ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 228 - (show annotations)
Tue Aug 28 13:37:37 2007 UTC (14 years, 1 month ago) by ph10
File size: 138559 byte(s)
Final version number changes for the 7.3 release.
1 ChangeLog for PCRE
2 ------------------
4 Version 7.3 28-Aug-07
5 ---------------------
7 1. In the rejigging of the build system that eventually resulted in 7.1, the
8 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
9 brackets there is not right, since it causes compilers to look for an
10 installed pcre.h, not the version that is in the source that is being
11 compiled (which of course may be different). I have changed it back to:
13 #include "pcre.h"
15 I have a vague recollection that the change was concerned with compiling in
16 different directories, but in the new build system, that is taken care of
17 by the VPATH setting the Makefile.
19 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
20 when the subject happened to end in the byte 0x85 (e.g. if the last
21 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
22 characters but of course it shouldn't be taken as a newline when it is part
23 of another character. The bug was that, for an unlimited repeat of . in
24 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
25 characters when looking for a newline.
27 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
29 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
30 in debug output.
32 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
33 long printing in the pcrecpp unittest when running under MinGW.
35 6. ESC_K was left out of the EBCDIC table.
37 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
38 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
39 limit also applies to "virtual nesting" when a pattern is recursive, and in
40 this case 1000 isn't so big. I have been able to remove this limit at the
41 expense of backing off one optimization in certain circumstances. Normally,
42 when pcre_exec() would call its internal match() function recursively and
43 immediately return the result unconditionally, it uses a "tail recursion"
44 feature to save stack. However, when a subpattern that can match an empty
45 string has an unlimited repetition quantifier, it no longer makes this
46 optimization. That gives it a stack frame in which to save the data for
47 checking that an empty string has been matched. Previously this was taken
48 from the 1000-entry workspace that had been reserved. So now there is no
49 explicit limit, but more stack is used.
51 8. Applied Daniel's patches to solve problems with the import/export magic
52 syntax that is required for Windows, and which was going wrong for the
53 pcreposix and pcrecpp parts of the library. These were overlooked when this
54 problem was solved for the main library.
56 9. There were some crude static tests to avoid integer overflow when computing
57 the size of patterns that contain repeated groups with explicit upper
58 limits. As the maximum quantifier is 65535, the maximum group length was
59 set at 30,000 so that the product of these two numbers did not overflow a
60 32-bit integer. However, it turns out that people want to use groups that
61 are longer than 30,000 bytes (though not repeat them that many times).
62 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
63 made it possible to implement the integer overflow checks in a much more
64 dynamic way, which I have now done. The artificial limitation on group
65 length has been removed - we now have only the limit on the total length of
66 the compiled pattern, which depends on the LINK_SIZE setting.
68 10. Fixed a bug in the documentation for get/copy named substring when
69 duplicate names are permitted. If none of the named substrings are set, the
70 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
71 empty string.
73 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
74 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
75 because the ] is interpreted as the first data character and the
76 terminating ] is not found. PCRE has been made compatible with Perl in this
77 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
78 cause memory overwriting.
80 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
81 string has been matched (to stop an infinite loop). It was not recognizing
82 a conditional subpattern that could match an empty string if that
83 subpattern was within another subpattern. For example, it looped when
84 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
85 condition was not nested. This bug has been fixed.
87 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
88 past the start of the subject in the presence of bytes with the top bit
89 set, for example "\x8aBCD".
91 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
92 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
94 14. Optimized (?!) to (*FAIL).
96 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
97 This restricts code points to be within the range 0 to 0x10FFFF, excluding
98 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
99 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
100 does: it's just the validity check that is more restrictive.
102 16. Inserted checks for integer overflows during escape sequence (backslash)
103 processing, and also fixed erroneous offset values for syntax errors during
104 backslash processing.
106 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
107 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
109 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
110 caused an overrun.
112 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
113 something other than just ASCII characters) inside a group that had an
114 unlimited repeat caused a loop at compile time (while checking to see
115 whether the group could match an empty string).
117 20. Debugging a pattern containing \p or \P could cause a crash. For example,
118 [\P{Any}] did so. (Error in the code for printing property names.)
120 21. An orphan \E inside a character class could cause a crash.
122 22. A repeated capturing bracket such as (A)? could cause a wild memory
123 reference during compilation.
125 23. There are several functions in pcre_compile() that scan along a compiled
126 expression for various reasons (e.g. to see if it's fixed length for look
127 behind). There were bugs in these functions when a repeated \p or \P was
128 present in the pattern. These operators have additional parameters compared
129 with \d, etc, and these were not being taken into account when moving along
130 the compiled data. Specifically:
132 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
133 length.
135 (b) An item such as \pL+ within a repeated group could cause crashes or
136 loops.
138 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
139 "reference to non-existent subpattern" error.
141 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
143 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
144 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
146 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
148 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
149 character were causing crashes (broken optimization).
151 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
152 \p or \P) caused a compile-time loop.
154 28. More problems have arisen in unanchored patterns when CRLF is a valid line
155 break. For example, the unstudied pattern [\r\n]A does not match the string
156 "\r\nA" because change 7.0/46 below moves the current point on by two
157 characters after failing to match at the start. However, the pattern \nA
158 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
159 the same is true. There doesn't seem any very clean way out of this, but
160 what I have chosen to do makes the common cases work: PCRE now takes note
161 of whether there can be an explicit match for \r or \n anywhere in the
162 pattern, and if so, 7.0/46 no longer applies. As part of this change,
163 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
164 pattern has explicit CR or LF references.
166 29. Added (*CR) etc for changing newline setting at start of pattern.
169 Version 7.2 19-Jun-07
170 ---------------------
172 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
173 which is apparently normally available under Windows.
175 2. Re-jig the pcregrep tests with different newline settings in an attempt
176 to make them independent of the local environment's newline setting.
178 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
180 4. Some of the "internals" tests were previously cut out when the link size
181 was not 2, because the output contained actual offsets. The recent new
182 "Z" feature of pcretest means that these can be cut out, making the tests
183 usable with all link sizes.
185 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
186 stack recursion. This gives a massive performance boost under BSD, but just
187 a small improvement under Linux. However, it saves one field in the frame
188 in all cases.
190 6. Added more features from the forthcoming Perl 5.10:
192 (a) (?-n) (where n is a string of digits) is a relative subroutine or
193 recursion call. It refers to the nth most recently opened parentheses.
195 (b) (?+n) is also a relative subroutine call; it refers to the nth next
196 to be opened parentheses.
198 (c) Conditions that refer to capturing parentheses can be specified
199 relatively, for example, (?(-2)... or (?(+3)...
201 (d) \K resets the start of the current match so that everything before
202 is not part of it.
204 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
206 (f) \g{name} is another synonym - part of Perl 5.10's unification of
207 reference syntax.
209 (g) (?| introduces a group in which the numbering of parentheses in each
210 alternative starts with the same number.
212 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
214 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
217 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
218 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
219 for detecting groups that can match an empty string.
221 9. A pattern with a very large number of alternatives (more than several
222 hundred) was running out of internal workspace during the pre-compile
223 phase, where pcre_compile() figures out how much memory will be needed. A
224 bit of new cunning has reduced the workspace needed for groups with
225 alternatives. The 1000-alternative test pattern now uses 12 bytes of
226 workspace instead of running out of the 4096 that are available.
228 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
230 11. Applied patch from Google to remove an optimization that didn't quite work.
231 The report of the bug said:
233 pcrecpp::RE("a*").FullMatch("aaa") matches, while
234 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
235 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
237 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
238 it matched the wrong number of bytes.
241 Version 7.1 24-Apr-07
242 ---------------------
244 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
245 that is more "standard", making use of automake and other Autotools. There
246 is some re-arrangement of the files and adjustment of comments consequent
247 on this.
249 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
250 for recursive directory scanning broke on some systems because the files
251 are not scanned in any specific order and on different systems the order
252 was different. A call to "sort" has been inserted into RunGrepTest for the
253 approprate test as a short-term fix. In the longer term there may be an
254 alternative.
256 3. I had an email from Eric Raymond about problems translating some of PCRE's
257 man pages to HTML (despite the fact that I distribute HTML pages, some
258 people do their own conversions for various reasons). The problems
259 concerned the use of low-level troff macros .br and .in. I have therefore
260 removed all such uses from the man pages (some were redundant, some could
261 be replaced by .nf/.fi pairs). The 132html script that I use to generate
262 HTML has been updated to handle .nf/.fi and to complain if it encounters
263 .br or .in.
265 4. Updated comments in configure.ac that get placed in config.h.in and also
266 arranged for config.h to be included in the distribution, with the name
267 config.h.generic, for the benefit of those who have to compile without
268 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
270 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
271 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
272 makevp.bat for the current PCRE, using the additional files
273 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
275 6. A Windows user reported a minor discrepancy with test 2, which turned out
276 to be caused by a trailing space on an input line that had got lost in his
277 copy. The trailing space was an accident, so I've just removed it.
279 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
280 that is needed.
282 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
283 as "const" (a) because they are and (b) because it helps the PHP
284 maintainers who have recently made a script to detect big data structures
285 in the php code that should be moved to the .rodata section. I remembered
286 to update Builducptable as well, so it won't revert if ucptable.h is ever
287 re-created.
289 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
290 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
291 order to be able to cut out the UTF-8 tables in the latter when UTF-8
292 support is not required. This saves 1.5-2K of code, which is important in
293 some applications.
295 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
296 so as not to refer to the tables, even though these functions will never be
297 called when UTF-8 support is disabled. Otherwise there are problems with a
298 shared library.
300 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
302 (a) It was defining its arguments as char * instead of void *.
304 (b) It was assuming that all moves were upwards in memory; this was true
305 a long time ago when I wrote it, but is no longer the case.
307 The emulated memove() is provided for those environments that have neither
308 memmove() nor bcopy(). I didn't think anyone used it these days, but that
309 is clearly not the case, as these two bugs were recently reported.
311 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
312 and Detrail to create the HTML documentation, the .txt form of the man
313 pages, and it removes trailing spaces from listed files. It also creates
314 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
315 case, it wraps all the #defines with #ifndefs. This script should be run
316 before "make dist".
318 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
319 with Unicode property support.
321 (a) For a maximizing quantifier, if the two different cases of the
322 character were of different lengths in their UTF-8 codings (there are
323 some cases like this - I found 11), and the matching function had to
324 back up over a mixture of the two cases, it incorrectly assumed they
325 were both the same length.
327 (b) When PCRE was configured to use the heap rather than the stack for
328 recursion during matching, it was not correctly preserving the data for
329 the other case of a UTF-8 character when checking ahead for a match
330 while processing a minimizing repeat. If the check also involved
331 matching a wide character, but failed, corruption could cause an
332 erroneous result when trying to check for a repeat of the original
333 character.
335 13. Some tidying changes to the testing mechanism:
337 (a) The RunTest script now detects the internal link size and whether there
338 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
339 values substituted by "configure". (The RunGrepTest script already did
340 this for UTF-8.) The configure.ac script no longer substitutes the
341 relevant variables.
343 (b) The debugging options /B and /D in pcretest show the compiled bytecode
344 with length and offset values. This means that the output is different
345 for different internal link sizes. Test 2 is skipped for link sizes
346 other than 2 because of this, bypassing the problem. Unfortunately,
347 there was also a test in test 3 (the locale tests) that used /B and
348 failed for link sizes other than 2. Rather than cut the whole test out,
349 I have added a new /Z option to pcretest that replaces the length and
350 offset values with spaces. This is now used to make test 3 independent
351 of link size. (Test 2 will be tidied up later.)
353 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
354 segmentation fault instead of returning the appropriate error message.
356 15. In multiline mode when the newline sequence was set to "any", the pattern
357 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
358 This doesn't seem right; it now treats the CRLF combination as the line
359 ending, and so does not match in that case. It's only a pattern such as ^$
360 that would hit this one: something like ^ABC$ would have failed after \r
361 and then tried again after \r\n.
363 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
364 in an attempt to make files that differ only in their line terminators
365 compare equal. This works on Linux.
367 17. Under certain error circumstances pcregrep might try to free random memory
368 as it exited. This is now fixed, thanks to valgrind.
370 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
371 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
372 was because its rules for how to advance for /g after matching an empty
373 string at the end of a line did not allow for this case. They now check for
374 it specially.
376 20. pcretest is supposed to handle patterns and data of any length, by
377 extending its buffers when necessary. It was getting this wrong when the
378 buffer for a data line had to be extended.
380 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
381 CRLF as a newline sequence.
383 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
384 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
385 I have nevertheless tidied it up.
387 23. Added some casts to kill warnings from HP-UX ia64 compiler.
389 24. Added a man page for pcre-config.
392 Version 7.0 19-Dec-06
393 ---------------------
395 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
396 moving to gcc 4.1.1.
398 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
399 sys/time.h, which is documented as needed for this function. It doesn't
400 seem to matter on Linux, but it showed up on some releases of OS X.
402 3. It seems that there are systems where bytes whose values are greater than
403 127 match isprint() in the "C" locale. The "C" locale should be the
404 default when a C program starts up. In most systems, only ASCII printing
405 characters match isprint(). This difference caused the output from pcretest
406 to vary, making some of the tests fail. I have changed pcretest so that:
408 (a) When it is outputting text in the compiled version of a pattern, bytes
409 other than 32-126 are always shown as hex escapes.
411 (b) When it is outputting text that is a matched part of a subject string,
412 it does the same, unless a different locale has been set for the match
413 (using the /L modifier). In this case, it uses isprint() to decide.
415 4. Fixed a major bug that caused incorrect computation of the amount of memory
416 required for a compiled pattern when options that changed within the
417 pattern affected the logic of the preliminary scan that determines the
418 length. The relevant options are -x, and -i in UTF-8 mode. The result was
419 that the computed length was too small. The symptoms of this bug were
420 either the PCRE error "internal error: code overflow" from pcre_compile(),
421 or a glibc crash with a message such as "pcretest: free(): invalid next
422 size (fast)". Examples of patterns that provoked this bug (shown in
423 pcretest format) are:
425 /(?-x: )/x
426 /(?x)(?-x: \s*#\s*)/
427 /((?i)[\x{c0}])/8
428 /(?i:[\x{c0}])/8
430 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
431 is now done differently.
433 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
434 wrapper classes; (b) implement a new function in the C++ scanner that is
435 more efficient than the old way of doing things because it avoids levels of
436 recursion in the regex matching; (c) add a paragraph to the documentation
437 for the FullMatch() function.
439 6. The escape sequence \n was being treated as whatever was defined as
440 "newline". Not only was this contrary to the documentation, which states
441 that \n is character 10 (hex 0A), but it also went horribly wrong when
442 "newline" was defined as CRLF. This has been fixed.
444 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
445 was being set to -1 for the "end of line" case (supposedly a value that no
446 character can have). Though this value is never used (the check for end of
447 line is "zero bytes in current character"), it caused compiler complaints.
448 I've changed it to 0xffffffff.
450 8. In pcre_version.c, the version string was being built by a sequence of
451 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
452 string (as it is for production releases) called a macro with an empty
453 argument. The C standard says the result of this is undefined. The gcc
454 compiler treats it as an empty string (which was what was wanted) but it is
455 reported that Visual C gives an error. The source has been hacked around to
456 avoid this problem.
458 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
459 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
460 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
461 of them did).
463 10. Originally, pcretest opened its input and output without "b"; then I was
464 told that "b" was needed in some environments, so it was added for release
465 5.0 to both the input and output. (It makes no difference on Unix-like
466 systems.) Later I was told that it is wrong for the input on Windows. I've
467 now abstracted the modes into two macros, to make it easier to fiddle with
468 them, and removed "b" from the input mode under Windows.
470 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
472 12. Added -help and --help to pcretest as an official way of being reminded
473 of the options.
475 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
476 and pcrecpp.cc because they annoy compilers at high warning levels.
478 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
480 15. Fixed an occurrence of == in configure.ac that should have been = (shell
481 scripts are not C programs :-) and which was not noticed because it works
482 on Linux.
484 16. pcretest is supposed to handle any length of pattern and data line (as one
485 line or as a continued sequence of lines) by extending its input buffer if
486 necessary. This feature was broken for very long pattern lines, leading to
487 a string of junk being passed to pcre_compile() if the pattern was longer
488 than about 50K.
490 17. I have done a major re-factoring of the way pcre_compile() computes the
491 amount of memory needed for a compiled pattern. Previously, there was code
492 that made a preliminary scan of the pattern in order to do this. That was
493 OK when PCRE was new, but as the facilities have expanded, it has become
494 harder and harder to keep it in step with the real compile phase, and there
495 have been a number of bugs (see for example, 4 above). I have now found a
496 cunning way of running the real compile function in a "fake" mode that
497 enables it to compute how much memory it would need, while actually only
498 ever using a few hundred bytes of working memory and without too many
499 tests of the mode. This should make future maintenance and development
500 easier. A side effect of this work is that the limit of 200 on the nesting
501 depth of parentheses has been removed (though this was never a serious
502 limitation, I suspect). However, there is a downside: pcre_compile() now
503 runs more slowly than before (30% or more, depending on the pattern). I
504 hope this isn't a big issue. There is no effect on runtime performance.
506 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
507 newline (only possible for the last line of a file) and it was a
508 pattern that set a locale (followed by /Lsomething), pcretest crashed.
510 19. Added additional timing features to pcretest. (1) The -tm option now times
511 matching only, not compiling. (2) Both -t and -tm can be followed, as a
512 separate command line item, by a number that specifies the number of
513 repeats to use when timing. The default is 50000; this gives better
514 precision, but takes uncomfortably long for very large patterns.
516 20. Extended pcre_study() to be more clever in cases where a branch of a
517 subpattern has no definite first character. For example, (a*|b*)[cd] would
518 previously give no result from pcre_study(). Now it recognizes that the
519 first character must be a, b, c, or d.
521 21. There was an incorrect error "recursive call could loop indefinitely" if
522 a subpattern (or the entire pattern) that was being tested for matching an
523 empty string contained only one non-empty item after a nested subpattern.
524 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
525 incorrectly, because the \d was being skipped in the check.
527 22. The pcretest program now has a new pattern option /B and a command line
528 option -b, which is equivalent to adding /B to every pattern. This causes
529 it to show the compiled bytecode, without the additional information that
530 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
531 is the same as /B/I).
533 23. A new optimization is now able automatically to treat some sequences such
534 as a*b as a*+b. More specifically, if something simple (such as a character
535 or a simple class like \d) has an unlimited quantifier, and is followed by
536 something that cannot possibly match the quantified thing, the quantifier
537 is automatically "possessified".
539 24. A recursive reference to a subpattern whose number was greater than 39
540 went wrong under certain circumstances in UTF-8 mode. This bug could also
541 have affected the operation of pcre_study().
543 25. Realized that a little bit of performance could be had by replacing
544 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
546 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
548 27. Possessive quantifiers such as a++ were previously implemented by turning
549 them into atomic groups such as ($>a+). Now they have their own opcodes,
550 which improves performance. This includes the automatically created ones
551 from 23 above.
553 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
554 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
555 the first matched character to be a colon. This applied both to named and
556 numbered groups.
558 29. The ucpinternal.h header file was missing its idempotency #ifdef.
560 30. I was sent a "project" file called libpcre.a.dev which I understand makes
561 building PCRE on Windows easier, so I have included it in the distribution.
563 31. There is now a check in pcretest against a ridiculously large number being
564 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
565 loop, the loop is abandoned.
567 32. Forward references to subpatterns in conditions such as (?(2)...) where
568 subpattern 2 is defined later cause pcre_compile() to search forwards in
569 the pattern for the relevant set of parentheses. This search went wrong
570 when there were unescaped parentheses in a character class, parentheses
571 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
573 33. "Subroutine" calls and backreferences were previously restricted to
574 referencing subpatterns earlier in the regex. This restriction has now
575 been removed.
577 34. Added a number of extra features that are going to be in Perl 5.10. On the
578 whole, these are just syntactic alternatives for features that PCRE had
579 previously implemented using the Python syntax or my own invention. The
580 other formats are all retained for compatibility.
582 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
583 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
584 also .NET compatible.
586 (b) A recursion or subroutine call to a named group can now be defined as
587 (?&name) as well as (?P>name).
589 (c) A backreference to a named group can now be defined as \k<name> or
590 \k'name' as well as (?P=name). The new forms, as well as being in Perl
591 5.10, are also .NET compatible.
593 (d) A conditional reference to a named group can now use the syntax
594 (?(<name>) or (?('name') as well as (?(name).
596 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
597 groups (named and numbered) that are never evaluated inline, but can be
598 called as "subroutines" from elsewhere. In effect, the DEFINE condition
599 is always false. There may be only one alternative in such a group.
601 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
602 as the simple (?(R). The condition is true only if the most recent
603 recursion is that of the given number or name. It does not search out
604 through the entire recursion stack.
606 (g) The escape \gN or \g{N} has been added, where N is a positive or
607 negative number, specifying an absolute or relative reference.
609 35. Tidied to get rid of some further signed/unsigned compiler warnings and
610 some "unreachable code" warnings.
612 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
613 things, this adds five new scripts.
615 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
616 There were also incompatibilities regarding the handling of \Q..\E inside
617 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
618 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
620 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
621 matches an empty string, and forcibly breaks the loop. There were bugs in
622 this code in non-simple cases. For a pattern such as ^(a()*)* matched
623 against aaaa the result was just "a" rather than "aaaa", for example. Two
624 separate and independent bugs (that affected different cases) have been
625 fixed.
627 39. Refactored the code to abolish the use of different opcodes for small
628 capturing bracket numbers. This is a tidy that I avoided doing when I
629 removed the limit on the number of capturing brackets for 3.5 back in 2001.
630 The new approach is not only tidier, it makes it possible to reduce the
631 memory needed to fix the previous bug (38).
633 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
634 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
635 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
636 mode.
638 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
639 report.
641 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
642 copy construction and assignment in the C++ wrapper.
644 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
645 couple of bugs that could have given wrong results in the "--newline=crlf"
646 case.
648 44. Added a number of casts and did some reorganization of signed/unsigned int
649 variables following suggestions from Dair Grant. Also renamed the variable
650 "this" as "item" because it is a C++ keyword.
652 45. Arranged for dftables to add
654 #include "pcre_internal.h"
656 to pcre_chartables.c because without it, gcc 4.x may remove the array
657 definition from the final binary if PCRE is built into a static library and
658 dead code stripping is activated.
660 46. For an unanchored pattern, if a match attempt fails at the start of a
661 newline sequence, and the newline setting is CRLF or ANY, and the next two
662 characters are CRLF, advance by two characters instead of one.
665 Version 6.7 04-Jul-06
666 ---------------------
668 1. In order to handle tests when input lines are enormously long, pcretest has
669 been re-factored so that it automatically extends its buffers when
670 necessary. The code is crude, but this _is_ just a test program. The
671 default size has been increased from 32K to 50K.
673 2. The code in pcre_study() was using the value of the re argument before
674 testing it for NULL. (Of course, in any sensible call of the function, it
675 won't be NULL.)
677 3. The memmove() emulation function in pcre_internal.h, which is used on
678 systems that lack both memmove() and bcopy() - that is, hardly ever -
679 was missing a "static" storage class specifier.
681 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
682 containing an extended class (one that cannot be represented by a bitmap
683 because it contains high-valued characters or Unicode property items, e.g.
684 [\pZ]). Almost always one would set UTF-8 mode when processing such a
685 pattern, but PCRE should not loop if you do not (it no longer does).
686 [Detail: two cases were found: (a) a repeated subpattern containing an
687 extended class; (b) a recursive reference to a subpattern that followed a
688 previous extended class. It wasn't skipping over the extended class
689 correctly when UTF-8 mode was not set.]
691 5. A negated single-character class was not being recognized as fixed-length
692 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
693 compile error "lookbehind assertion is not fixed length".
695 6. The RunPerlTest auxiliary script was showing an unexpected difference
696 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
697 write a Perl script that can interpret lines of an input file either as
698 byte characters or as UTF-8, which is what "perltest" was being required to
699 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
700 can't do is switch easily at run time between having the "use utf8;" pragma
701 or not. In the end, I fudged it by using the RunPerlTest script to insert
702 "use utf8;" explicitly for the UTF-8 tests.
704 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
705 the end of the subject string, contrary to the documentation and to what
706 Perl does. This was true of both matching functions. Now it matches only at
707 the start of the subject and immediately after *internal* newlines.
709 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
710 a pointer to an int instead of a pointer to an unsigned long int. This
711 caused problems on 64-bit systems.
713 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
714 instance of the 'standard' template library not being so standard".
716 10. There was no check on the number of named subpatterns nor the maximum
717 length of a subpattern name. The product of these values is used to compute
718 the size of the memory block for a compiled pattern. By supplying a very
719 long subpattern name and a large number of named subpatterns, the size
720 computation could be caused to overflow. This is now prevented by limiting
721 the length of names to 32 characters, and the number of named subpatterns
722 to 10,000.
724 11. Subpatterns that are repeated with specific counts have to be replicated in
725 the compiled pattern. The size of memory for this was computed from the
726 length of the subpattern and the repeat count. The latter is limited to
727 65535, but there was no limit on the former, meaning that integer overflow
728 could in principle occur. The compiled length of a repeated subpattern is
729 now limited to 30,000 bytes in order to prevent this.
731 12. Added the optional facility to have named substrings with the same name.
733 13. Added the ability to use a named substring as a condition, using the
734 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
735 are numbers (not recommended). Forward references are permitted.
737 14. Added forward references in named backreferences (if you see what I mean).
739 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
740 pattern could run off the end of the subject. For example, the pattern
741 "(?s)(.{1,5})"8 did this with the subject "ab".
743 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
744 PCRE_CASELESS was set when matching characters that were quantified with ?
745 or *.
747 17. A character class other than a single negated character that had a minimum
748 but no maximum quantifier - for example [ab]{6,} - was not handled
749 correctly by pce_dfa_exec(). It would match only one character.
751 18. A valid (though odd) pattern that looked like a POSIX character
752 class but used an invalid character after [ (for example [[,abc,]]) caused
753 pcre_compile() to give the error "Failed: internal error: code overflow" or
754 in some cases to crash with a glibc free() error. This could even happen if
755 the pattern terminated after [[ but there just happened to be a sequence of
756 letters, a binary zero, and a closing ] in the memory that followed.
758 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
759 over the years. Originally (before any Unicode support), just the bottom 8
760 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
761 output from "man perlunicode" includes this:
763 The regular expression compiler produces polymorphic opcodes. That
764 is, the pattern adapts to the data and automatically switches to
765 the Unicode character scheme when presented with Unicode data--or
766 instead uses a traditional byte scheme when presented with byte
767 data.
769 Sadly, a wide octal escape does not cause a switch, and in a string with
770 no other multibyte characters, these octal escapes are treated as before.
771 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
772 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
773 Unicode string.
775 I have not perpetrated such confusion in PCRE. Up till now, it took just
776 the bottom 8 bits, as in old Perl. I have now made octal escapes with
777 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
778 translate to the appropriate multibyte character.
780 29. Applied some refactoring to reduce the number of warnings from Microsoft
781 and Borland compilers. This has included removing the fudge introduced
782 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
783 a warning about an unused variable.
785 21. PCRE has not included VT (character 0x0b) in the set of whitespace
786 characters since release 4.0, because Perl (from release 5.004) does not.
787 [Or at least, is documented not to: some releases seem to be in conflict
788 with the documentation.] However, when a pattern was studied with
789 pcre_study() and all its branches started with \s, PCRE still included VT
790 as a possible starting character. Of course, this did no harm; it just
791 caused an unnecessary match attempt.
793 22. Removed a now-redundant internal flag bit that recorded the fact that case
794 dependency changed within the pattern. This was once needed for "required
795 byte" processing, but is no longer used. This recovers a now-scarce options
796 bit. Also moved the least significant internal flag bit to the most-
797 significant bit of the word, which was not previously used (hangover from
798 the days when it was an int rather than a uint) to free up another bit for
799 the future.
801 23. Added support for CRLF line endings as well as CR and LF. As well as the
802 default being selectable at build time, it can now be changed at runtime
803 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
804 specify that it is scanning data with non-default line endings.
806 24. Changed the definition of CXXLINK to make it agree with the definition of
807 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
809 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
810 recursions. This makes a big different to stack usage for some patterns.
812 26. If a subpattern containing a named recursion or subroutine reference such
813 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
814 the space required for the compiled pattern went wrong and gave too small a
815 value. Depending on the environment, this could lead to "Failed: internal
816 error: code overflow at offset 49" or "glibc detected double free or
817 corruption" errors.
819 27. Applied patches from Google (a) to support the new newline modes and (b) to
820 advance over multibyte UTF-8 characters in GlobalReplace.
822 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
823 difference for some implementation of PCRE in some Windows version.
825 29. Added some extra testing facilities to pcretest:
827 \q<number> in a data line sets the "match limit" value
828 \Q<number> in a data line sets the "match recursion limt" value
829 -S <number> sets the stack size, where <number> is in megabytes
831 The -S option isn't available for Windows.
834 Version 6.6 06-Feb-06
835 ---------------------
837 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
838 in pcreposix.h. I have copied the definition from pcre.h.
840 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
841 because pcre.h is no longer a built file.
843 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
844 not normally included in the compiled code.
847 Version 6.5 01-Feb-06
848 ---------------------
850 1. When using the partial match feature with pcre_dfa_exec(), it was not
851 anchoring the second and subsequent partial matches at the new starting
852 point. This could lead to incorrect results. For example, with the pattern
853 /1234/, partially matching against "123" and then "a4" gave a match.
855 2. Changes to pcregrep:
857 (a) All non-match returns from pcre_exec() were being treated as failures
858 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
859 error message is output. Some extra information is given for the
861 probably the only errors that are likely to be caused by users (by
862 specifying a regex that has nested indefinite repeats, for instance).
863 If there are more than 20 of these errors, pcregrep is abandoned.
865 (b) A binary zero was treated as data while matching, but terminated the
866 output line if it was written out. This has been fixed: binary zeroes
867 are now no different to any other data bytes.
869 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
870 used to set a locale for matching. The --locale=xxxx long option has
871 been added (no short equivalent) to specify a locale explicitly on the
872 pcregrep command, overriding the environment variables.
874 (d) When -B was used with -n, some line numbers in the output were one less
875 than they should have been.
877 (e) Added the -o (--only-matching) option.
879 (f) If -A or -C was used with -c (count only), some lines of context were
880 accidentally printed for the final match.
882 (g) Added the -H (--with-filename) option.
884 (h) The combination of options -rh failed to suppress file names for files
885 that were found from directory arguments.
887 (i) Added the -D (--devices) and -d (--directories) options.
889 (j) Added the -F (--fixed-strings) option.
891 (k) Allow "-" to be used as a file name for -f as well as for a data file.
893 (l) Added the --colo(u)r option.
895 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
896 is not present by default.
898 3. A nasty bug was discovered in the handling of recursive patterns, that is,
899 items such as (?R) or (?1), when the recursion could match a number of
900 alternatives. If it matched one of the alternatives, but subsequently,
901 outside the recursion, there was a failure, the code tried to back up into
902 the recursion. However, because of the way PCRE is implemented, this is not
903 possible, and the result was an incorrect result from the match.
905 In order to prevent this happening, the specification of recursion has
906 been changed so that all such subpatterns are automatically treated as
907 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
909 4. I had overlooked the fact that, in some locales, there are characters for
910 which isalpha() is true but neither isupper() nor islower() are true. In
911 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
912 and ordfeminine) are like this. This affected the treatment of \w and \W
913 when they appeared in character classes, but not when they appeared outside
914 a character class. The bit map for "word" characters is now created
915 separately from the results of isalnum() instead of just taking it from the
916 upper, lower, and digit maps. (Plus the underscore character, of course.)
918 5. The above bug also affected the handling of POSIX character classes such as
919 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
920 permanent tables. Instead, the bit maps for such a class were previously
921 created as the appropriate unions of the upper, lower, and digit bitmaps.
922 Now they are created by subtraction from the [[:word:]] class, which has
923 its own bitmap.
925 6. The [[:blank:]] character class matches horizontal, but not vertical space.
926 It is created by subtracting the vertical space characters (\x09, \x0a,
927 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
928 subtraction was done in the overall bitmap for a character class, meaning
929 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
930 be recognized. This bug has been fixed.
932 7. Patches from the folks at Google:
934 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
935 real life, but is still worth protecting against".
937 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
938 regular expressions".
940 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
941 have it.
943 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
944 "configure" and the latter not, in order to fix a problem somebody had
945 with compiling the Arg class on HP-UX.
947 (e) Improve the error-handling of the C++ wrapper a little bit.
949 (f) New tests for checking recursion limiting.
951 8. The pcre_memmove() function, which is used only if the environment does not
952 have a standard memmove() function (and is therefore rarely compiled),
953 contained two bugs: (a) use of int instead of size_t, and (b) it was not
954 returning a result (though PCRE never actually uses the result).
956 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
957 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
958 returned instead of calling malloc() with an overflowing number that would
959 most likely cause subsequent chaos.
961 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
963 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
964 with this option is matched, the nmatch and pmatch options of regexec() are
965 ignored.
967 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
968 provided in case anyone wants to the the POSIX interface with UTF-8
969 strings.
971 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
972 C++ linking (needed for some HP-UX environments).
974 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
975 (unused parameter) and in the pcre_printint() function (omitted "default"
976 switch label when the default is to do nothing).
978 15. Added some code to make it possible, when PCRE is compiled as a C++
979 library, to replace subject pointers for pcre_exec() with a smart pointer
980 class, thus making it possible to process discontinuous strings.
982 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
983 much the same function. They were added by different people who were trying
984 to make PCRE easy to compile on non-Unix systems. It has been suggested
985 that PCRE_EXPORT be abolished now that there is more automatic apparatus
986 for compiling on Windows systems. I have therefore replaced it with
987 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
988 defaults to "extern" for C or "extern C" for C++, which works fine on
989 Unix-like systems. It is now possible to override the value of PCRE_DATA_
990 SCOPE with something explicit in config.h. In addition:
992 (a) pcreposix.h still had just "extern" instead of either of these macros;
993 I have replaced it with PCRE_DATA_SCOPE.
995 (b) Functions such as _pcre_xclass(), which are internal to the library,
996 but external in the C sense, all had PCRE_EXPORT in their definitions.
997 This is apparently wrong for the Windows case, so I have removed it.
998 (It makes no difference on Unix-like systems.)
1000 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1001 of recursive calls to match(). This is different to MATCH_LIMIT because
1002 that limits the total number of calls to match(), not all of which increase
1003 the depth of recursion. Limiting the recursion depth limits the amount of
1004 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1005 when PCRE is compiled, and changed at run time. A patch from Google adds
1006 this functionality to the C++ interface.
1008 18. Changes to the handling of Unicode character properties:
1010 (a) Updated the table to Unicode 4.1.0.
1012 (b) Recognize characters that are not in the table as "Cn" (undefined).
1014 (c) I revised the way the table is implemented to a much improved format
1015 which includes recognition of ranges. It now supports the ranges that
1016 are defined in UnicodeData.txt, and it also amalgamates other
1017 characters into ranges. This has reduced the number of entries in the
1018 table from around 16,000 to around 3,000, thus reducing its size
1019 considerably. I realized I did not need to use a tree structure after
1020 all - a binary chop search is just as efficient. Having reduced the
1021 number of entries, I extended their size from 6 bytes to 8 bytes to
1022 allow for more data.
1024 (d) Added support for Unicode script names via properties such as \p{Han}.
1026 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1027 matching that character.
1029 20. When matching a repeated Unicode property with a minimum greater than zero,
1030 (for example \pL{2,}), PCRE could look past the end of the subject if it
1031 reached it while seeking the minimum number of characters. This could
1032 happen only if some of the characters were more than one byte long, because
1033 there is a check for at least the minimum number of bytes.
1035 21. Refactored the implementation of \p and \P so as to be more general, to
1036 allow for more different types of property in future. This has changed the
1037 compiled form incompatibly. Anybody with saved compiled patterns that use
1038 \p or \P will have to recompile them.
1040 22. Added "Any" and "L&" to the supported property types.
1042 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1043 but give a compile time error if the value is greater than 0xff.
1045 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1046 accidentally not being installed or uninstalled.
1048 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1049 made were to insert the current release number. This seemed silly, because
1050 it made things harder for people building PCRE on systems that don't run
1051 "configure". I have turned pcre.h into a distributed file, no longer built
1052 by "configure", with the version identification directly included. There is
1053 no longer a pcre.h.in file.
1055 However, this change necessitated a change to the pcre-config script as
1056 well. It is built from pcre-config.in, and one of the substitutions was the
1057 release number. I have updated configure.ac so that ./configure now finds
1058 the release number by grepping pcre.h.
1060 26. Added the ability to run the tests under valgrind.
1063 Version 6.4 05-Sep-05
1064 ---------------------
1066 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1067 "--" to be printed when multiple files were scanned, even when none of the
1068 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1069 consider it to be a bug, and have restored the previous behaviour.
1071 2. A couple of code tidies to get rid of compiler warnings.
1073 3. The pcretest program used to cheat by referring to symbols in the library
1074 whose names begin with _pcre_. These are internal symbols that are not
1075 really supposed to be visible externally, and in some environments it is
1076 possible to suppress them. The cheating is now confined to including
1077 certain files from the library's source, which is a bit cleaner.
1079 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1080 file's purpose clearer.
1082 5. Reorganized pcre_ucp_findchar().
1085 Version 6.3 15-Aug-05
1086 ---------------------
1088 1. The file libpcre.pc.in did not have general read permission in the tarball.
1090 2. There were some problems when building without C++ support:
1092 (a) If C++ support was not built, "make install" and "make test" still
1093 tried to test it.
1095 (b) There were problems when the value of CXX was explicitly set. Some
1096 changes have been made to try to fix these, and ...
1098 (c) --disable-cpp can now be used to explicitly disable C++ support.
1100 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1101 backslash in a target when C++ was disabled. This confuses some
1102 versions of "make", apparently. Using an intermediate variable solves
1103 this. (Same for CPP_LOBJ.)
1105 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1106 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1107 necessary on certain architectures.
1109 4. Added a setting of -export-symbols-regex to the link command to remove
1110 those symbols that are exported in the C sense, but actually are local
1111 within the library, and not documented. Their names all begin with
1112 "_pcre_". This is not a perfect job, because (a) we have to except some
1113 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1114 available (and never for static libraries). I have made a note to try to
1115 find a way round (a) in the future.
1118 Version 6.2 01-Aug-05
1119 ---------------------
1121 1. There was no test for integer overflow of quantifier values. A construction
1122 such as {1111111111111111} would give undefined results. What is worse, if
1123 a minimum quantifier for a parenthesized subpattern overflowed and became
1124 negative, the calculation of the memory size went wrong. This could have
1125 led to memory overwriting.
1127 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1129 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1130 operating environments where this matters.
1132 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1133 PCRE options from within the C++ wrapper.
1135 5. Named capturing subpatterns were not being correctly counted when a pattern
1136 was compiled. This caused two problems: (a) If there were more than 100
1137 such subpatterns, the calculation of the memory needed for the whole
1138 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1139 back references of the form \12, where the number was greater than 9, were
1140 not recognized as back references, even though there were sufficient
1141 previous subpatterns.
1143 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1144 versions of gcc, e.g. 2.95.4.
1147 Version 6.1 21-Jun-05
1148 ---------------------
1150 1. There was one reference to the variable "posix" in pcretest.c that was not
1151 surrounded by "#if !defined NOPOSIX".
1153 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1154 the cross-check on the old pcre_info() function, for the benefit of the
1155 cut-down version of PCRE that is currently imported into Exim.
1157 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1158 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1159 patterns that start (?i) but I don't think that matters. The use of (?i) is
1160 just an example; this all applies to the other options as well.
1162 4. Since libtool seems to echo the compile commands it is issuing, the output
1163 from "make" can be reduced a bit by putting "@" in front of each libtool
1164 compile command.
1166 5. Patch from the folks at Google for configure.in to be a bit more thorough
1167 in checking for a suitable C++ installation before trying to compile the
1168 C++ stuff. This should fix a reported problem when a compiler was present,
1169 but no suitable headers.
1171 6. The man pages all had just "PCRE" as their title. I have changed them to
1172 be the relevant file name. I have also arranged that these names are
1173 retained in the file doc/pcre.txt, which is a concatenation in text format
1174 of all the man pages except the little individual ones for each function.
1176 7. The NON-UNIX-USE file had not been updated for the different set of source
1177 files that come with release 6. I also added a few comments about the C++
1178 wrapper.
1181 Version 6.0 07-Jun-05
1182 ---------------------
1184 1. Some minor internal re-organization to help with my DFA experiments.
1186 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1187 didn't matter for the library itself when fully configured, but did matter
1188 when compiling without UCP support, or within Exim, where the ucp files are
1189 not imported.
1191 3. Refactoring of the library code to split up the various functions into
1192 different source modules. The addition of the new DFA matching code (see
1193 below) to a single monolithic source would have made it really too
1194 unwieldy, quite apart from causing all the code to be include in a
1195 statically linked application, when only some functions are used. This is
1196 relevant even without the DFA addition now that patterns can be compiled in
1197 one application and matched in another.
1199 The downside of splitting up is that there have to be some external
1200 functions and data tables that are used internally in different modules of
1201 the library but which are not part of the API. These have all had their
1202 names changed to start with "_pcre_" so that they are unlikely to clash
1203 with other external names.
1205 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1206 a different (DFA) algorithm. Although it is slower than the original
1207 function, it does have some advantages for certain types of matching
1208 problem.
1210 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1211 including restarting after a partial match.
1213 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1214 defined when compiling for Windows was sent to me. I have put it into the
1215 code, though I have no means of testing or verifying it.
1217 7. Added the pcre_refcount() auxiliary function.
1219 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1220 match before or at the first newline in the subject string. In pcretest,
1221 the /f option on a pattern can be used to set this.
1223 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1224 would behave wrongly. This has been present in PCRE since release 4.0.
1226 10. A number of changes to the pcregrep command:
1228 (a) Refactored how -x works; insert ^(...)$ instead of setting
1229 PCRE_ANCHORED and checking the length, in preparation for adding
1230 something similar for -w.
1232 (b) Added the -w (match as a word) option.
1234 (c) Refactored the way lines are read and buffered so as to have more
1235 than one at a time available.
1237 (d) Implemented a pcregrep test script.
1239 (e) Added the -M (multiline match) option. This allows patterns to match
1240 over several lines of the subject. The buffering ensures that at least
1241 8K, or the rest of the document (whichever is the shorter) is available
1242 for matching (and similarly the previous 8K for lookbehind assertions).
1244 (f) Changed the --help output so that it now says
1246 -w, --word-regex(p)
1248 instead of two lines, one with "regex" and the other with "regexp"
1249 because that confused at least one person since the short forms are the
1250 same. (This required a bit of code, as the output is generated
1251 automatically from a table. It wasn't just a text change.)
1253 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1254 option but starts with a hyphen. Could be a pattern or a path name
1255 starting with a hyphen, for instance.
1257 (h) "-" can be given as a file name to represent stdin.
1259 (i) When file names are being printed, "(standard input)" is used for
1260 the standard input, for compatibility with GNU grep. Previously
1261 "<stdin>" was used.
1263 (j) The option --label=xxx can be used to supply a name to be used for
1264 stdin when file names are being printed. There is no short form.
1266 (k) Re-factored the options decoding logic because we are going to add
1267 two more options that take data. Such options can now be given in four
1268 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1270 (l) Added the -A, -B, and -C options for requesting that lines of context
1271 around matches be printed.
1273 (m) Added the -L option to print the names of files that do not contain
1274 any matching lines, that is, the complement of -l.
1276 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1277 continue to scan other files.
1279 (o) The -s option was incorrectly implemented. For compatibility with other
1280 greps, it now suppresses the error message for a non-existent or non-
1281 accessible file (but not the return code). There is a new option called
1282 -q that suppresses the output of matching lines, which was what -s was
1283 previously doing.
1285 (p) Added --include and --exclude options to specify files for inclusion
1286 and exclusion when recursing.
1288 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1289 Hopefully, it now does.
1291 12. Missing cast in pcre_study().
1293 13. Added an "uninstall" target to the makefile.
1295 14. Replaced "extern" in the function prototypes in Makefile.in with
1296 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1297 world, but is set differently for Windows.
1299 15. Added a second compiling function called pcre_compile2(). The only
1300 difference is that it has an extra argument, which is a pointer to an
1301 integer error code. When there is a compile-time failure, this is set
1302 non-zero, in addition to the error test pointer being set to point to an
1303 error message. The new argument may be NULL if no error number is required
1304 (but then you may as well call pcre_compile(), which is now just a
1305 wrapper). This facility is provided because some applications need a
1306 numeric error indication, but it has also enabled me to tidy up the way
1307 compile-time errors are handled in the POSIX wrapper.
1309 16. Added VPATH=.libs to the makefile; this should help when building with one
1310 prefix path and installing with another. (Or so I'm told by someone who
1311 knows more about this stuff than I do.)
1313 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1314 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1315 match everything, including newlines. This is not POSIX-compatible, but
1316 somebody wanted the feature. From pcretest it can be activated by using
1317 both the P and the s flags.
1319 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1321 19. libpcre.pc was being incorrectly installed as executable.
1323 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1324 it now also looks for '\r' so that it will work unmodified on Windows.
1326 21. Added Google's contributed C++ wrapper to the distribution.
1328 22. Added some untidy missing memory free() calls in pcretest, to keep
1329 Electric Fence happy when testing.
1333 Version 5.0 13-Sep-04
1334 ---------------------
1336 1. Internal change: literal characters are no longer packed up into items
1337 containing multiple characters in a single byte-string. Each character
1338 is now matched using a separate opcode. However, there may be more than one
1339 byte in the character in UTF-8 mode.
1341 2. The pcre_callout_block structure has two new fields: pattern_position and
1342 next_item_length. These contain the offset in the pattern to the next match
1343 item, and its length, respectively.
1345 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1346 insertion of callouts before each pattern item. Added the /C option to
1347 pcretest to make use of this.
1349 4. On the advice of a Windows user, the lines
1351 #if defined(_WIN32) || defined(WIN32)
1352 _setmode( _fileno( stdout ), 0x8000 );
1353 #endif /* defined(_WIN32) || defined(WIN32) */
1355 have been added to the source of pcretest. This apparently does useful
1356 magic in relation to line terminators.
1358 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1359 for the benefit of those environments where the "b" makes a difference.
1361 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1362 to know about it. I have put a hack into configure.in that adds in code
1363 to set GCC=yes if CC=icc. This seems to end up at a point in the
1364 generated configure script that is early enough to affect the setting of
1365 compiler options, which is what is needed, but I have no means of testing
1366 whether it really works. (The user who reported this had patched the
1367 generated configure script, which of course I cannot do.)
1369 LATER: After change 22 below (new libtool files), the configure script
1370 seems to know about icc (and also ecc). Therefore, I have commented out
1371 this hack in configure.in.
1373 7. Added support for pkg-config (2 patches were sent in).
1375 8. Negated POSIX character classes that used a combination of internal tables
1376 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1377 [[:^ascii]]. Typically, they would match almost any characters. The other
1378 POSIX classes were not broken in this way.
1380 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1381 to find the match, as PCRE was deluded into thinking that the match had to
1382 start at the start point or following a newline. The same bug applied to
1383 patterns with negative forward assertions or any backward assertions
1384 preceding ".*" at the start, unless the pattern required a fixed first
1385 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1387 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1388 starting at the last subject character, bytes beyond the end of the subject
1389 string were read.
1391 11. Renamed the variable "class" as "classbits" to make life easier for C++
1392 users. (Previously there was a macro definition, but it apparently wasn't
1393 enough.)
1395 12. Added the new field "tables" to the extra data so that tables can be passed
1396 in at exec time, or the internal tables can be re-selected. This allows
1397 a compiled regex to be saved and re-used at a later time by a different
1398 program that might have everything at different addresses.
1400 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1401 -R library as well as a -L library.
1403 14. The debugging options of pcretest (-d on the command line or D on a
1404 pattern) showed incorrect output for anything following an extended class
1405 that contained multibyte characters and which was followed by a quantifier.
1407 15. Added optional support for general category Unicode character properties
1408 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1409 support. It adds about 90K to the size of the library. The meanings of the
1410 inbuilt class escapes such as \d and \s have NOT been changed.
1412 16. Updated pcredemo.c to include calls to free() to release the memory for the
1413 compiled pattern.
1415 17. The generated file chartables.c was being created in the source directory
1416 instead of in the building directory. This caused the build to fail if the
1417 source directory was different from the building directory, and was
1418 read-only.
1420 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1421 file. No doubt somebody will tell me if they don't make sense... Also added
1422 Dan Mooney's comments about building on OpenVMS.
1424 19. Added support for partial matching via the PCRE_PARTIAL option for
1425 pcre_exec() and the \P data escape in pcretest.
1427 20. Extended pcretest with 3 new pattern features:
1429 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1430 write the compiled pattern to the file whose name is "rest-of-line".
1431 This is a straight binary dump of the data, with the saved pointer to
1432 the character tables forced to be NULL. The study data, if any, is
1433 written too. After writing, pcretest reads a new pattern.
1435 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1436 compiled pattern from the given file. There must not be any
1437 occurrences of "<" in the file name (pretty unlikely); if there are,
1438 pcretest will instead treat the initial "<" as a pattern delimiter.
1439 After reading in the pattern, pcretest goes on to read data lines as
1440 usual.
1442 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1443 and 16-bit fields in a compiled pattern, to simulate a pattern that
1444 was compiled on a host of opposite endianness.
1446 21. The pcre-exec() function can now cope with patterns that were compiled on
1447 hosts of opposite endianness, with this restriction:
1449 As for any compiled expression that is saved and used later, the tables
1450 pointer field cannot be preserved; the extra_data field in the arguments
1451 to pcre_exec() should be used to pass in a tables address if a value
1452 other than the default internal tables were used at compile time.
1454 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1455 now diagnosed as an error. Previously, most of the time, a negative number
1456 would have been treated as zero, but if in addition "ovector" was passed as
1457 NULL, a crash could occur.
1459 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1460 new versions from the libtool 1.5 distribution (the last one is a copy of
1461 a file called libtool.m4). This seems to have fixed the need to patch
1462 "configure" to support Darwin 1.3 (which I used to do). However, I still
1463 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1464 workstation).
1466 24. Changed the PCRE licence to be the more standard "BSD" licence.
1469 Version 4.5 01-Dec-03
1470 ---------------------
1472 1. There has been some re-arrangement of the code for the match() function so
1473 that it can be compiled in a version that does not call itself recursively.
1474 Instead, it keeps those local variables that need separate instances for
1475 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1476 needs to "recurse". Keeping track of where control must go is done by means
1477 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1478 hide most of the details from the main code, and operates only if
1479 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1480 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1481 operating.
1483 To make it easier for callers to provide specially tailored get/free
1484 functions for this usage, two new functions, pcre_stack_malloc, and
1485 pcre_stack_free, are used. They are always called in strict stacking order,
1486 and the size of block requested is always the same.
1488 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1489 PCRE has been compiled to use the stack or the heap for recursion. The
1490 -C option of pcretest uses this to show which version is compiled.
1492 A new data escape \S, is added to pcretest; it causes the amounts of store
1493 obtained and freed by both kinds of malloc/free at match time to be added
1494 to the output.
1496 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1497 what's available on my current Linux desktop machine.
1499 3. When matching a UTF-8 string, the test for a valid string at the start has
1500 been extended. If start_offset is not zero, PCRE now checks that it points
1501 to a byte that is the start of a UTF-8 character. If not, it returns
1502 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1503 this is necessary because there may be backward assertions in the pattern.
1504 When matching the same subject several times, it may save resources to use
1505 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1507 4. The code for checking the validity of UTF-8 strings has been tightened so
1508 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1509 containing "overlong sequences".
1511 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1512 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1513 should have been "&", but it just so happened that all the cases this let
1514 through by mistake were picked up later in the function.
1516 6. I had used a variable called "isblank" - this is a C99 function, causing
1517 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1519 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1520 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1521 so the version is shown; (c) stop "make test" echoing "./RunTest".
1523 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1525 9. The prototype for memmove() for systems that don't have it was using
1526 size_t, but the inclusion of the header that defines size_t was later. I've
1527 moved the #includes for the C headers earlier to avoid this.
1529 10. Added some adjustments to the code to make it easier to compiler on certain
1530 special systems:
1532 (a) Some "const" qualifiers were missing.
1533 (b) Added the macro EXPORT before all exported functions; by default this
1534 is defined to be empty.
1535 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1536 that it reads its output file name as an argument instead of writing
1537 to the standard output and assuming this can be redirected.
1539 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1540 class containing characters with values greater than 255, PCRE compilation
1541 went into a loop.
1543 12. A recursive reference to a subpattern that was within another subpattern
1544 that had a minimum quantifier of zero caused PCRE to crash. For example,
1545 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1546 recursion. If the recursively-called subpattern itself had a zero repeat,
1547 that was OK.
1549 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1550 buffer into which it was copied (for escape processing) was still set at
1551 1024, so long lines caused crashes.
1553 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1554 "internal error: code overflow...". This applied to any character class
1555 that was followed by a possessive quantifier.
1557 15. Modified the Makefile to add libpcre.la as a prerequisite for
1558 libpcreposix.la because I was told this is needed for a parallel build to
1559 work.
1561 16. If a pattern that contained .* following optional items at the start was
1562 studied, the wrong optimizing data was generated, leading to matching
1563 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1564 matching string must start with a or b or c. The correct conclusion for
1565 this pattern is that a match can start with any character.
1568 Version 4.4 13-Aug-03
1569 ---------------------
1571 1. In UTF-8 mode, a character class containing characters with values between
1572 127 and 255 was not handled correctly if the compiled pattern was studied.
1573 In fixing this, I have also improved the studying algorithm for such
1574 classes (slightly).
1576 2. Three internal functions had redundant arguments passed to them. Removal
1577 might give a very teeny performance improvement.
1579 3. Documentation bug: the value of the capture_top field in a callout is *one
1580 more than* the number of the hightest numbered captured substring.
1582 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1583 in incorrectly linking with a previously installed version. They now link
1584 explicitly with libpcre.la.
1586 5. configure.in no longer needs to recognize Cygwin specially.
1588 6. A problem in pcre.in for Windows platforms is fixed.
1590 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1591 pcretest, it used to include the size of the study block as part of its
1592 output. Unfortunately, the structure contains a field that has a different
1593 size on different hardware architectures. This meant that the tests that
1594 showed this size failed. As the block is currently always of a fixed size,
1595 this information isn't actually particularly useful in pcretest output, so
1596 I have just removed it.
1598 8. Three pre-processor statements accidentally did not start in column 1.
1599 Sadly, there are *still* compilers around that complain, even though
1600 standard C has not required this for well over a decade. Sigh.
1602 9. In pcretest, the code for checking callouts passed small integers in the
1603 callout_data field, which is a void * field. However, some picky compilers
1604 complained about the casts involved for this on 64-bit systems. Now
1605 pcretest passes the address of the small integer instead, which should get
1606 rid of the warnings.
1608 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1609 both compile and run time, and gives an error if an invalid UTF-8 sequence
1610 is found. There is a option for disabling this check in cases where the
1611 string is known to be correct and/or the maximum performance is wanted.
1613 11. In response to a bug report, I changed one line in Makefile.in from
1615 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1616 to
1617 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1619 to look similar to other lines, but I have no way of telling whether this
1620 is the right thing to do, as I do not use Windows. No doubt I'll get told
1621 if it's wrong...
1624 Version 4.3 21-May-03
1625 ---------------------
1627 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1628 Makefile.
1630 2. Some refactoring to improve the quality of the code:
1632 (i) The utf8_table... variables are now declared "const".
1634 (ii) The code for \cx, which used the "case flipping" table to upper case
1635 lower case letters, now just substracts 32. This is ASCII-specific,
1636 but the whole concept of \cx is ASCII-specific, so it seems
1637 reasonable.
1639 (iii) PCRE was using its character types table to recognize decimal and
1640 hexadecimal digits in the pattern. This is silly, because it handles
1641 only 0-9, a-f, and A-F, but the character types table is locale-
1642 specific, which means strange things might happen. A private
1643 table is now used for this - though it costs 256 bytes, a table is
1644 much faster than multiple explicit tests. Of course, the standard
1645 character types table is still used for matching digits in subject
1646 strings against \d.
1648 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1649 ending in _t are). So I've renamed it as ESC_tee.
1651 3. The first argument for regexec() in the POSIX wrapper should have been
1652 defined as "const".
1654 4. Changed pcretest to use malloc() for its buffers so that they can be
1655 Electric Fenced for debugging.
1657 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1658 to read one or more bytes before the start of the subject string. Often this
1659 had no effect on PCRE's behaviour, but in some circumstances it could
1660 provoke a segmentation fault.
1662 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1663 to try to read one or more bytes before the start of the subject string.
1665 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1666 UTF-8 support could misbehave in various ways if the subject string
1667 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1668 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1669 back over UTF-8 characters.)
1672 Version 4.2 14-Apr-03
1673 ---------------------
1675 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1677 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1678 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1679 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1680 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1681 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1683 Note: automatic setting of the BUILD variables is not yet working
1684 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1685 compile-time but not at link-time
1686 [LINK]: use for linking executables only
1687 make different versions for Windows and non-Windows
1688 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1689 libraries
1690 [LINK_FOR_BUILD]: new variable
1691 [OBJEXT]: use throughout
1692 [EXEEXT]: use throughout
1693 <winshared>: new target
1694 <wininstall>: new target
1695 <dftables.o>: use native compiler
1696 <dftables>: use native linker
1697 <install>: handle Windows platform correctly
1698 <clean>: ditto
1699 <check>: ditto
1700 copy DLL to top builddir before testing
1702 As part of these changes, -no-undefined was removed again. This was reported
1703 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1704 in any case.
1706 3. Some tidies to get rid of compiler warnings:
1708 . In the match_data structure, match_limit was an unsigned long int, whereas
1709 match_call_count was an int. I've made them both unsigned long ints.
1711 . In pcretest the fact that a const uschar * doesn't automatically cast to
1712 a void * provoked a warning.
1714 . Turning on some more compiler warnings threw up some "shadow" variables
1715 and a few more missing casts.
1717 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1718 option, a class that contained a single character with a value between 128
1719 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1721 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1722 option, a class that contained several characters, but with at least one
1723 whose value was between 128 and 255 caused PCRE to crash.
1726 Version 4.1 12-Mar-03
1727 ---------------------
1729 1. Compiling with gcc -pedantic found a couple of places where casts were
1730 needed, and a string in dftables.c that was longer than standard compilers are
1731 required to support.
1733 2. Compiling with Sun's compiler found a few more places where the code could
1734 be tidied up in order to avoid warnings.
1736 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1737 first of these names is deprecated in the latest Autoconf in favour of the name
1738 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1739 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1740 analogy I have changed it to CFLAGS_FOR_BUILD.
1742 4. Added -no-undefined to the linking command in the Makefile, because this is
1743 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1744 linking step for the pcreposix library.
1746 5. PCRE was failing to diagnose the case of two named groups with the same
1747 name.
1749 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1750 literal character that is needed in the subject for a match, and scans along to
1751 ensure that it is present before embarking on the full matching process. This
1752 saves time in cases of nested unlimited repeats that are never going to match.
1753 Problem: the scan can take a lot of time if the subject is very long (e.g.
1754 megabytes), thus penalizing straightforward matches. It is now done only if the
1755 amount of subject to be scanned is less than 1000 bytes.
1757 7. A lesser problem with the same optimization is that it was recording the
1758 first character of an anchored pattern as "needed", thus provoking a search
1759 right along the subject, even when the first match of the pattern was going to
1760 fail. The "needed" character is now not set for anchored patterns, unless it
1761 follows something in the pattern that is of non-fixed length. Thus, it still
1762 fulfils its original purpose of finding quick non-matches in cases of nested
1763 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1766 Version 4.0 17-Feb-03
1767 ---------------------
1769 1. If a comment in an extended regex that started immediately after a meta-item
1770 extended to the end of string, PCRE compiled incorrect data. This could lead to
1771 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1773 2. Moved to autoconf 2.53 and libtool 1.4.2.
1775 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1776 the special perltest8 script is no longer needed - all the tests can be run
1777 from a single perltest script.
1779 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1780 by \s. It has now been removed in PCRE. This means it isn't recognized as
1781 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1782 class [:space:] *does* include VT, thereby creating a mess.
1784 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1785 space and tab.
1787 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1788 its new features into the main test script, reducing the number of scripts.
1790 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1791 were backward compatible, and made the (?i) apply to the whole pattern, as if
1792 /i were given. Now it behaves more logically, and applies the option setting
1793 only to what follows. PCRE has been changed to follow suit. However, if it
1794 finds options settings right at the start of the pattern, it extracts them into
1795 the global options, as before. Thus, they show up in the info data.
1797 8. Added support for the \Q...\E escape sequence. Characters in between are
1798 treated as literals. This is slightly different from Perl in that $ and @ are
1799 also handled as literals inside the quotes. In Perl, they will cause variable
1800 interpolation. Note the following examples:
1802 Pattern PCRE matches Perl matches
1804 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1805 \Qabc\$xyz\E abc\$xyz abc\$xyz
1806 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1808 For compatibility with Perl, \Q...\E sequences are recognized inside character
1809 classes as well as outside them.
1811 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1812 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1813 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1814 signed/unsigned warnings.
1816 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1817 option for pcretest, so I've replaced it by a simple function that does just
1818 that job.
1820 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1821 "pcregrep -".
1823 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1824 Java package. This provides some syntactic sugar for simple cases of what my
1825 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1826 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1827 item, you can use this simplified notation. Note that only makes sense with
1828 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1829 greediness, whatever the setting of the PCRE_UNGREEDY option.
1831 13. A change of greediness default within a pattern was not taking effect at
1832 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1833 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1834 was abstracted outside.
1836 14. PCRE now supports the \G assertion. It is true when the current matching
1837 position is at the start point of the match. This differs from \A when the
1838 starting offset is non-zero. Used with the /g option of pcretest (or similar
1839 code), it works in the same way as it does for Perl's /g option. If all
1840 alternatives of a regex begin with \G, the expression is anchored to the start
1841 match position, and the "anchored" flag is set in the compiled expression.
1843 15. Some bugs concerning the handling of certain option changes within patterns
1844 have been fixed. These applied to options other than (?ims). For example,
1845 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1846 been the other way round. Some of this was related to change 7 above.
1848 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1849 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1850 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1851 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1853 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1854 mode. Unlike ".", it always matches newline, whatever the setting of
1855 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1856 assertions. Perl allows it, but it doesn't (in general) work because it can't
1857 calculate the length of the lookbehind. At least, that's the case for Perl
1858 5.8.0 - I've been told they are going to document that it doesn't work in
1859 future.
1861 18. Added an error diagnosis for escapes that PCRE does not support: these are
1862 \L, \l, \N, \P, \p, \U, \u, and \X.
1864 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1865 reading past the end of the pattern in cases such as /[abcd/.
1867 20. PCRE was getting more memory than necessary for patterns with classes that
1868 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1870 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1871 compiling PCRE for use with Virtual Pascal.
1873 22. Small fix to the Makefile to make it work properly if the build is done
1874 outside the source tree.
1876 23. Added a new extension: a condition to go with recursion. If a conditional
1877 subpattern starts with (?(R) the "true" branch is used if recursion has
1878 happened, whereas the "false" branch is used only at the top level.
1880 24. When there was a very long string of literal characters (over 255 bytes
1881 without UTF support, over 250 bytes with UTF support), the computation of how
1882 much memory was required could be incorrect, leading to segfaults or other
1883 strange effects.
1885 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1886 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1887 there was a subsequent back reference to those brackets. This meant that, for
1888 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1889 possible to check for precisely this case. All we can do is abandon the
1890 optimization if .* occurs inside capturing brackets when there are any back
1891 references whatsoever. (See below for a better fix that came later.)
1893 26. The handling of the optimization for finding the first character of a
1894 non-anchored pattern, and for finding a character that is required later in the
1895 match were failing in some cases. This didn't break the matching; it just
1896 failed to optimize when it could. The way this is done has been re-implemented.
1898 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1900 28. Added a new feature that provides some of the functionality that Perl
1901 provides with (?{...}). The facility is termed a "callout". The way it is done
1902 in PCRE is for the caller to provide an optional function, by setting
1903 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1904 global variable. By default it is unset, which disables all calling out. To get
1905 the function called, the regex must include (?C) at appropriate points. This
1906 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1907 This provides a means of identifying different callout points. When PCRE
1908 reaches such a point in the regex, if pcre_callout has been set, the external
1909 function is called. It is provided with data in a structure called
1910 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1911 matching continues; if it returns a non-zero value, the match at the current
1912 point fails. However, backtracking will occur if possible. [This was changed
1913 later and other features added - see item 49 below.]
1915 29. pcretest is upgraded to test the callout functionality. It provides a
1916 callout function that displays information. By default, it shows the start of
1917 the match and the current position in the text. There are some new data escapes
1918 to vary what happens:
1920 \C+ in addition, show current contents of captured substrings
1921 \C- do not supply a callout function
1922 \C!n return 1 when callout number n is reached
1923 \C!n!m return 1 when callout number n is reached for the mth time
1925 30. If pcregrep was called with the -l option and just a single file name, it
1926 output "<stdin>" if a match was found, instead of the file name.
1928 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1929 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1930 pcre_exec(). This saves a malloc/free per call. The default value of
1931 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1932 when configuring.
1934 32. The default maximum size of a compiled pattern is 64K. There have been a
1935 few cases of people hitting this limit. The code now uses macros to handle the
1936 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1937 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1938 configuring. Tests 2 and 5 work only with 2-byte links because they output
1939 debugging information about compiled patterns.
1941 33. Internal code re-arrangements:
1943 (a) Moved the debugging function for printing out a compiled regex into
1944 its own source file (printint.c) and used #include to pull it into
1945 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1946 separate copies.
1948 (b) Defined the list of op-code names for debugging as a macro in
1949 internal.h so that it is next to the definition of the opcodes.
1951 (c) Defined a table of op-code lengths for simpler skipping along compiled
1952 code. This is again a macro in internal.h so that it is next to the
1953 definition of the opcodes.
1955 34. Added support for recursive calls to individual subpatterns, along the
1956 lines of Robin Houston's patch (but implemented somewhat differently).
1958 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1959 allow it to read and process whole directories in Win32. This code was
1960 contributed by Lionel Fourquaux; it has not been tested by me.
1962 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1963 used to name a group. Names consist of alphanumerics and underscores, and must
1964 be unique. Back references use the syntax (?P=name) and recursive calls use
1965 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1966 numbers. The function pcre_fullinfo() can be used after compilation to extract
1967 a name/number map. There are three relevant calls:
1969 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1970 PCRE_INFO_NAMECOUNT yields the number of entries
1971 PCRE_INFO_NAMETABLE yields a pointer to the map.
1973 The map is a vector of fixed-size entries. The size of each entry depends on
1974 the length of the longest name used. The first two bytes of each entry are the
1975 group number, most significant byte first. There follows the corresponding
1976 name, zero terminated. The names are in alphabetical order.
1978 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1979 case instead of 255. Making it the same both with and without UTF-8 support
1980 means that the same test output works with both.
1982 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1983 calling malloc() with a zero argument.
1985 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1986 optimization. I've improved things by keeping a bitmap of backreferences with
1987 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1988 fact referenced, the optimization can be applied. It is unlikely that a
1989 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1990 the match to follow \n) will appear inside brackets with a number greater than
1991 31, but if it does, any back reference > 31 suppresses the optimization.
1993 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1994 of disabling numbered capturing parentheses. Any opening parenthesis that is
1995 not followed by ? behaves as if it were followed by ?: but named parentheses
1996 can still be used for capturing (and they will acquire numbers in the usual
1997 way).
1999 41. Redesigned the return codes from the match() function into yes/no/error so
2000 that errors can be passed back from deep inside the nested calls. A malloc
2001 failure while inside a recursive subpattern call now causes the
2002 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2004 42. It is now possible to set a limit on the number of times the match()
2005 function is called in a call to pcre_exec(). This facility makes it possible to
2006 limit the amount of recursion and backtracking, though not in a directly
2007 obvious way, because the match() function is used in a number of different
2008 circumstances. The count starts from zero for each position in the subject
2009 string (for non-anchored patterns). The default limit is, for compatibility, a
2010 large number, namely 10 000 000. You can change this in two ways:
2012 (a) When configuring PCRE before making, you can use --with-match-limit=n
2013 to set a default value for the compiled library.
2015 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2016 a different value is set. See 45 below.
2018 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2020 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2021 of things that can be changed at compile time. The first argument specifies
2022 what is wanted and the second points to where the information is to be placed.
2023 The current list of available information is:
2027 The output is an integer that is set to one if UTF-8 support is available;
2028 otherwise it is set to zero.
2032 The output is an integer that it set to the value of the code that is used for
2033 newline. It is either LF (10) or CR (13).
2037 The output is an integer that contains the number of bytes used for internal
2038 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2042 The output is an integer that contains the threshold above which the POSIX
2043 interface uses malloc() for output vectors. See item 31 above.
2047 The output is an unsigned integer that contains the default limit of the number
2048 of match() calls in a pcre_exec() execution. See 42 above.
2050 44. pcretest has been upgraded by the addition of the -C option. This causes it
2051 to extract all the available output from the new pcre_config() function, and to
2052 output it. The program then exits immediately.
2054 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2055 order to support additional features. One way would have been to define
2056 pcre_exec2() (for example) with extra arguments, but this would not have been
2057 extensible, and would also have required all calls to the original function to
2058 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2059 is used for passing in "extra" data from pcre_study().
2061 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2062 contains the following fields:
2064 flags a bitmap indicating which of the following fields are set
2065 study_data opaque data from pcre_study()
2066 match_limit a way of specifying a limit on match() calls for a specific
2067 call to pcre_exec()
2068 callout_data data for callouts (see 49 below)
2070 The flag bits are also defined in pcre.h, and are
2076 The pcre_study() function now returns one of these new pcre_extra blocks, with
2077 the actual study data pointed to by the study_data field, and the
2078 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2079 before. That is, this change is entirely upwards-compatible and requires no
2080 change to existing code.
2082 If you want to pass in additional data to pcre_exec(), you can either place it
2083 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2084 block.
2086 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2087 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2088 times with different match limits, until it finds the minimum value needed for
2089 pcre_exec() to complete. The value is then output. This can be instructive; for
2090 most simple matches the number is quite small, but for pathological cases it
2091 gets very large very quickly.
2093 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2094 returns the size of the data block pointed to by the study_data field in a
2095 pcre_extra block, that is, the value that was passed as the argument to
2096 pcre_malloc() when PCRE was getting memory in which to place the information
2097 created by pcre_study(). The fourth argument should point to a size_t variable.
2098 pcretest has been extended so that this information is shown after a successful
2099 pcre_study() call when information about the compiled regex is being displayed.
2101 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2102 because what follows is always an absolute path. (Later: it turns out that this
2103 is more than cosmetic for MinGW, because it doesn't like empty path
2104 components.)
2106 49. Some changes have been made to the callout feature (see 28 above):
2108 (i) A callout function now has three choices for what it returns:
2110 0 => success, carry on matching
2111 > 0 => failure at this point, but backtrack if possible
2112 < 0 => serious error, return this value from pcre_exec()
2114 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2115 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2116 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2117 use by callout functions. It will never be used by PCRE itself.
2119 (ii) The pcre_extra structure (see 45 above) has a void * field called
2120 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2121 pcre_callout_block structure has a field of the same name. The contents of
2122 the field passed in the pcre_extra structure are passed to the callout
2123 function in the corresponding field in the callout block. This makes it
2124 easier to use the same callout-containing regex from multiple threads. For
2125 testing, the pcretest program has a new data escape
2127 \C*n pass the number n (may be negative) as callout_data
2129 If the callout function in pcretest receives a non-zero value as
2130 callout_data, it returns that value.
2132 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2133 there were some redundant $(CFLAGS) in commands that are now specified as
2134 $(LINK), which already includes $(CFLAGS).
2136 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2137 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2138 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2139 one-byte characters throughout. Note that case-insensitive matching applies
2140 only to characters whose values are less than 256. PCRE doesn't support the
2141 notion of cases for higher-valued characters.
2143 (i) A character class whose characters are all within 0-255 is handled as
2144 a bit map, and the map is inverted for negative classes. Previously, a
2145 character > 255 always failed to match such a class; however it should
2146 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2148 (ii) A negated character class with a single character < 255 is coded as
2149 "not this character" (OP_NOT). This wasn't working properly when the test
2150 character was multibyte, either singly or repeated.
2152 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2153 mode, for example: \x{100}{2,3}.
2155 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2156 singly or repeated) now correctly test multibyte characters. However,
2157 PCRE doesn't recognize any characters with values greater than 255 as
2158 digits, spaces, or word characters. Such characters always match \D, \S,
2159 and \W, and never match \d, \s, or \w.
2161 (v) Classes may now contain characters and character ranges with values
2162 greater than 255. For example: [ab\x{100}-\x{400}].
2164 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2165 PCRE in UTF-8 mode.
2167 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2168 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2169 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2170 value.)
2172 53. The single man page has become too large. I have therefore split it up into
2173 a number of separate man pages. These also give rise to individual HTML pages;
2174 these are now put in a separate directory, and there is an index.html page that
2175 lists them all. Some hyperlinking between the pages has been installed.
2177 54. Added convenience functions for handling named capturing parentheses.
2179 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2180 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2181 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2182 are faulted.
2184 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2185 calling configure. These values are used when compiling the dftables.c program
2186 which is run to generate the source of the default character tables. They
2187 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2188 you will need to set these values.
2190 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2193 Version 3.9 02-Jan-02
2194 ---------------------
2196 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2198 2. If --disable-static was given, the building process failed when trying to
2199 build pcretest and pcregrep. (For some reason it was using libtool to compile
2200 them, which is not right, as they aren't part of the library.)
2203 Version 3.8 18-Dec-01
2204 ---------------------
2206 1. The experimental UTF-8 code was completely screwed up. It was packing the
2207 bytes in the wrong order. How dumb can you get?
2210 Version 3.7 29-Oct-01
2211 ---------------------
2213 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2214 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2215 this didn't happen under Solaris 8, where I normally test things.
2217 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2218 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2219 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2220 this fix an hour or so after the initial 3.7 release.)
2223 Version 3.6 23-Oct-01
2224 ---------------------
2226 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2227 offsets passed as NULL with zero offset count.
2229 2. The config.guess and config.sub files had not been updated when I moved to
2230 the latest autoconf.
2233 Version 3.5 15-Aug-01
2234 ---------------------
2236 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2237 had been forgotten.
2239 2. By using declared but undefined structures, we can avoid using "void"
2240 definitions in pcre.h while keeping the internal definitions of the structures
2241 private.
2243 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2244 user point of view, this means that both static and shared libraries are built
2245 by default, but this can be individually controlled. More of the work of
2246 handling this static/shared cases is now inside libtool instead of PCRE's make
2247 file.
2249 4. The pcretest utility is now installed along with pcregrep because it is
2250 useful for users (to test regexs) and by doing this, it automatically gets
2251 relinked by libtool. The documentation has been turned into a man page, so
2252 there are now .1, .txt, and .html versions in /doc.
2254 5. Upgrades to pcregrep:
2255 (i) Added long-form option names like gnu grep.
2256 (ii) Added --help to list all options with an explanatory phrase.
2257 (iii) Added -r, --recursive to recurse into sub-directories.
2258 (iv) Added -f, --file to read patterns from a file.
2260 6. pcre_exec() was referring to its "code" argument before testing that
2261 argument for NULL (and giving an error if it was NULL).
2263 7. Upgraded Makefile.in to allow for compiling in a different directory from
2264 the source directory.
2266 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2267 options bits, the pointer it was passed was to an int instead of to an unsigned
2268 long int. This mattered only on 64-bit systems.
2270 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2271 generated) instead of pcre.in, which it its source. Also made the same change
2272 in several of the .c files.
2274 10. A new release of gcc defines printf() as a macro, which broke pcretest
2275 because it had an ifdef in the middle of a string argument for printf(). Fixed
2276 by using separate calls to printf().
2278 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2279 script, to force use of CR or LF instead of \n in the source. On non-Unix
2280 systems, the value can be set in config.h.
2282 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2283 absolute limit. Changed the text of the error message to make this clear, and
2284 likewise updated the man page.
2286 13. The limit of 99 on the number of capturing subpatterns has been removed.
2287 The new limit is 65535, which I hope will not be a "real" limit.
2290 Version 3.4 22-Aug-00
2291 ---------------------
2293 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2295 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2298 Version 3.3 01-Aug-00
2299 ---------------------
2301 1. If an octal character was given, but the value was greater than \377, it
2302 was not getting masked to the least significant bits, as documented. This could
2303 lead to crashes in some systems.
2305 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2306 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2308 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2309 These just pass their arguments on to (pcre_free)(), but they are provided
2310 because some uses of PCRE bind it to non-C systems that can call its functions,
2311 but cannot call free() or pcre_free() directly.
2313 4. Add "make test" as a synonym for "make check". Corrected some comments in
2314 the Makefile.
2316 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2317 Makefile.
2319 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2320 command called pgrep for grepping around the active processes.
2322 7. Added the beginnings of support for UTF-8 character strings.
2324 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2325 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2326 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2327 out for the ar command.)
2330 Version 3.2 12-May-00
2331 ---------------------
2333 This is purely a bug fixing release.
2335 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2336 of ZA. This was just one example of several cases that could provoke this bug,
2337 which was introduced by change 9 of version 2.00. The code for breaking
2338 infinite loops after an iteration that matches an empty string was't working
2339 correctly.
2341 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2342 when matched against abbab (for example). After matching an empty string, it
2343 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2344 caused it to match further down the string than it should.
2346 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2347 was there because it doesn't seem to be needed, and it causes trouble on some
2348 systems, as it is not a Standard C header. It has been removed.
2350 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2351 were reported on the Macintosh. The changes were from
2353 while ((c = *(++ptr)) != 0 && c != '\n');
2354 to
2355 while ((c = *(++ptr)) != 0 && c != '\n') ;
2357 Totally extraordinary, but if that's what it takes...
2359 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2360 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2361 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2362 assumes the way PCRE uses memmove() (always moving upwards).
2364 6. PCRE is being used in one environment where strchr() is not available. There
2365 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2366 faster code anyway.
2369 Version 3.1 09-Feb-00
2370 ---------------------
2372 The only change in this release is the fixing of some bugs in Makefile.in for
2373 the "install" target:
2375 (1) It was failing to install pcreposix.h.
2377 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2380 Version 3.0 01-Feb-00
2381 ---------------------
2383 1. Add support for the /+ modifier to perltest (to output $` like it does in
2384 pcretest).
2386 2. Add support for the /g modifier to perltest.
2388 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2389 matches null strings.
2391 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2392 pattern. Perl treats empty patterns specially - it reuses the most recent
2393 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2394 effect.
2396 5. The POSIX interface was broken in that it was just handing over the POSIX
2397 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2398 required a bigger vector, with some working space on the end. This means that
2399 the POSIX wrapper now has to get and free some memory, and copy the results.
2401 6. Added some simple autoconf support, placing the test data and the
2402 documentation in separate directories, re-organizing some of the
2403 information files, and making it build pcre-config (a GNU standard). Also added
2404 libtool support for building PCRE as a shared library, which is now the
2405 default.
2407 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2408 09 are not valid octal constants. Single digits will be used for minor values
2409 less than 10.
2411 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2412 existing programs that set these in the POSIX interface can use PCRE without
2413 modification.
2415 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2416 return all that pcre_info() returns, plus additional data. The pcre_info()
2417 function is retained for compatibility, but is considered to be obsolete.
2419 10. Added experimental recursion feature (?R) to handle one common case that
2420 Perl 5.6 will be able to do with (?p{...}).
2422 11. Added support for POSIX character classes like [:alpha:], which Perl is
2423 adopting.
2426 Version 2.08 31-Aug-99
2427 ----------------------
2429 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2430 trying to match at the startoffset position, but instead was moving forward to
2431 the next newline as if a previous match had failed.
2433 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2434 and could get into a loop if a null string was matched other than at the start
2435 of the subject.
2437 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2438 be distinguished at compile time, and for completeness also added PCRE_DATE.
2440 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2441 in GnuWin32 environments.
2444 Version 2.07 29-Jul-99
2445 ----------------------
2447 1. The documentation is now supplied in plain text form and HTML as well as in
2448 the form of man page sources.
2450 2. C++ compilers don't like assigning (void *) values to other pointer types.
2451 In particular this affects malloc(). Although there is no problem in Standard
2452 C, I've put in casts to keep C++ compilers happy.
2454 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2455 should be (const char *).
2457 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2458 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2459 However, I haven't made this a standard facility. The documentation doesn't
2460 mention it, and the Makefile doesn't support it.
2462 5. The Makefile now contains an "install" target, with editable destinations at
2463 the top of the file. The pcretest program is not installed.
2465 6. pgrep -V now gives the PCRE version number and date.
2467 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2468 causing the entire string to be ignored, instead of just the last character.
2470 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2471 non-matching string, it can take a very, very long time, even for strings of
2472 quite modest length, because of the nested recursion. PCRE now does better in
2473 some of these cases. It does this by remembering the last required literal
2474 character in the pattern, and pre-searching the subject to ensure it is present
2475 before running the real match. In other words, it applies a heuristic to detect
2476 some types of certain failure quickly, and in the above example, if presented
2477 with a string that has no trailing " it gives "no match" very quickly.
2479 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2480 other alternatives are tried instead.
2483 Version 2.06 09-Jun-99
2484 ----------------------
2486 1. Change pcretest's output for amount of store used to show just the code
2487 space, because the remainder (the data block) varies in size between 32-bit and
2488 64-bit systems.
2490 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2491 start matching at. This allows lookbehinds to work when searching for multiple
2492 occurrences in a string.
2494 3. Added additional options to pcretest for testing multiple occurrences:
2496 /+ outputs the rest of the string that follows a match
2497 /g loops for multiple occurrences, using the new startoffset argument
2498 /G loops for multiple occurrences by passing an incremented pointer
2500 4. PCRE wasn't doing the "first character" optimization for patterns starting
2501 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2502 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2503 the letter 'x'. On long subject strings, this gives a significant speed-up.
2506 Version 2.05 21-Apr-99
2507 ----------------------
2509 1. Changed the type of magic_number from int to long int so that it works
2510 properly on 16-bit systems.
2512 2. Fixed a bug which caused patterns starting with .* not to work correctly
2513 when the subject string contained newline characters. PCRE was assuming
2514 anchoring for such patterns in all cases, which is not correct because .* will
2515 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2516 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2517 must be retried after every newline in the subject.
2520 Version 2.04 18-Feb-99
2521 ----------------------
2523 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2524 computation of the store needed to hold the pattern was incorrect (too large).
2525 If such patterns were nested a few deep, this could multiply and become a real
2526 problem.
2528 2. Added /M option to pcretest to show the memory requirement of a specific
2529 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2531 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2532 compiled in such a way that the backtracking after subsequent failure was
2533 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2534 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2537 Version 2.03 02-Feb-99
2538 ----------------------
2540 1. Fixed typo and small mistake in man page.
2542 2. Added 4th condition (GPL supersedes if conflict) and created separate
2543 LICENCE file containing the conditions.
2545 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2546 Perl, that is the internal \ allows the delimiter to be included in the
2547 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2548 the final delimiter, add \ to the end of the pattern (to test the error).
2550 4. Added the convenience functions for extracting substrings after a successful
2551 match. Updated pcretest to make it able to test these functions.
2554 Version 2.02 14-Jan-99
2555 ----------------------
2557 1. Initialized the working variables associated with each extraction so that
2558 their saving and restoring doesn't refer to uninitialized store.
2560 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2561 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2562 fix the problem.
2564 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2565 calls, and wasn't printing the correct value for compiling calls. Increased the
2566 default value of LOOPREPEAT, and the number of significant figures in the
2567 times.
2569 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2571 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2572 a building problem on Windows NT with a FAT file system.
2575 Version 2.01 21-Oct-98
2576 ----------------------
2578 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2579 to character tables built by pcre_maketables() in the current locale. If NULL
2580 is passed, the default tables are used.
2583 Version 2.00 24-Sep-98
2584 ----------------------
2586 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2587 it any more.
2589 2. Allow quantification of (?>) groups, and make it work correctly.
2591 3. The first character computation wasn't working for (?>) groups.
2593 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2594 end of the subject) and add 5.005's \z, which really does match only at the
2595 very end of the subject.
2597 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2599 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2600 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2601 localized options. All options to pcre_study() were also removed.
2603 7. Add other new features from 5.005:
2605 $(?<= positive lookbehind
2606 $(?<! negative lookbehind
2607 (?imsx-imsx) added the unsetting capability
2608 such a setting is global if at outer level; local otherwise
2609 (?imsx-imsx:) non-capturing groups with option setting
2610 (?(cond)re|re) conditional pattern matching
2612 A backreference to itself in a repeated group matches the previous
2613 captured string.
2615 8. General tidying up of studying (both automatic and via "study")
2616 consequential on the addition of new assertions.
2618 9. As in 5.005, unlimited repeated groups that could match an empty substring
2619 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2620 runtime if any iteration does actually match an empty substring.
2622 10. Include the RunTest script in the distribution.
2624 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2625 discrepancies, some of which were old and were also with respect to 5.004. They
2626 have now been fixed.
2629 Version 1.09 28-Apr-98
2630 ----------------------
2632 1. A negated single character class followed by a quantifier with a minimum
2633 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2634 program crashes, or just wrong answers. This did not apply to negated classes
2635 containing more than one character, or to minima other than one.
2638 Version 1.08 27-Mar-98
2639 ----------------------
2641 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2643 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2644 latter must appear before anything that relies on it in the pattern.
2647 Version 1.07 16-Feb-98
2648 ----------------------
2650 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2651 repeat of a potentially empty string).
2654 Version 1.06 23-Jan-98
2655 ----------------------
2657 1. Added Markus Oberhumer's little patches for C++.
2659 2. Literal strings longer than 255 characters were broken.
2662 Version 1.05 23-Dec-97
2663 ----------------------
2665 1. Negated character classes containing more than one character were failing if
2666 PCRE_CASELESS was set at run time.
2669 Version 1.04 19-Dec-97
2670 ----------------------
2672 1. Corrected the man page, where some "const" qualifiers had been omitted.
2674 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2675 input syntax.
2677 3. Fixed memory leak which occurred when a regex with back references was
2678 matched with an offsets vector that wasn't big enough. The temporary memory
2679 that is used in this case wasn't being freed if the match failed.
2681 4. Tidied pcretest to ensure it frees memory that it gets.
2683 5. Temporary memory was being obtained in the case where the passed offsets
2684 vector was exactly big enough.
2686 6. Corrected definition of offsetof() from change 5 below.
2688 7. I had screwed up change 6 below and broken the rules for the use of
2689 setjmp(). Now fixed.
2692 Version 1.03 18-Dec-97
2693 ----------------------
2695 1. A erroneous regex with a missing opening parenthesis was correctly
2696 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2697 on some systems.
2699 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2700 it was reported that one broken compiler failed on the former because "code" is
2701 also an independent variable.
2703 3. The erroneous regex a[]b caused an array overrun reference.
2705 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2706 fail on data ending with that character. (It was going on too far, and checking
2707 the next character, typically a binary zero.) This was specific to the
2708 optimized code for single-character negative classes.
2710 5. Added a contributed patch from the TIN world which does the following:
2712 + Add an undef for memmove, in case the the system defines a macro for it.
2714 + Add a definition of offsetof(), in case there isn't one. (I don't know
2715 the reason behind this - offsetof() is part of the ANSI standard - but
2716 it does no harm).
2718 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2719 most of the places where whitespace preceded '#'. I have given up and
2720 allowed the remaining 2 cases to be at the margin.
2722 + Rename some variables in pcre to eliminate shadowing. This seems very
2723 pedantic, but does no harm, of course.
2725 6. Moved the call to setjmp() into its own function, to get rid of warnings
2726 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2728 7. Constructs such as \d{8,} were compiling into the equivalent of
2729 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2730 outcome, but in this particular case used more store than had been allocated,
2731 which caused the bug to be discovered because it threw up an internal error.
2733 8. The debugging code in both pcre and pcretest for outputting the compiled
2734 form of a regex was going wrong in the case of back references followed by
2735 curly-bracketed repeats.
2738 Version 1.02 12-Dec-97
2739 ----------------------
2741 1. Typos in pcre.3 and comments in the source fixed.
2743 2. Applied a contributed patch to get rid of places where it used to remove
2744 'const' from variables, and fixed some signed/unsigned and uninitialized
2745 variable warnings.
2747 3. Added the "runtest" target to Makefile.
2749 4. Set default compiler flag to -O2 rather than just -O.
2752 Version 1.01 19-Nov-97
2753 ----------------------
2755 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2756 like /([ab]*)*/, that is, for classes with more than one character in them.
2758 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2759 as /((?>a*))*/ (a PCRE_EXTRA facility).
2762 Version 1.00 18-Nov-97
2763 ----------------------
2765 1. Added compile-time macros to support systems such as SunOS4 which don't have
2766 memmove() or strerror() but have other things that can be used instead.
2768 2. Arranged that "make clean" removes the executables.
2771 Version 0.99 27-Oct-97
2772 ----------------------
2774 1. Fixed bug in code for optimizing classes with only one character. It was
2775 initializing a 32-byte map regardless, which could cause it to run off the end
2776 of the memory it had got.
2778 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2781 Version 0.98 22-Oct-97
2782 ----------------------
2784 1. Fixed bug in code for handling temporary memory usage when there are more
2785 back references than supplied space in the ovector. This could cause segfaults.
2788 Version 0.97 21-Oct-97
2789 ----------------------
2791 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2793 2. Optimized negated single characters not to use a bit map.
2795 3. Brought error texts together as macro definitions; clarified some of them;
2796 fixed one that was wrong - it said "range out of order" when it meant "invalid
2797 escape sequence".
2799 4. Changed some char * arguments to const char *.
2801 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2803 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2804 pcretest.
2807 Version 0.96 16-Oct-97
2808 ----------------------
2810 1. Added a simple "pgrep" utility to the distribution.
2812 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2813 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2814 where "ddd" means "one or more decimal digits".
2816 3. Fixed serious bug. If a pattern had a back reference, but the call to
2817 pcre_exec() didn't supply a large enough ovector to record the related
2818 identifying subpattern, the match always failed. PCRE now remembers the number
2819 of the largest back reference, and gets some temporary memory in which to save
2820 the offsets during matching if necessary, in order to ensure that
2821 backreferences always work.
2823 4. Increased the compatibility with Perl in a number of ways:
2825 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2826 to request this handling. The option can be set at compile or exec time.
2828 (b) $ matches before a terminating newline by default; an option
2829 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2830 mode). The option can be set at compile or exec time.
2832 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2833 the same as Perl's. If the decimal number it represents is less than 10
2834 or there aren't that many previous left capturing parentheses, an octal
2835 escape is read. Inside a character class, it's always an octal escape,
2836 even if it is a single digit.
2838 (d) An escaped but undefined alphabetic character is taken as a literal,
2839 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2840 escapes.
2842 (e) {0} is now permitted. (The previous item is removed from the compiled
2843 pattern).
2845 5. Changed all the names of code files so that the basic parts are no longer
2846 than 10 characters, and abolished the teeny "globals.c" file.
2848 6. Changed the handling of character classes; they are now done with a 32-byte
2849 bit map always.
2851 7. Added the -d and /D options to pcretest to make it possible to look at the
2852 internals of compilation without having to recompile pcre.
2855 Version 0.95 23-Sep-97
2856 ----------------------
2858 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2859 \x20 at the start of a run of normal characters. These were being treated as
2860 real characters, instead of the source characters being re-checked.
2863 Version 0.94 18-Sep-97
2864 ----------------------
2866 1. The functions are now thread-safe, with the caveat that the global variables
2867 containing pointers to malloc() and free() or alternative functions are the
2868 same for all threads.
2870 2. Get pcre_study() to generate a bitmap of initial characters for non-
2871 anchored patterns when this is possible, and use it if passed to pcre_exec().
2874 Version 0.93 15-Sep-97
2875 ----------------------
2877 1. /(b)|(:+)/ was computing an incorrect first character.
2879 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2880 but not actually doing anything yet.
2882 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2883 as Perl does (e.g. [-az] or [az-]).
2885 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2886 all possible positions.
2888 5. Split up into different modules to avoid including unneeded functions in a
2889 compiled binary. However, compile and exec are still in one module. The "study"
2890 function is split off.
2892 6. The character tables are now in a separate module whose source is generated
2893 by an auxiliary program - but can then be edited by hand if required. There are
2894 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2895 toupper() in the code.
2897 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2898 make them global. Abolish the function for setting them, as the caller can now
2899 set them directly.
2902 Version 0.92 11-Sep-97
2903 ----------------------
2905 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2906 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2908 2. Caseless matching was not working in character classes if the characters in
2909 the pattern were in upper case.
2911 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2913 4. Make PCRE_ANCHORED public and accept as a compile option.
2915 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2916 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2917 pass them.
2919 6. Give an error if bad option bits passed at compile or run time.
2921 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2922 pcretest to cause it to pass that flag.
2924 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2925 options, and the first character, if set.
2927 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2930 Version 0.91 10-Sep-97
2931 ----------------------
2933 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2934 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2936 2. PCRE was looping on encountering an indefinitely repeated back reference to
2937 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2938 Perl does - treats the match as successful.
2940 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5