follow \n) will appear inside brackets with a number greater than
1856 31, but if it does, any back reference > 31 suppresses the optimization.
1858 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1859 of disabling numbered capturing parentheses. Any opening parenthesis that is
1860 not followed by ? behaves as if it were followed by ?: but named parentheses
1861 can still be used for capturing (and they will acquire numbers in the usual
1862 way).
1864 41. Redesigned the return codes from the match() function into yes/no/error so
1865 that errors can be passed back from deep inside the nested calls. A malloc
1866 failure while inside a recursive subpattern call now causes the
1867 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1869 42. It is now possible to set a limit on the number of times the match()
1870 function is called in a call to pcre_exec(). This facility makes it possible to
1871 limit the amount of recursion and backtracking, though not in a directly
1872 obvious way, because the match() function is used in a number of different
1873 circumstances. The count starts from zero for each position in the subject
1874 string (for non-anchored patterns). The default limit is, for compatibility, a
1875 large number, namely 10 000 000. You can change this in two ways:
1877 (a) When configuring PCRE before making, you can use --with-match-limit=n
1878 to set a default value for the compiled library.
1880 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1881 a different value is set. See 45 below.
1883 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1885 43. Added a new function pcre_config(int, void *) to enable run-time extraction
1886 of things that can be changed at compile time. The first argument specifies
1887 what is wanted and the second points to where the information is to be placed.
1888 The current list of available information is:
1892 The output is an integer that is set to one if UTF-8 support is available;
1893 otherwise it is set to zero.
1897 The output is an integer that it set to the value of the code that is used for
1898 newline. It is either LF (10) or CR (13).
1902 The output is an integer that contains the number of bytes used for internal
1903 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1907 The output is an integer that contains the threshold above which the POSIX
1908 interface uses malloc() for output vectors. See item 31 above.
1912 The output is an unsigned integer that contains the default limit of the number
1913 of match() calls in a pcre_exec() execution. See 42 above.
1915 44. pcretest has been upgraded by the addition of the -C option. This causes it
1916 to extract all the available output from the new pcre_config() function, and to
1917 output it. The program then exits immediately.
1919 45. A need has arisen to pass over additional data with calls to pcre_exec() in
1920 order to support additional features. One way would have been to define
1921 pcre_exec2() (for example) with extra arguments, but this would not have been
1922 extensible, and would also have required all calls to the original function to
1923 be mapped to the new one. Instead, I have chosen to extend the mechanism that
1924 is used for passing in "extra" data from pcre_study().
1926 The pcre_extra structure is now exposed and defined in pcre.h. It currently
1927 contains the following fields:
1929 flags a bitmap indicating which of the following fields are set
1930 study_data opaque data from pcre_study()
1931 match_limit a way of specifying a limit on match() calls for a specific
1932 call to pcre_exec()
1933 callout_data data for callouts (see 49 below)
1935 The flag bits are also defined in pcre.h, and are
1941 The pcre_study() function now returns one of these new pcre_extra blocks, with
1942 the actual study data pointed to by the study_data field, and the
1943 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1944 before. That is, this change is entirely upwards-compatible and requires no
1945 change to existing code.
1947 If you want to pass in additional data to pcre_exec(), you can either place it
1948 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1949 block.
1951 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1952 data string contains the escape sequence \M, pcretest calls pcre_exec() several
1953 times with different match limits, until it finds the minimum value needed for
1954 pcre_exec() to complete. The value is then output. This can be instructive; for
1955 most simple matches the number is quite small, but for pathological cases it
1956 gets very large very quickly.
1958 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1959 returns the size of the data block pointed to by the study_data field in a
1960 pcre_extra block, that is, the value that was passed as the argument to
1961 pcre_malloc() when PCRE was getting memory in which to place the information
1962 created by pcre_study(). The fourth argument should point to a size_t variable.
1963 pcretest has been extended so that this information is shown after a successful
1964 pcre_study() call when information about the compiled regex is being displayed.
1966 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1967 because what follows is always an absolute path. (Later: it turns out that this
1968 is more than cosmetic for MinGW, because it doesn't like empty path
1969 components.)
1971 49. Some changes have been made to the callout feature (see 28 above):
1973 (i) A callout function now has three choices for what it returns:
1975 0 => success, carry on matching
1976 > 0 => failure at this point, but backtrack if possible
1977 < 0 => serious error, return this value from pcre_exec()
1979 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1980 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1981 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1982 use by callout functions. It will never be used by PCRE itself.
1984 (ii) The pcre_extra structure (see 45 above) has a void * field called
1985 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1986 pcre_callout_block structure has a field of the same name. The contents of
1987 the field passed in the pcre_extra structure are passed to the callout
1988 function in the corresponding field in the callout block. This makes it
1989 easier to use the same callout-containing regex from multiple threads. For
1990 testing, the pcretest program has a new data escape
1992 \C*n pass the number n (may be negative) as callout_data
1994 If the callout function in pcretest receives a non-zero value as
1995 callout_data, it returns that value.
1997 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1998 there were some redundant $(CFLAGS) in commands that are now specified as
1999 $(LINK), which already includes $(CFLAGS).
2001 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2002 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2003 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2004 one-byte characters throughout. Note that case-insensitive matching applies
2005 only to characters whose values are less than 256. PCRE doesn't support the
2006 notion of cases for higher-valued characters.
2008 (i) A character class whose characters are all within 0-255 is handled as
2009 a bit map, and the map is inverted for negative classes. Previously, a
2010 character > 255 always failed to match such a class; however it should
2011 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2013 (ii) A negated character class with a single character < 255 is coded as
2014 "not this character" (OP_NOT). This wasn't working properly when the test
2015 character was multibyte, either singly or repeated.
2017 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2018 mode, for example: \x{100}{2,3}.
2020 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2021 singly or repeated) now correctly test multibyte characters. However,
2022 PCRE doesn't recognize any characters with values greater than 255 as
2023 digits, spaces, or word characters. Such characters always match \D, \S,
2024 and \W, and never match \d, \s, or \w.
2026 (v) Classes may now contain characters and character ranges with values
2027 greater than 255. For example: [ab\x{100}-\x{400}].
2029 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2030 PCRE in UTF-8 mode.
2032 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2033 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2034 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2035 value.)
2037 53. The single man page has become too large. I have therefore split it up into
2038 a number of separate man pages. These also give rise to individual HTML pages;
2039 these are now put in a separate directory, and there is an index.html page that
2040 lists them all. Some hyperlinking between the pages has been installed.
2042 54. Added convenience functions for handling named capturing parentheses.
2044 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2045 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2046 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2047 are faulted.
2049 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2050 calling configure. These values are used when compiling the dftables.c program
2051 which is run to generate the source of the default character tables. They
2052 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2053 you will need to set these values.
2055 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2058 Version 3.9 02-Jan-02
2059 ---------------------
2061 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2063 2. If --disable-static was given, the building process failed when trying to
2064 build pcretest and pcregrep. (For some reason it was using libtool to compile
2065 them, which is not right, as they aren't part of the library.)
2068 Version 3.8 18-Dec-01
2069 ---------------------
2071 1. The experimental UTF-8 code was completely screwed up. It was packing the
2072 bytes in the wrong order. How dumb can you get?
2075 Version 3.7 29-Oct-01
2076 ---------------------
2078 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2079 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2080 this didn't happen under Solaris 8, where I normally test things.
2082 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2083 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2084 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2085 this fix an hour or so after the initial 3.7 release.)
2088 Version 3.6 23-Oct-01
2089 ---------------------
2091 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2092 offsets passed as NULL with zero offset count.
2094 2. The config.guess and config.sub files had not been updated when I moved to
2095 the latest autoconf.
2098 Version 3.5 15-Aug-01
2099 ---------------------
2101 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2102 had been forgotten.
2104 2. By using declared but undefined structures, we can avoid using "void"
2105 definitions in pcre.h while keeping the internal definitions of the structures
2106 private.
2108 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2109 user point of view, this means that both static and shared libraries are built
2110 by default, but this can be individually controlled. More of the work of
2111 handling this static/shared cases is now inside libtool instead of PCRE's make
2112 file.
2114 4. The pcretest utility is now installed along with pcregrep because it is
2115 useful for users (to test regexs) and by doing this, it automatically gets
2116 relinked by libtool. The documentation has been turned into a man page, so
2117 there are now .1, .txt, and .html versions in /doc.
2119 5. Upgrades to pcregrep:
2120 (i) Added long-form option names like gnu grep.
2121 (ii) Added --help to list all options with an explanatory phrase.
2122 (iii) Added -r, --recursive to recurse into sub-directories.
2123 (iv) Added -f, --file to read patterns from a file.
2125 6. pcre_exec() was referring to its "code" argument before testing that
2126 argument for NULL (and giving an error if it was NULL).
2128 7. Upgraded Makefile.in to allow for compiling in a different directory from
2129 the source directory.
2131 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2132 options bits, the pointer it was passed was to an int instead of to an unsigned
2133 long int. This mattered only on 64-bit systems.
2135 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2136 generated) instead of pcre.in, which it its source. Also made the same change
2137 in several of the .c files.
2139 10. A new release of gcc defines printf() as a macro, which broke pcretest
2140 because it had an ifdef in the middle of a string argument for printf(). Fixed
2141 by using separate calls to printf().
2143 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2144 script, to force use of CR or LF instead of \n in the source. On non-Unix
2145 systems, the value can be set in config.h.
2147 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2148 absolute limit. Changed the text of the error message to make this clear, and
2149 likewise updated the man page.
2151 13. The limit of 99 on the number of capturing subpatterns has been removed.
2152 The new limit is 65535, which I hope will not be a "real" limit.
2155 Version 3.4 22-Aug-00
2156 ---------------------
2158 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2160 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2163 Version 3.3 01-Aug-00
2164 ---------------------
2166 1. If an octal character was given, but the value was greater than \377, it
2167 was not getting masked to the least significant bits, as documented. This could
2168 lead to crashes in some systems.
2170 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2171 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2173 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2174 These just pass their arguments on to (pcre_free)(), but they are provided
2175 because some uses of PCRE bind it to non-C systems that can call its functions,
2176 but cannot call free() or pcre_free() directly.
2178 4. Add "make test" as a synonym for "make check". Corrected some comments in
2179 the Makefile.
2181 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2182 Makefile.
2184 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2185 command called pgrep for grepping around the active processes.
2187 7. Added the beginnings of support for UTF-8 character strings.
2189 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2190 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2191 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2192 out for the ar command.)
2195 Version 3.2 12-May-00
2196 ---------------------
2198 This is purely a bug fixing release.
2200 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2201 of ZA. This was just one example of several cases that could provoke this bug,
2202 which was introduced by change 9 of version 2.00. The code for breaking
2203 infinite loops after an iteration that matches an empty string was't working
2204 correctly.
2206 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2207 when matched against abbab (for example). After matching an empty string, it
2208 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2209 caused it to match further down the string than it should.
2211 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2212 was there because it doesn't seem to be needed, and it causes trouble on some
2213 systems, as it is not a Standard C header. It has been removed.
2215 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2216 were reported on the Macintosh. The changes were from
2218 while ((c = *(++ptr)) != 0 && c != '\n');
2219 to
2220 while ((c = *(++ptr)) != 0 && c != '\n') ;
2222 Totally extraordinary, but if that's what it takes...
2224 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2225 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2226 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2227 assumes the way PCRE uses memmove() (always moving upwards).
2229 6. PCRE is being used in one environment where strchr() is not available. There
2230 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2231 faster code anyway.
2234 Version 3.1 09-Feb-00
2235 ---------------------
2237 The only change in this release is the fixing of some bugs in Makefile.in for
2238 the "install" target:
2240 (1) It was failing to install pcreposix.h.
2242 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2245 Version 3.0 01-Feb-00
2246 ---------------------
2248 1. Add support for the /+ modifier to perltest (to output $` like it does in
2249 pcretest).
2251 2. Add support for the /g modifier to perltest.
2253 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2254 matches null strings.
2256 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2257 pattern. Perl treats empty patterns specially - it reuses the most recent
2258 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2259 effect.
2261 5. The POSIX interface was broken in that it was just handing over the POSIX
2262 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2263 required a bigger vector, with some working space on the end. This means that
2264 the POSIX wrapper now has to get and free some memory, and copy the results.
2266 6. Added some simple autoconf support, placing the test data and the
2267 documentation in separate directories, re-organizing some of the
2268 information files, and making it build pcre-config (a GNU standard). Also added
2269 libtool support for building PCRE as a shared library, which is now the
2270 default.
2272 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2273 09 are not valid octal constants. Single digits will be used for minor values
2274 less than 10.
2276 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2277 existing programs that set these in the POSIX interface can use PCRE without
2278 modification.
2280 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2281 return all that pcre_info() returns, plus additional data. The pcre_info()
2282 function is retained for compatibility, but is considered to be obsolete.
2284 10. Added experimental recursion feature (?R) to handle one common case that
2285 Perl 5.6 will be able to do with (?p{...}).
2287 11. Added support for POSIX character classes like [:alpha:], which Perl is
2288 adopting.
2291 Version 2.08 31-Aug-99
2292 ----------------------
2294 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2295 trying to match at the startoffset position, but instead was moving forward to
2296 the next newline as if a previous match had failed.
2298 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2299 and could get into a loop if a null string was matched other than at the start
2300 of the subject.
2302 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2303 be distinguished at compile time, and for completeness also added PCRE_DATE.
2305 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2306 in GnuWin32 environments.
2309 Version 2.07 29-Jul-99
2310 ----------------------
2312 1. The documentation is now supplied in plain text form and HTML as well as in
2313 the form of man page sources.
2315 2. C++ compilers don't like assigning (void *) values to other pointer types.
2316 In particular this affects malloc(). Although there is no problem in Standard
2317 C, I've put in casts to keep C++ compilers happy.
2319 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2320 should be (const char *).
2322 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2323 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2324 However, I haven't made this a standard facility. The documentation doesn't
2325 mention it, and the Makefile doesn't support it.
2327 5. The Makefile now contains an "install" target, with editable destinations at
2328 the top of the file. The pcretest program is not installed.
2330 6. pgrep -V now gives the PCRE version number and date.
2332 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2333 causing the entire string to be ignored, instead of just the last character.
2335 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2336 non-matching string, it can take a very, very long time, even for strings of
2337 quite modest length, because of the nested recursion. PCRE now does better in
2338 some of these cases. It does this by remembering the last required literal
2339 character in the pattern, and pre-searching the subject to ensure it is present
2340 before running the real match. In other words, it applies a heuristic to detect
2341 some types of certain failure quickly, and in the above example, if presented
2342 with a string that has no trailing " it gives "no match" very quickly.
2344 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2345 other alternatives are tried instead.
2348 Version 2.06 09-Jun-99
2349 ----------------------
2351 1. Change pcretest's output for amount of store used to show just the code
2352 space, because the remainder (the data block) varies in size between 32-bit and
2353 64-bit systems.
2355 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2356 start matching at. This allows lookbehinds to work when searching for multiple
2357 occurrences in a string.
2359 3. Added additional options to pcretest for testing multiple occurrences:
2361 /+ outputs the rest of the string that follows a match
2362 /g loops for multiple occurrences, using the new startoffset argument
2363 /G loops for multiple occurrences by passing an incremented pointer
2365 4. PCRE wasn't doing the "first character" optimization for patterns starting
2366 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2367 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2368 the letter 'x'. On long subject strings, this gives a significant speed-up.
2371 Version 2.05 21-Apr-99
2372 ----------------------
2374 1. Changed the type of magic_number from int to long int so that it works
2375 properly on 16-bit systems.
2377 2. Fixed a bug which caused patterns starting with .* not to work correctly
2378 when the subject string contained newline characters. PCRE was assuming
2379 anchoring for such patterns in all cases, which is not correct because .* will
2380 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2381 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2382 must be retried after every newline in the subject.
2385 Version 2.04 18-Feb-99
2386 ----------------------
2388 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2389 computation of the store needed to hold the pattern was incorrect (too large).
2390 If such patterns were nested a few deep, this could multiply and become a real
2391 problem.
2393 2. Added /M option to pcretest to show the memory requirement of a specific
2394 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2396 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2397 compiled in such a way that the backtracking after subsequent failure was
2398 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2399 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2402 Version 2.03 02-Feb-99
2403 ----------------------
2405 1. Fixed typo and small mistake in man page.
2407 2. Added 4th condition (GPL supersedes if conflict) and created separate
2408 LICENCE file containing the conditions.
2410 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2411 Perl, that is the internal \ allows the delimiter to be included in the
2412 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2413 the final delimiter, add \ to the end of the pattern (to test the error).
2415 4. Added the convenience functions for extracting substrings after a successful
2416 match. Updated pcretest to make it able to test these functions.
2419 Version 2.02 14-Jan-99
2420 ----------------------
2422 1. Initialized the working variables associated with each extraction so that
2423 their saving and restoring doesn't refer to uninitialized store.
2425 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2426 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2427 fix the problem.
2429 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2430 calls, and wasn't printing the correct value for compiling calls. Increased the
2431 default value of LOOPREPEAT, and the number of significant figures in the
2432 times.
2434 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2436 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2437 a building problem on Windows NT with a FAT file system.
2440 Version 2.01 21-Oct-98
2441 ----------------------
2443 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2444 to character tables built by pcre_maketables() in the current locale. If NULL
2445 is passed, the default tables are used.
2448 Version 2.00 24-Sep-98
2449 ----------------------
2451 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2452 it any more.
2454 2. Allow quantification of (?>) groups, and make it work correctly.
2456 3. The first character computation wasn't working for (?>) groups.
2458 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2459 end of the subject) and add 5.005's \z, which really does match only at the
2460 very end of the subject.
2462 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2464 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2465 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2466 localized options. All options to pcre_study() were also removed.
2468 7. Add other new features from 5.005:
2470 $(?<= positive lookbehind
2471 $(?<! negative lookbehind
2472 (?imsx-imsx) added the unsetting capability
2473 such a setting is global if at outer level; local otherwise
2474 (?imsx-imsx:) non-capturing groups with option setting
2475 (?(cond)re|re) conditional pattern matching
2477 A backreference to itself in a repeated group matches the previous
2478 captured string.
2480 8. General tidying up of studying (both automatic and via "study")
2481 consequential on the addition of new assertions.
2483 9. As in 5.005, unlimited repeated groups that could match an empty substring
2484 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2485 runtime if any iteration does actually match an empty substring.
2487 10. Include the RunTest script in the distribution.
2489 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2490 discrepancies, some of which were old and were also with respect to 5.004. They
2491 have now been fixed.
2494 Version 1.09 28-Apr-98
2495 ----------------------
2497 1. A negated single character class followed by a quantifier with a minimum
2498 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2499 program crashes, or just wrong answers. This did not apply to negated classes
2500 containing more than one character, or to minima other than one.
2503 Version 1.08 27-Mar-98
2504 ----------------------
2506 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2508 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2509 latter must appear before anything that relies on it in the pattern.
2512 Version 1.07 16-Feb-98
2513 ----------------------
2515 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2516 repeat of a potentially empty string).
2519 Version 1.06 23-Jan-98
2520 ----------------------
2522 1. Added Markus Oberhumer's little patches for C++.
2524 2. Literal strings longer than 255 characters were broken.
2527 Version 1.05 23-Dec-97
2528 ----------------------
2530 1. Negated character classes containing more than one character were failing if
2531 PCRE_CASELESS was set at run time.
2534 Version 1.04 19-Dec-97
2535 ----------------------
2537 1. Corrected the man page, where some "const" qualifiers had been omitted.
2539 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2540 input syntax.
2542 3. Fixed memory leak which occurred when a regex with back references was
2543 matched with an offsets vector that wasn't big enough. The temporary memory
2544 that is used in this case wasn't being freed if the match failed.
2546 4. Tidied pcretest to ensure it frees memory that it gets.
2548 5. Temporary memory was being obtained in the case where the passed offsets
2549 vector was exactly big enough.
2551 6. Corrected definition of offsetof() from change 5 below.
2553 7. I had screwed up change 6 below and broken the rules for the use of
2554 setjmp(). Now fixed.
2557 Version 1.03 18-Dec-97
2558 ----------------------
2560 1. A erroneous regex with a missing opening parenthesis was correctly
2561 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2562 on some systems.
2564 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2565 it was reported that one broken compiler failed on the former because "code" is
2566 also an independent variable.
2568 3. The erroneous regex a[]b caused an array overrun reference.
2570 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2571 fail on data ending with that character. (It was going on too far, and checking
2572 the next character, typically a binary zero.) This was specific to the
2573 optimized code for single-character negative classes.
2575 5. Added a contributed patch from the TIN world which does the following:
2577 + Add an undef for memmove, in case the the system defines a macro for it.
2579 + Add a definition of offsetof(), in case there isn't one. (I don't know
2580 the reason behind this - offsetof() is part of the ANSI standard - but
2581 it does no harm).
2583 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2584 most of the places where whitespace preceded '#'. I have given up and
2585 allowed the remaining 2 cases to be at the margin.
2587 + Rename some variables in pcre to eliminate shadowing. This seems very
2588 pedantic, but does no harm, of course.
2590 6. Moved the call to setjmp() into its own function, to get rid of warnings
2591 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2593 7. Constructs such as \d{8,} were compiling into the equivalent of
2594 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2595 outcome, but in this particular case used more store than had been allocated,
2596 which caused the bug to be discovered because it threw up an internal error.
2598 8. The debugging code in both pcre and pcretest for outputting the compiled
2599 form of a regex was going wrong in the case of back references followed by
2600 curly-bracketed repeats.
2603 Version 1.02 12-Dec-97
2604 ----------------------
2606 1. Typos in pcre.3 and comments in the source fixed.
2608 2. Applied a contributed patch to get rid of places where it used to remove
2609 'const' from variables, and fixed some signed/unsigned and uninitialized
2610 variable warnings.
2612 3. Added the "runtest" target to Makefile.
2614 4. Set default compiler flag to -O2 rather than just -O.
2617 Version 1.01 19-Nov-97
2618 ----------------------
2620 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2621 like /([ab]*)*/, that is, for classes with more than one character in them.
2623 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2624 as /((?>a*))*/ (a PCRE_EXTRA facility).
2627 Version 1.00 18-Nov-97
2628 ----------------------
2630 1. Added compile-time macros to support systems such as SunOS4 which don't have
2631 memmove() or strerror() but have other things that can be used instead.
2633 2. Arranged that "make clean" removes the executables.
2636 Version 0.99 27-Oct-97
2637 ----------------------
2639 1. Fixed bug in code for optimizing classes with only one character. It was
2640 initializing a 32-byte map regardless, which could cause it to run off the end
2641 of the memory it had got.
2643 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2646 Version 0.98 22-Oct-97
2647 ----------------------
2649 1. Fixed bug in code for handling temporary memory usage when there are more
2650 back references than supplied space in the ovector. This could cause segfaults.
2653 Version 0.97 21-Oct-97
2654 ----------------------
2656 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2658 2. Optimized negated single characters not to use a bit map.
2660 3. Brought error texts together as macro definitions; clarified some of them;
2661 fixed one that was wrong - it said "range out of order" when it meant "invalid
2662 escape sequence".
2664 4. Changed some char * arguments to const char *.
2666 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2668 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2669 pcretest.
2672 Version 0.96 16-Oct-97
2673 ----------------------
2675 1. Added a simple "pgrep" utility to the distribution.
2677 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2678 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2679 where "ddd" means "one or more decimal digits".
2681 3. Fixed serious bug. If a pattern had a back reference, but the call to
2682 pcre_exec() didn't supply a large enough ovector to record the related
2683 identifying subpattern, the match always failed. PCRE now remembers the number
2684 of the largest back reference, and gets some temporary memory in which to save
2685 the offsets during matching if necessary, in order to ensure that
2686 backreferences always work.
2688 4. Increased the compatibility with Perl in a number of ways:
2690 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2691 to request this handling. The option can be set at compile or exec time.
2693 (b) $ matches before a terminating newline by default; an option
2694 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2695 mode). The option can be set at compile or exec time.
2697 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2698 the same as Perl's. If the decimal number it represents is less than 10
2699 or there aren't that many previous left capturing parentheses, an octal
2700 escape is read. Inside a character class, it's always an octal escape,
2701 even if it is a single digit.
2703 (d) An escaped but undefined alphabetic character is taken as a literal,
2704 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2705 escapes.
2707 (e) {0} is now permitted. (The previous item is removed from the compiled
2708 pattern).
2710 5. Changed all the names of code files so that the basic parts are no longer
2711 than 10 characters, and abolished the teeny "globals.c" file.
2713 6. Changed the handling of character classes; they are now done with a 32-byte
2714 bit map always.
2716 7. Added the -d and /D options to pcretest to make it possible to look at the
2717 internals of compilation without having to recompile pcre.
2720 Version 0.95 23-Sep-97
2721 ----------------------
2723 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2724 \x20 at the start of a run of normal characters. These were being treated as
2725 real characters, instead of the source characters being re-checked.
2728 Version 0.94 18-Sep-97
2729 ----------------------
2731 1. The functions are now thread-safe, with the caveat that the global variables
2732 containing pointers to malloc() and free() or alternative functions are the
2733 same for all threads.
2735 2. Get pcre_study() to generate a bitmap of initial characters for non-
2736 anchored patterns when this is possible, and use it if passed to pcre_exec().
2739 Version 0.93 15-Sep-97
2740 ----------------------
2742 1. /(b)|(:+)/ was computing an incorrect first character.
2744 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2745 but not actually doing anything yet.
2747 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2748 as Perl does (e.g. [-az] or [az-]).
2750 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2751 all possible positions.
2753 5. Split up into different modules to avoid including unneeded functions in a
2754 compiled binary. However, compile and exec are still in one module. The "study"
2755 function is split off.
2757 6. The character tables are now in a separate module whose source is generated
2758 by an auxiliary program - but can then be edited by hand if required. There are
2759 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2760 toupper() in the code.
2762 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2763 make them global. Abolish the function for setting them, as the caller can now
2764 set them directly.
2767 Version 0.92 11-Sep-97
2768 ----------------------
2770 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2771 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2773 2. Caseless matching was not working in character classes if the characters in
2774 the pattern were in upper case.
2776 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2778 4. Make PCRE_ANCHORED public and accept as a compile option.
2780 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2781 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2782 pass them.
2784 6. Give an error if bad option bits passed at compile or run time.
2786 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2787 pcretest to cause it to pass that flag.
2789 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2790 options, and the first character, if set.
2792 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2795 Version 0.91 10-Sep-97
2796 ----------------------
2798 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2799 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2801 2. PCRE was looping on encountering an indefinitely repeated back reference to
2802 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2803 Perl does - treats the match as successful.
2805 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5