ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 275 - (show annotations)
Wed Nov 21 15:35:09 2007 UTC (13 years, 11 months ago) by ph10
File size: 143867 byte(s)
s/backslash/backspace in two comments where it was wrong.
1 ChangeLog for PCRE
2 ------------------
4 Version 7.5 12-Nov-07
5 ---------------------
7 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
8 values in parens when parsing an RE using the C++ wrapper."
10 2. Negative specials like \S did not work in character classes in UTF-8 mode.
11 Characters greater than 255 were excluded from the class instead of being
12 included.
14 3. The same bug as (2) above applied to negated POSIX classes such as
15 [:^space:].
17 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
18 defined or documented. It seems to have been a typo for PCRE_STATIC, so
19 I have changed it.
21 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
22 first named subpattern) and a construct such as (?&a) would reference the
23 first named subpattern whose name started with "a" (in other words, the
24 length check was missing). Both these problems are fixed. "Subpattern name
25 expected" is now given for (?&) (a zero-length name), and this patch also
26 makes it give the same error for \k'' (previously it complained that that
27 was a reference to a non-existent subpattern).
29 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
30 this is right because (?- can be followed by option settings as well as by
31 digits. I have, however, made the messages clearer.
33 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
34 than the number used in the conditional) now cause a compile-time error.
35 This is actually not compatible with Perl, which accepts such patterns, but
36 treats the conditional as always being FALSE (as PCRE used to), but it
37 seems to me that giving a diagnostic is better.
39 8. Change "alphameric" to the more common word "alphanumeric" in comments
40 and messages.
42 9. Fix two occurrences of "backslash" in comments that should have been
43 "backspace".
46 Version 7.4 21-Sep-07
47 ---------------------
49 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
50 means that a class such as [\s] counted as "explicit reference to CR or
51 LF". That isn't really right - the whole point of the change was to try to
52 help when there was an actual mention of one of the two characters. So now
53 the change happens only if \r or \n (or a literal CR or LF) character is
54 encountered.
56 2. The 32-bit options word was also used for 6 internal flags, but the numbers
57 of both had grown to the point where there were only 3 bits left.
58 Fortunately, there was spare space in the data structure, and so I have
59 moved the internal flags into a new 16-bit field to free up more option
60 bits.
62 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
63 but did not set the internal JCHANGED flag - either of these is enough to
64 control the way the "get" function works - but the PCRE_INFO_JCHANGED
65 facility is supposed to tell if (?J) was ever used, so now (?J) at the
66 start sets both bits.
68 4. Added options (at build time, compile time, exec time) to change \R from
69 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
71 5. doc/pcresyntax.html was missing from the distribution.
73 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
74 compatibility, even though it is no longer used.
76 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
77 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
78 windows.h file is present (where different names are used). [This was
79 reversed later after testing - see 16 below.]
81 8. Changed all #include <config.h> to #include "config.h". There were also
82 some further <pcre.h> cases that I changed to "pcre.h".
84 9. When pcregrep was used with the --colour option, it missed the line ending
85 sequence off the lines that it output.
87 10. It was pointed out to me that arrays of string pointers cause lots of
88 relocations when a shared library is dynamically loaded. A technique of
89 using a single long string with a table of offsets can drastically reduce
90 these. I have refactored PCRE in four places to do this. The result is
91 dramatic:
93 Originally: 290
94 After changing UCP table: 187
95 After changing error message table: 43
96 After changing table of "verbs" 36
97 After changing table of Posix names 22
99 Thanks to the folks working on Gregex for glib for this insight.
101 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
102 unicode-properties was also set.
104 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
106 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
107 checked only for CRLF.
109 14. Added casts to pcretest.c to avoid compiler warnings.
111 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
113 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
114 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
115 entirely. This removes changes made in 7 above.
117 17. The CMake files have been updated, and there is now more information about
118 building with CMake in the NON-UNIX-USE document.
121 Version 7.3 28-Aug-07
122 ---------------------
124 1. In the rejigging of the build system that eventually resulted in 7.1, the
125 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
126 brackets there is not right, since it causes compilers to look for an
127 installed pcre.h, not the version that is in the source that is being
128 compiled (which of course may be different). I have changed it back to:
130 #include "pcre.h"
132 I have a vague recollection that the change was concerned with compiling in
133 different directories, but in the new build system, that is taken care of
134 by the VPATH setting the Makefile.
136 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
137 when the subject happened to end in the byte 0x85 (e.g. if the last
138 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
139 characters but of course it shouldn't be taken as a newline when it is part
140 of another character. The bug was that, for an unlimited repeat of . in
141 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
142 characters when looking for a newline.
144 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
146 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
147 in debug output.
149 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
150 long printing in the pcrecpp unittest when running under MinGW.
152 6. ESC_K was left out of the EBCDIC table.
154 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
155 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
156 limit also applies to "virtual nesting" when a pattern is recursive, and in
157 this case 1000 isn't so big. I have been able to remove this limit at the
158 expense of backing off one optimization in certain circumstances. Normally,
159 when pcre_exec() would call its internal match() function recursively and
160 immediately return the result unconditionally, it uses a "tail recursion"
161 feature to save stack. However, when a subpattern that can match an empty
162 string has an unlimited repetition quantifier, it no longer makes this
163 optimization. That gives it a stack frame in which to save the data for
164 checking that an empty string has been matched. Previously this was taken
165 from the 1000-entry workspace that had been reserved. So now there is no
166 explicit limit, but more stack is used.
168 8. Applied Daniel's patches to solve problems with the import/export magic
169 syntax that is required for Windows, and which was going wrong for the
170 pcreposix and pcrecpp parts of the library. These were overlooked when this
171 problem was solved for the main library.
173 9. There were some crude static tests to avoid integer overflow when computing
174 the size of patterns that contain repeated groups with explicit upper
175 limits. As the maximum quantifier is 65535, the maximum group length was
176 set at 30,000 so that the product of these two numbers did not overflow a
177 32-bit integer. However, it turns out that people want to use groups that
178 are longer than 30,000 bytes (though not repeat them that many times).
179 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
180 made it possible to implement the integer overflow checks in a much more
181 dynamic way, which I have now done. The artificial limitation on group
182 length has been removed - we now have only the limit on the total length of
183 the compiled pattern, which depends on the LINK_SIZE setting.
185 10. Fixed a bug in the documentation for get/copy named substring when
186 duplicate names are permitted. If none of the named substrings are set, the
187 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
188 empty string.
190 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
191 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
192 because the ] is interpreted as the first data character and the
193 terminating ] is not found. PCRE has been made compatible with Perl in this
194 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
195 cause memory overwriting.
197 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
198 string has been matched (to stop an infinite loop). It was not recognizing
199 a conditional subpattern that could match an empty string if that
200 subpattern was within another subpattern. For example, it looped when
201 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
202 condition was not nested. This bug has been fixed.
204 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
205 past the start of the subject in the presence of bytes with the top bit
206 set, for example "\x8aBCD".
208 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
209 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
211 14. Optimized (?!) to (*FAIL).
213 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
214 This restricts code points to be within the range 0 to 0x10FFFF, excluding
215 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
216 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
217 does: it's just the validity check that is more restrictive.
219 16. Inserted checks for integer overflows during escape sequence (backslash)
220 processing, and also fixed erroneous offset values for syntax errors during
221 backslash processing.
223 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
224 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
226 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
227 caused an overrun.
229 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
230 something other than just ASCII characters) inside a group that had an
231 unlimited repeat caused a loop at compile time (while checking to see
232 whether the group could match an empty string).
234 20. Debugging a pattern containing \p or \P could cause a crash. For example,
235 [\P{Any}] did so. (Error in the code for printing property names.)
237 21. An orphan \E inside a character class could cause a crash.
239 22. A repeated capturing bracket such as (A)? could cause a wild memory
240 reference during compilation.
242 23. There are several functions in pcre_compile() that scan along a compiled
243 expression for various reasons (e.g. to see if it's fixed length for look
244 behind). There were bugs in these functions when a repeated \p or \P was
245 present in the pattern. These operators have additional parameters compared
246 with \d, etc, and these were not being taken into account when moving along
247 the compiled data. Specifically:
249 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
250 length.
252 (b) An item such as \pL+ within a repeated group could cause crashes or
253 loops.
255 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
256 "reference to non-existent subpattern" error.
258 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
260 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
261 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
263 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
265 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
266 character were causing crashes (broken optimization).
268 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
269 \p or \P) caused a compile-time loop.
271 28. More problems have arisen in unanchored patterns when CRLF is a valid line
272 break. For example, the unstudied pattern [\r\n]A does not match the string
273 "\r\nA" because change 7.0/46 below moves the current point on by two
274 characters after failing to match at the start. However, the pattern \nA
275 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
276 the same is true. There doesn't seem any very clean way out of this, but
277 what I have chosen to do makes the common cases work: PCRE now takes note
278 of whether there can be an explicit match for \r or \n anywhere in the
279 pattern, and if so, 7.0/46 no longer applies. As part of this change,
280 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
281 pattern has explicit CR or LF references.
283 29. Added (*CR) etc for changing newline setting at start of pattern.
286 Version 7.2 19-Jun-07
287 ---------------------
289 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
290 which is apparently normally available under Windows.
292 2. Re-jig the pcregrep tests with different newline settings in an attempt
293 to make them independent of the local environment's newline setting.
295 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
297 4. Some of the "internals" tests were previously cut out when the link size
298 was not 2, because the output contained actual offsets. The recent new
299 "Z" feature of pcretest means that these can be cut out, making the tests
300 usable with all link sizes.
302 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
303 stack recursion. This gives a massive performance boost under BSD, but just
304 a small improvement under Linux. However, it saves one field in the frame
305 in all cases.
307 6. Added more features from the forthcoming Perl 5.10:
309 (a) (?-n) (where n is a string of digits) is a relative subroutine or
310 recursion call. It refers to the nth most recently opened parentheses.
312 (b) (?+n) is also a relative subroutine call; it refers to the nth next
313 to be opened parentheses.
315 (c) Conditions that refer to capturing parentheses can be specified
316 relatively, for example, (?(-2)... or (?(+3)...
318 (d) \K resets the start of the current match so that everything before
319 is not part of it.
321 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
323 (f) \g{name} is another synonym - part of Perl 5.10's unification of
324 reference syntax.
326 (g) (?| introduces a group in which the numbering of parentheses in each
327 alternative starts with the same number.
329 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
331 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
334 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
335 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
336 for detecting groups that can match an empty string.
338 9. A pattern with a very large number of alternatives (more than several
339 hundred) was running out of internal workspace during the pre-compile
340 phase, where pcre_compile() figures out how much memory will be needed. A
341 bit of new cunning has reduced the workspace needed for groups with
342 alternatives. The 1000-alternative test pattern now uses 12 bytes of
343 workspace instead of running out of the 4096 that are available.
345 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
347 11. Applied patch from Google to remove an optimization that didn't quite work.
348 The report of the bug said:
350 pcrecpp::RE("a*").FullMatch("aaa") matches, while
351 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
352 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
354 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
355 it matched the wrong number of bytes.
358 Version 7.1 24-Apr-07
359 ---------------------
361 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
362 that is more "standard", making use of automake and other Autotools. There
363 is some re-arrangement of the files and adjustment of comments consequent
364 on this.
366 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
367 for recursive directory scanning broke on some systems because the files
368 are not scanned in any specific order and on different systems the order
369 was different. A call to "sort" has been inserted into RunGrepTest for the
370 approprate test as a short-term fix. In the longer term there may be an
371 alternative.
373 3. I had an email from Eric Raymond about problems translating some of PCRE's
374 man pages to HTML (despite the fact that I distribute HTML pages, some
375 people do their own conversions for various reasons). The problems
376 concerned the use of low-level troff macros .br and .in. I have therefore
377 removed all such uses from the man pages (some were redundant, some could
378 be replaced by .nf/.fi pairs). The 132html script that I use to generate
379 HTML has been updated to handle .nf/.fi and to complain if it encounters
380 .br or .in.
382 4. Updated comments in configure.ac that get placed in config.h.in and also
383 arranged for config.h to be included in the distribution, with the name
384 config.h.generic, for the benefit of those who have to compile without
385 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
387 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
388 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
389 makevp.bat for the current PCRE, using the additional files
390 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
392 6. A Windows user reported a minor discrepancy with test 2, which turned out
393 to be caused by a trailing space on an input line that had got lost in his
394 copy. The trailing space was an accident, so I've just removed it.
396 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
397 that is needed.
399 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
400 as "const" (a) because they are and (b) because it helps the PHP
401 maintainers who have recently made a script to detect big data structures
402 in the php code that should be moved to the .rodata section. I remembered
403 to update Builducptable as well, so it won't revert if ucptable.h is ever
404 re-created.
406 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
407 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
408 order to be able to cut out the UTF-8 tables in the latter when UTF-8
409 support is not required. This saves 1.5-2K of code, which is important in
410 some applications.
412 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
413 so as not to refer to the tables, even though these functions will never be
414 called when UTF-8 support is disabled. Otherwise there are problems with a
415 shared library.
417 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
419 (a) It was defining its arguments as char * instead of void *.
421 (b) It was assuming that all moves were upwards in memory; this was true
422 a long time ago when I wrote it, but is no longer the case.
424 The emulated memove() is provided for those environments that have neither
425 memmove() nor bcopy(). I didn't think anyone used it these days, but that
426 is clearly not the case, as these two bugs were recently reported.
428 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
429 and Detrail to create the HTML documentation, the .txt form of the man
430 pages, and it removes trailing spaces from listed files. It also creates
431 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
432 case, it wraps all the #defines with #ifndefs. This script should be run
433 before "make dist".
435 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
436 with Unicode property support.
438 (a) For a maximizing quantifier, if the two different cases of the
439 character were of different lengths in their UTF-8 codings (there are
440 some cases like this - I found 11), and the matching function had to
441 back up over a mixture of the two cases, it incorrectly assumed they
442 were both the same length.
444 (b) When PCRE was configured to use the heap rather than the stack for
445 recursion during matching, it was not correctly preserving the data for
446 the other case of a UTF-8 character when checking ahead for a match
447 while processing a minimizing repeat. If the check also involved
448 matching a wide character, but failed, corruption could cause an
449 erroneous result when trying to check for a repeat of the original
450 character.
452 13. Some tidying changes to the testing mechanism:
454 (a) The RunTest script now detects the internal link size and whether there
455 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
456 values substituted by "configure". (The RunGrepTest script already did
457 this for UTF-8.) The configure.ac script no longer substitutes the
458 relevant variables.
460 (b) The debugging options /B and /D in pcretest show the compiled bytecode
461 with length and offset values. This means that the output is different
462 for different internal link sizes. Test 2 is skipped for link sizes
463 other than 2 because of this, bypassing the problem. Unfortunately,
464 there was also a test in test 3 (the locale tests) that used /B and
465 failed for link sizes other than 2. Rather than cut the whole test out,
466 I have added a new /Z option to pcretest that replaces the length and
467 offset values with spaces. This is now used to make test 3 independent
468 of link size. (Test 2 will be tidied up later.)
470 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
471 segmentation fault instead of returning the appropriate error message.
473 15. In multiline mode when the newline sequence was set to "any", the pattern
474 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
475 This doesn't seem right; it now treats the CRLF combination as the line
476 ending, and so does not match in that case. It's only a pattern such as ^$
477 that would hit this one: something like ^ABC$ would have failed after \r
478 and then tried again after \r\n.
480 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
481 in an attempt to make files that differ only in their line terminators
482 compare equal. This works on Linux.
484 17. Under certain error circumstances pcregrep might try to free random memory
485 as it exited. This is now fixed, thanks to valgrind.
487 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
488 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
489 was because its rules for how to advance for /g after matching an empty
490 string at the end of a line did not allow for this case. They now check for
491 it specially.
493 20. pcretest is supposed to handle patterns and data of any length, by
494 extending its buffers when necessary. It was getting this wrong when the
495 buffer for a data line had to be extended.
497 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
498 CRLF as a newline sequence.
500 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
501 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
502 I have nevertheless tidied it up.
504 23. Added some casts to kill warnings from HP-UX ia64 compiler.
506 24. Added a man page for pcre-config.
509 Version 7.0 19-Dec-06
510 ---------------------
512 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
513 moving to gcc 4.1.1.
515 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
516 sys/time.h, which is documented as needed for this function. It doesn't
517 seem to matter on Linux, but it showed up on some releases of OS X.
519 3. It seems that there are systems where bytes whose values are greater than
520 127 match isprint() in the "C" locale. The "C" locale should be the
521 default when a C program starts up. In most systems, only ASCII printing
522 characters match isprint(). This difference caused the output from pcretest
523 to vary, making some of the tests fail. I have changed pcretest so that:
525 (a) When it is outputting text in the compiled version of a pattern, bytes
526 other than 32-126 are always shown as hex escapes.
528 (b) When it is outputting text that is a matched part of a subject string,
529 it does the same, unless a different locale has been set for the match
530 (using the /L modifier). In this case, it uses isprint() to decide.
532 4. Fixed a major bug that caused incorrect computation of the amount of memory
533 required for a compiled pattern when options that changed within the
534 pattern affected the logic of the preliminary scan that determines the
535 length. The relevant options are -x, and -i in UTF-8 mode. The result was
536 that the computed length was too small. The symptoms of this bug were
537 either the PCRE error "internal error: code overflow" from pcre_compile(),
538 or a glibc crash with a message such as "pcretest: free(): invalid next
539 size (fast)". Examples of patterns that provoked this bug (shown in
540 pcretest format) are:
542 /(?-x: )/x
543 /(?x)(?-x: \s*#\s*)/
544 /((?i)[\x{c0}])/8
545 /(?i:[\x{c0}])/8
547 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
548 is now done differently.
550 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
551 wrapper classes; (b) implement a new function in the C++ scanner that is
552 more efficient than the old way of doing things because it avoids levels of
553 recursion in the regex matching; (c) add a paragraph to the documentation
554 for the FullMatch() function.
556 6. The escape sequence \n was being treated as whatever was defined as
557 "newline". Not only was this contrary to the documentation, which states
558 that \n is character 10 (hex 0A), but it also went horribly wrong when
559 "newline" was defined as CRLF. This has been fixed.
561 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
562 was being set to -1 for the "end of line" case (supposedly a value that no
563 character can have). Though this value is never used (the check for end of
564 line is "zero bytes in current character"), it caused compiler complaints.
565 I've changed it to 0xffffffff.
567 8. In pcre_version.c, the version string was being built by a sequence of
568 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
569 string (as it is for production releases) called a macro with an empty
570 argument. The C standard says the result of this is undefined. The gcc
571 compiler treats it as an empty string (which was what was wanted) but it is
572 reported that Visual C gives an error. The source has been hacked around to
573 avoid this problem.
575 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
576 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
577 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
578 of them did).
580 10. Originally, pcretest opened its input and output without "b"; then I was
581 told that "b" was needed in some environments, so it was added for release
582 5.0 to both the input and output. (It makes no difference on Unix-like
583 systems.) Later I was told that it is wrong for the input on Windows. I've
584 now abstracted the modes into two macros, to make it easier to fiddle with
585 them, and removed "b" from the input mode under Windows.
587 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
589 12. Added -help and --help to pcretest as an official way of being reminded
590 of the options.
592 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
593 and pcrecpp.cc because they annoy compilers at high warning levels.
595 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
597 15. Fixed an occurrence of == in configure.ac that should have been = (shell
598 scripts are not C programs :-) and which was not noticed because it works
599 on Linux.
601 16. pcretest is supposed to handle any length of pattern and data line (as one
602 line or as a continued sequence of lines) by extending its input buffer if
603 necessary. This feature was broken for very long pattern lines, leading to
604 a string of junk being passed to pcre_compile() if the pattern was longer
605 than about 50K.
607 17. I have done a major re-factoring of the way pcre_compile() computes the
608 amount of memory needed for a compiled pattern. Previously, there was code
609 that made a preliminary scan of the pattern in order to do this. That was
610 OK when PCRE was new, but as the facilities have expanded, it has become
611 harder and harder to keep it in step with the real compile phase, and there
612 have been a number of bugs (see for example, 4 above). I have now found a
613 cunning way of running the real compile function in a "fake" mode that
614 enables it to compute how much memory it would need, while actually only
615 ever using a few hundred bytes of working memory and without too many
616 tests of the mode. This should make future maintenance and development
617 easier. A side effect of this work is that the limit of 200 on the nesting
618 depth of parentheses has been removed (though this was never a serious
619 limitation, I suspect). However, there is a downside: pcre_compile() now
620 runs more slowly than before (30% or more, depending on the pattern). I
621 hope this isn't a big issue. There is no effect on runtime performance.
623 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
624 newline (only possible for the last line of a file) and it was a
625 pattern that set a locale (followed by /Lsomething), pcretest crashed.
627 19. Added additional timing features to pcretest. (1) The -tm option now times
628 matching only, not compiling. (2) Both -t and -tm can be followed, as a
629 separate command line item, by a number that specifies the number of
630 repeats to use when timing. The default is 50000; this gives better
631 precision, but takes uncomfortably long for very large patterns.
633 20. Extended pcre_study() to be more clever in cases where a branch of a
634 subpattern has no definite first character. For example, (a*|b*)[cd] would
635 previously give no result from pcre_study(). Now it recognizes that the
636 first character must be a, b, c, or d.
638 21. There was an incorrect error "recursive call could loop indefinitely" if
639 a subpattern (or the entire pattern) that was being tested for matching an
640 empty string contained only one non-empty item after a nested subpattern.
641 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
642 incorrectly, because the \d was being skipped in the check.
644 22. The pcretest program now has a new pattern option /B and a command line
645 option -b, which is equivalent to adding /B to every pattern. This causes
646 it to show the compiled bytecode, without the additional information that
647 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
648 is the same as /B/I).
650 23. A new optimization is now able automatically to treat some sequences such
651 as a*b as a*+b. More specifically, if something simple (such as a character
652 or a simple class like \d) has an unlimited quantifier, and is followed by
653 something that cannot possibly match the quantified thing, the quantifier
654 is automatically "possessified".
656 24. A recursive reference to a subpattern whose number was greater than 39
657 went wrong under certain circumstances in UTF-8 mode. This bug could also
658 have affected the operation of pcre_study().
660 25. Realized that a little bit of performance could be had by replacing
661 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
663 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
665 27. Possessive quantifiers such as a++ were previously implemented by turning
666 them into atomic groups such as ($>a+). Now they have their own opcodes,
667 which improves performance. This includes the automatically created ones
668 from 23 above.
670 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
671 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
672 the first matched character to be a colon. This applied both to named and
673 numbered groups.
675 29. The ucpinternal.h header file was missing its idempotency #ifdef.
677 30. I was sent a "project" file called libpcre.a.dev which I understand makes
678 building PCRE on Windows easier, so I have included it in the distribution.
680 31. There is now a check in pcretest against a ridiculously large number being
681 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
682 loop, the loop is abandoned.
684 32. Forward references to subpatterns in conditions such as (?(2)...) where
685 subpattern 2 is defined later cause pcre_compile() to search forwards in
686 the pattern for the relevant set of parentheses. This search went wrong
687 when there were unescaped parentheses in a character class, parentheses
688 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
690 33. "Subroutine" calls and backreferences were previously restricted to
691 referencing subpatterns earlier in the regex. This restriction has now
692 been removed.
694 34. Added a number of extra features that are going to be in Perl 5.10. On the
695 whole, these are just syntactic alternatives for features that PCRE had
696 previously implemented using the Python syntax or my own invention. The
697 other formats are all retained for compatibility.
699 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
700 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
701 also .NET compatible.
703 (b) A recursion or subroutine call to a named group can now be defined as
704 (?&name) as well as (?P>name).
706 (c) A backreference to a named group can now be defined as \k<name> or
707 \k'name' as well as (?P=name). The new forms, as well as being in Perl
708 5.10, are also .NET compatible.
710 (d) A conditional reference to a named group can now use the syntax
711 (?(<name>) or (?('name') as well as (?(name).
713 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
714 groups (named and numbered) that are never evaluated inline, but can be
715 called as "subroutines" from elsewhere. In effect, the DEFINE condition
716 is always false. There may be only one alternative in such a group.
718 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
719 as the simple (?(R). The condition is true only if the most recent
720 recursion is that of the given number or name. It does not search out
721 through the entire recursion stack.
723 (g) The escape \gN or \g{N} has been added, where N is a positive or
724 negative number, specifying an absolute or relative reference.
726 35. Tidied to get rid of some further signed/unsigned compiler warnings and
727 some "unreachable code" warnings.
729 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
730 things, this adds five new scripts.
732 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
733 There were also incompatibilities regarding the handling of \Q..\E inside
734 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
735 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
737 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
738 matches an empty string, and forcibly breaks the loop. There were bugs in
739 this code in non-simple cases. For a pattern such as ^(a()*)* matched
740 against aaaa the result was just "a" rather than "aaaa", for example. Two
741 separate and independent bugs (that affected different cases) have been
742 fixed.
744 39. Refactored the code to abolish the use of different opcodes for small
745 capturing bracket numbers. This is a tidy that I avoided doing when I
746 removed the limit on the number of capturing brackets for 3.5 back in 2001.
747 The new approach is not only tidier, it makes it possible to reduce the
748 memory needed to fix the previous bug (38).
750 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
751 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
752 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
753 mode.
755 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
756 report.
758 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
759 copy construction and assignment in the C++ wrapper.
761 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
762 couple of bugs that could have given wrong results in the "--newline=crlf"
763 case.
765 44. Added a number of casts and did some reorganization of signed/unsigned int
766 variables following suggestions from Dair Grant. Also renamed the variable
767 "this" as "item" because it is a C++ keyword.
769 45. Arranged for dftables to add
771 #include "pcre_internal.h"
773 to pcre_chartables.c because without it, gcc 4.x may remove the array
774 definition from the final binary if PCRE is built into a static library and
775 dead code stripping is activated.
777 46. For an unanchored pattern, if a match attempt fails at the start of a
778 newline sequence, and the newline setting is CRLF or ANY, and the next two
779 characters are CRLF, advance by two characters instead of one.
782 Version 6.7 04-Jul-06
783 ---------------------
785 1. In order to handle tests when input lines are enormously long, pcretest has
786 been re-factored so that it automatically extends its buffers when
787 necessary. The code is crude, but this _is_ just a test program. The
788 default size has been increased from 32K to 50K.
790 2. The code in pcre_study() was using the value of the re argument before
791 testing it for NULL. (Of course, in any sensible call of the function, it
792 won't be NULL.)
794 3. The memmove() emulation function in pcre_internal.h, which is used on
795 systems that lack both memmove() and bcopy() - that is, hardly ever -
796 was missing a "static" storage class specifier.
798 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
799 containing an extended class (one that cannot be represented by a bitmap
800 because it contains high-valued characters or Unicode property items, e.g.
801 [\pZ]). Almost always one would set UTF-8 mode when processing such a
802 pattern, but PCRE should not loop if you do not (it no longer does).
803 [Detail: two cases were found: (a) a repeated subpattern containing an
804 extended class; (b) a recursive reference to a subpattern that followed a
805 previous extended class. It wasn't skipping over the extended class
806 correctly when UTF-8 mode was not set.]
808 5. A negated single-character class was not being recognized as fixed-length
809 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
810 compile error "lookbehind assertion is not fixed length".
812 6. The RunPerlTest auxiliary script was showing an unexpected difference
813 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
814 write a Perl script that can interpret lines of an input file either as
815 byte characters or as UTF-8, which is what "perltest" was being required to
816 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
817 can't do is switch easily at run time between having the "use utf8;" pragma
818 or not. In the end, I fudged it by using the RunPerlTest script to insert
819 "use utf8;" explicitly for the UTF-8 tests.
821 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
822 the end of the subject string, contrary to the documentation and to what
823 Perl does. This was true of both matching functions. Now it matches only at
824 the start of the subject and immediately after *internal* newlines.
826 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
827 a pointer to an int instead of a pointer to an unsigned long int. This
828 caused problems on 64-bit systems.
830 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
831 instance of the 'standard' template library not being so standard".
833 10. There was no check on the number of named subpatterns nor the maximum
834 length of a subpattern name. The product of these values is used to compute
835 the size of the memory block for a compiled pattern. By supplying a very
836 long subpattern name and a large number of named subpatterns, the size
837 computation could be caused to overflow. This is now prevented by limiting
838 the length of names to 32 characters, and the number of named subpatterns
839 to 10,000.
841 11. Subpatterns that are repeated with specific counts have to be replicated in
842 the compiled pattern. The size of memory for this was computed from the
843 length of the subpattern and the repeat count. The latter is limited to
844 65535, but there was no limit on the former, meaning that integer overflow
845 could in principle occur. The compiled length of a repeated subpattern is
846 now limited to 30,000 bytes in order to prevent this.
848 12. Added the optional facility to have named substrings with the same name.
850 13. Added the ability to use a named substring as a condition, using the
851 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
852 are numbers (not recommended). Forward references are permitted.
854 14. Added forward references in named backreferences (if you see what I mean).
856 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
857 pattern could run off the end of the subject. For example, the pattern
858 "(?s)(.{1,5})"8 did this with the subject "ab".
860 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
861 PCRE_CASELESS was set when matching characters that were quantified with ?
862 or *.
864 17. A character class other than a single negated character that had a minimum
865 but no maximum quantifier - for example [ab]{6,} - was not handled
866 correctly by pce_dfa_exec(). It would match only one character.
868 18. A valid (though odd) pattern that looked like a POSIX character
869 class but used an invalid character after [ (for example [[,abc,]]) caused
870 pcre_compile() to give the error "Failed: internal error: code overflow" or
871 in some cases to crash with a glibc free() error. This could even happen if
872 the pattern terminated after [[ but there just happened to be a sequence of
873 letters, a binary zero, and a closing ] in the memory that followed.
875 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
876 over the years. Originally (before any Unicode support), just the bottom 8
877 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
878 output from "man perlunicode" includes this:
880 The regular expression compiler produces polymorphic opcodes. That
881 is, the pattern adapts to the data and automatically switches to
882 the Unicode character scheme when presented with Unicode data--or
883 instead uses a traditional byte scheme when presented with byte
884 data.
886 Sadly, a wide octal escape does not cause a switch, and in a string with
887 no other multibyte characters, these octal escapes are treated as before.
888 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
889 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
890 Unicode string.
892 I have not perpetrated such confusion in PCRE. Up till now, it took just
893 the bottom 8 bits, as in old Perl. I have now made octal escapes with
894 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
895 translate to the appropriate multibyte character.
897 29. Applied some refactoring to reduce the number of warnings from Microsoft
898 and Borland compilers. This has included removing the fudge introduced
899 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
900 a warning about an unused variable.
902 21. PCRE has not included VT (character 0x0b) in the set of whitespace
903 characters since release 4.0, because Perl (from release 5.004) does not.
904 [Or at least, is documented not to: some releases seem to be in conflict
905 with the documentation.] However, when a pattern was studied with
906 pcre_study() and all its branches started with \s, PCRE still included VT
907 as a possible starting character. Of course, this did no harm; it just
908 caused an unnecessary match attempt.
910 22. Removed a now-redundant internal flag bit that recorded the fact that case
911 dependency changed within the pattern. This was once needed for "required
912 byte" processing, but is no longer used. This recovers a now-scarce options
913 bit. Also moved the least significant internal flag bit to the most-
914 significant bit of the word, which was not previously used (hangover from
915 the days when it was an int rather than a uint) to free up another bit for
916 the future.
918 23. Added support for CRLF line endings as well as CR and LF. As well as the
919 default being selectable at build time, it can now be changed at runtime
920 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
921 specify that it is scanning data with non-default line endings.
923 24. Changed the definition of CXXLINK to make it agree with the definition of
924 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
926 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
927 recursions. This makes a big different to stack usage for some patterns.
929 26. If a subpattern containing a named recursion or subroutine reference such
930 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
931 the space required for the compiled pattern went wrong and gave too small a
932 value. Depending on the environment, this could lead to "Failed: internal
933 error: code overflow at offset 49" or "glibc detected double free or
934 corruption" errors.
936 27. Applied patches from Google (a) to support the new newline modes and (b) to
937 advance over multibyte UTF-8 characters in GlobalReplace.
939 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
940 difference for some implementation of PCRE in some Windows version.
942 29. Added some extra testing facilities to pcretest:
944 \q<number> in a data line sets the "match limit" value
945 \Q<number> in a data line sets the "match recursion limt" value
946 -S <number> sets the stack size, where <number> is in megabytes
948 The -S option isn't available for Windows.
951 Version 6.6 06-Feb-06
952 ---------------------
954 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
955 in pcreposix.h. I have copied the definition from pcre.h.
957 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
958 because pcre.h is no longer a built file.
960 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
961 not normally included in the compiled code.
964 Version 6.5 01-Feb-06
965 ---------------------
967 1. When using the partial match feature with pcre_dfa_exec(), it was not
968 anchoring the second and subsequent partial matches at the new starting
969 point. This could lead to incorrect results. For example, with the pattern
970 /1234/, partially matching against "123" and then "a4" gave a match.
972 2. Changes to pcregrep:
974 (a) All non-match returns from pcre_exec() were being treated as failures
975 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
976 error message is output. Some extra information is given for the
978 probably the only errors that are likely to be caused by users (by
979 specifying a regex that has nested indefinite repeats, for instance).
980 If there are more than 20 of these errors, pcregrep is abandoned.
982 (b) A binary zero was treated as data while matching, but terminated the
983 output line if it was written out. This has been fixed: binary zeroes
984 are now no different to any other data bytes.
986 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
987 used to set a locale for matching. The --locale=xxxx long option has
988 been added (no short equivalent) to specify a locale explicitly on the
989 pcregrep command, overriding the environment variables.
991 (d) When -B was used with -n, some line numbers in the output were one less
992 than they should have been.
994 (e) Added the -o (--only-matching) option.
996 (f) If -A or -C was used with -c (count only), some lines of context were
997 accidentally printed for the final match.
999 (g) Added the -H (--with-filename) option.
1001 (h) The combination of options -rh failed to suppress file names for files
1002 that were found from directory arguments.
1004 (i) Added the -D (--devices) and -d (--directories) options.
1006 (j) Added the -F (--fixed-strings) option.
1008 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1010 (l) Added the --colo(u)r option.
1012 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1013 is not present by default.
1015 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1016 items such as (?R) or (?1), when the recursion could match a number of
1017 alternatives. If it matched one of the alternatives, but subsequently,
1018 outside the recursion, there was a failure, the code tried to back up into
1019 the recursion. However, because of the way PCRE is implemented, this is not
1020 possible, and the result was an incorrect result from the match.
1022 In order to prevent this happening, the specification of recursion has
1023 been changed so that all such subpatterns are automatically treated as
1024 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1026 4. I had overlooked the fact that, in some locales, there are characters for
1027 which isalpha() is true but neither isupper() nor islower() are true. In
1028 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1029 and ordfeminine) are like this. This affected the treatment of \w and \W
1030 when they appeared in character classes, but not when they appeared outside
1031 a character class. The bit map for "word" characters is now created
1032 separately from the results of isalnum() instead of just taking it from the
1033 upper, lower, and digit maps. (Plus the underscore character, of course.)
1035 5. The above bug also affected the handling of POSIX character classes such as
1036 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1037 permanent tables. Instead, the bit maps for such a class were previously
1038 created as the appropriate unions of the upper, lower, and digit bitmaps.
1039 Now they are created by subtraction from the [[:word:]] class, which has
1040 its own bitmap.
1042 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1043 It is created by subtracting the vertical space characters (\x09, \x0a,
1044 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1045 subtraction was done in the overall bitmap for a character class, meaning
1046 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1047 be recognized. This bug has been fixed.
1049 7. Patches from the folks at Google:
1051 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1052 real life, but is still worth protecting against".
1054 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1055 regular expressions".
1057 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1058 have it.
1060 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1061 "configure" and the latter not, in order to fix a problem somebody had
1062 with compiling the Arg class on HP-UX.
1064 (e) Improve the error-handling of the C++ wrapper a little bit.
1066 (f) New tests for checking recursion limiting.
1068 8. The pcre_memmove() function, which is used only if the environment does not
1069 have a standard memmove() function (and is therefore rarely compiled),
1070 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1071 returning a result (though PCRE never actually uses the result).
1073 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1074 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1075 returned instead of calling malloc() with an overflowing number that would
1076 most likely cause subsequent chaos.
1078 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1080 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1081 with this option is matched, the nmatch and pmatch options of regexec() are
1082 ignored.
1084 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1085 provided in case anyone wants to the the POSIX interface with UTF-8
1086 strings.
1088 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1089 C++ linking (needed for some HP-UX environments).
1091 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1092 (unused parameter) and in the pcre_printint() function (omitted "default"
1093 switch label when the default is to do nothing).
1095 15. Added some code to make it possible, when PCRE is compiled as a C++
1096 library, to replace subject pointers for pcre_exec() with a smart pointer
1097 class, thus making it possible to process discontinuous strings.
1099 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1100 much the same function. They were added by different people who were trying
1101 to make PCRE easy to compile on non-Unix systems. It has been suggested
1102 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1103 for compiling on Windows systems. I have therefore replaced it with
1104 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1105 defaults to "extern" for C or "extern C" for C++, which works fine on
1106 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1107 SCOPE with something explicit in config.h. In addition:
1109 (a) pcreposix.h still had just "extern" instead of either of these macros;
1110 I have replaced it with PCRE_DATA_SCOPE.
1112 (b) Functions such as _pcre_xclass(), which are internal to the library,
1113 but external in the C sense, all had PCRE_EXPORT in their definitions.
1114 This is apparently wrong for the Windows case, so I have removed it.
1115 (It makes no difference on Unix-like systems.)
1117 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1118 of recursive calls to match(). This is different to MATCH_LIMIT because
1119 that limits the total number of calls to match(), not all of which increase
1120 the depth of recursion. Limiting the recursion depth limits the amount of
1121 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1122 when PCRE is compiled, and changed at run time. A patch from Google adds
1123 this functionality to the C++ interface.
1125 18. Changes to the handling of Unicode character properties:
1127 (a) Updated the table to Unicode 4.1.0.
1129 (b) Recognize characters that are not in the table as "Cn" (undefined).
1131 (c) I revised the way the table is implemented to a much improved format
1132 which includes recognition of ranges. It now supports the ranges that
1133 are defined in UnicodeData.txt, and it also amalgamates other
1134 characters into ranges. This has reduced the number of entries in the
1135 table from around 16,000 to around 3,000, thus reducing its size
1136 considerably. I realized I did not need to use a tree structure after
1137 all - a binary chop search is just as efficient. Having reduced the
1138 number of entries, I extended their size from 6 bytes to 8 bytes to
1139 allow for more data.
1141 (d) Added support for Unicode script names via properties such as \p{Han}.
1143 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1144 matching that character.
1146 20. When matching a repeated Unicode property with a minimum greater than zero,
1147 (for example \pL{2,}), PCRE could look past the end of the subject if it
1148 reached it while seeking the minimum number of characters. This could
1149 happen only if some of the characters were more than one byte long, because
1150 there is a check for at least the minimum number of bytes.
1152 21. Refactored the implementation of \p and \P so as to be more general, to
1153 allow for more different types of property in future. This has changed the
1154 compiled form incompatibly. Anybody with saved compiled patterns that use
1155 \p or \P will have to recompile them.
1157 22. Added "Any" and "L&" to the supported property types.
1159 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1160 but give a compile time error if the value is greater than 0xff.
1162 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1163 accidentally not being installed or uninstalled.
1165 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1166 made were to insert the current release number. This seemed silly, because
1167 it made things harder for people building PCRE on systems that don't run
1168 "configure". I have turned pcre.h into a distributed file, no longer built
1169 by "configure", with the version identification directly included. There is
1170 no longer a pcre.h.in file.
1172 However, this change necessitated a change to the pcre-config script as
1173 well. It is built from pcre-config.in, and one of the substitutions was the
1174 release number. I have updated configure.ac so that ./configure now finds
1175 the release number by grepping pcre.h.
1177 26. Added the ability to run the tests under valgrind.
1180 Version 6.4 05-Sep-05
1181 ---------------------
1183 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1184 "--" to be printed when multiple files were scanned, even when none of the
1185 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1186 consider it to be a bug, and have restored the previous behaviour.
1188 2. A couple of code tidies to get rid of compiler warnings.
1190 3. The pcretest program used to cheat by referring to symbols in the library
1191 whose names begin with _pcre_. These are internal symbols that are not
1192 really supposed to be visible externally, and in some environments it is
1193 possible to suppress them. The cheating is now confined to including
1194 certain files from the library's source, which is a bit cleaner.
1196 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1197 file's purpose clearer.
1199 5. Reorganized pcre_ucp_findchar().
1202 Version 6.3 15-Aug-05
1203 ---------------------
1205 1. The file libpcre.pc.in did not have general read permission in the tarball.
1207 2. There were some problems when building without C++ support:
1209 (a) If C++ support was not built, "make install" and "make test" still
1210 tried to test it.
1212 (b) There were problems when the value of CXX was explicitly set. Some
1213 changes have been made to try to fix these, and ...
1215 (c) --disable-cpp can now be used to explicitly disable C++ support.
1217 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1218 backslash in a target when C++ was disabled. This confuses some
1219 versions of "make", apparently. Using an intermediate variable solves
1220 this. (Same for CPP_LOBJ.)
1222 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1223 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1224 necessary on certain architectures.
1226 4. Added a setting of -export-symbols-regex to the link command to remove
1227 those symbols that are exported in the C sense, but actually are local
1228 within the library, and not documented. Their names all begin with
1229 "_pcre_". This is not a perfect job, because (a) we have to except some
1230 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1231 available (and never for static libraries). I have made a note to try to
1232 find a way round (a) in the future.
1235 Version 6.2 01-Aug-05
1236 ---------------------
1238 1. There was no test for integer overflow of quantifier values. A construction
1239 such as {1111111111111111} would give undefined results. What is worse, if
1240 a minimum quantifier for a parenthesized subpattern overflowed and became
1241 negative, the calculation of the memory size went wrong. This could have
1242 led to memory overwriting.
1244 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1246 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1247 operating environments where this matters.
1249 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1250 PCRE options from within the C++ wrapper.
1252 5. Named capturing subpatterns were not being correctly counted when a pattern
1253 was compiled. This caused two problems: (a) If there were more than 100
1254 such subpatterns, the calculation of the memory needed for the whole
1255 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1256 back references of the form \12, where the number was greater than 9, were
1257 not recognized as back references, even though there were sufficient
1258 previous subpatterns.
1260 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1261 versions of gcc, e.g. 2.95.4.
1264 Version 6.1 21-Jun-05
1265 ---------------------
1267 1. There was one reference to the variable "posix" in pcretest.c that was not
1268 surrounded by "#if !defined NOPOSIX".
1270 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1271 the cross-check on the old pcre_info() function, for the benefit of the
1272 cut-down version of PCRE that is currently imported into Exim.
1274 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1275 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1276 patterns that start (?i) but I don't think that matters. The use of (?i) is
1277 just an example; this all applies to the other options as well.
1279 4. Since libtool seems to echo the compile commands it is issuing, the output
1280 from "make" can be reduced a bit by putting "@" in front of each libtool
1281 compile command.
1283 5. Patch from the folks at Google for configure.in to be a bit more thorough
1284 in checking for a suitable C++ installation before trying to compile the
1285 C++ stuff. This should fix a reported problem when a compiler was present,
1286 but no suitable headers.
1288 6. The man pages all had just "PCRE" as their title. I have changed them to
1289 be the relevant file name. I have also arranged that these names are
1290 retained in the file doc/pcre.txt, which is a concatenation in text format
1291 of all the man pages except the little individual ones for each function.
1293 7. The NON-UNIX-USE file had not been updated for the different set of source
1294 files that come with release 6. I also added a few comments about the C++
1295 wrapper.
1298 Version 6.0 07-Jun-05
1299 ---------------------
1301 1. Some minor internal re-organization to help with my DFA experiments.
1303 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1304 didn't matter for the library itself when fully configured, but did matter
1305 when compiling without UCP support, or within Exim, where the ucp files are
1306 not imported.
1308 3. Refactoring of the library code to split up the various functions into
1309 different source modules. The addition of the new DFA matching code (see
1310 below) to a single monolithic source would have made it really too
1311 unwieldy, quite apart from causing all the code to be include in a
1312 statically linked application, when only some functions are used. This is
1313 relevant even without the DFA addition now that patterns can be compiled in
1314 one application and matched in another.
1316 The downside of splitting up is that there have to be some external
1317 functions and data tables that are used internally in different modules of
1318 the library but which are not part of the API. These have all had their
1319 names changed to start with "_pcre_" so that they are unlikely to clash
1320 with other external names.
1322 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1323 a different (DFA) algorithm. Although it is slower than the original
1324 function, it does have some advantages for certain types of matching
1325 problem.
1327 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1328 including restarting after a partial match.
1330 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1331 defined when compiling for Windows was sent to me. I have put it into the
1332 code, though I have no means of testing or verifying it.
1334 7. Added the pcre_refcount() auxiliary function.
1336 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1337 match before or at the first newline in the subject string. In pcretest,
1338 the /f option on a pattern can be used to set this.
1340 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1341 would behave wrongly. This has been present in PCRE since release 4.0.
1343 10. A number of changes to the pcregrep command:
1345 (a) Refactored how -x works; insert ^(...)$ instead of setting
1346 PCRE_ANCHORED and checking the length, in preparation for adding
1347 something similar for -w.
1349 (b) Added the -w (match as a word) option.
1351 (c) Refactored the way lines are read and buffered so as to have more
1352 than one at a time available.
1354 (d) Implemented a pcregrep test script.
1356 (e) Added the -M (multiline match) option. This allows patterns to match
1357 over several lines of the subject. The buffering ensures that at least
1358 8K, or the rest of the document (whichever is the shorter) is available
1359 for matching (and similarly the previous 8K for lookbehind assertions).
1361 (f) Changed the --help output so that it now says
1363 -w, --word-regex(p)
1365 instead of two lines, one with "regex" and the other with "regexp"
1366 because that confused at least one person since the short forms are the
1367 same. (This required a bit of code, as the output is generated
1368 automatically from a table. It wasn't just a text change.)
1370 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1371 option but starts with a hyphen. Could be a pattern or a path name
1372 starting with a hyphen, for instance.
1374 (h) "-" can be given as a file name to represent stdin.
1376 (i) When file names are being printed, "(standard input)" is used for
1377 the standard input, for compatibility with GNU grep. Previously
1378 "<stdin>" was used.
1380 (j) The option --label=xxx can be used to supply a name to be used for
1381 stdin when file names are being printed. There is no short form.
1383 (k) Re-factored the options decoding logic because we are going to add
1384 two more options that take data. Such options can now be given in four
1385 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1387 (l) Added the -A, -B, and -C options for requesting that lines of context
1388 around matches be printed.
1390 (m) Added the -L option to print the names of files that do not contain
1391 any matching lines, that is, the complement of -l.
1393 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1394 continue to scan other files.
1396 (o) The -s option was incorrectly implemented. For compatibility with other
1397 greps, it now suppresses the error message for a non-existent or non-
1398 accessible file (but not the return code). There is a new option called
1399 -q that suppresses the output of matching lines, which was what -s was
1400 previously doing.
1402 (p) Added --include and --exclude options to specify files for inclusion
1403 and exclusion when recursing.
1405 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1406 Hopefully, it now does.
1408 12. Missing cast in pcre_study().
1410 13. Added an "uninstall" target to the makefile.
1412 14. Replaced "extern" in the function prototypes in Makefile.in with
1413 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1414 world, but is set differently for Windows.
1416 15. Added a second compiling function called pcre_compile2(). The only
1417 difference is that it has an extra argument, which is a pointer to an
1418 integer error code. When there is a compile-time failure, this is set
1419 non-zero, in addition to the error test pointer being set to point to an
1420 error message. The new argument may be NULL if no error number is required
1421 (but then you may as well call pcre_compile(), which is now just a
1422 wrapper). This facility is provided because some applications need a
1423 numeric error indication, but it has also enabled me to tidy up the way
1424 compile-time errors are handled in the POSIX wrapper.
1426 16. Added VPATH=.libs to the makefile; this should help when building with one
1427 prefix path and installing with another. (Or so I'm told by someone who
1428 knows more about this stuff than I do.)
1430 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1431 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1432 match everything, including newlines. This is not POSIX-compatible, but
1433 somebody wanted the feature. From pcretest it can be activated by using
1434 both the P and the s flags.
1436 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1438 19. libpcre.pc was being incorrectly installed as executable.
1440 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1441 it now also looks for '\r' so that it will work unmodified on Windows.
1443 21. Added Google's contributed C++ wrapper to the distribution.
1445 22. Added some untidy missing memory free() calls in pcretest, to keep
1446 Electric Fence happy when testing.
1450 Version 5.0 13-Sep-04
1451 ---------------------
1453 1. Internal change: literal characters are no longer packed up into items
1454 containing multiple characters in a single byte-string. Each character
1455 is now matched using a separate opcode. However, there may be more than one
1456 byte in the character in UTF-8 mode.
1458 2. The pcre_callout_block structure has two new fields: pattern_position and
1459 next_item_length. These contain the offset in the pattern to the next match
1460 item, and its length, respectively.
1462 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1463 insertion of callouts before each pattern item. Added the /C option to
1464 pcretest to make use of this.
1466 4. On the advice of a Windows user, the lines
1468 #if defined(_WIN32) || defined(WIN32)
1469 _setmode( _fileno( stdout ), 0x8000 );
1470 #endif /* defined(_WIN32) || defined(WIN32) */
1472 have been added to the source of pcretest. This apparently does useful
1473 magic in relation to line terminators.
1475 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1476 for the benefit of those environments where the "b" makes a difference.
1478 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1479 to know about it. I have put a hack into configure.in that adds in code
1480 to set GCC=yes if CC=icc. This seems to end up at a point in the
1481 generated configure script that is early enough to affect the setting of
1482 compiler options, which is what is needed, but I have no means of testing
1483 whether it really works. (The user who reported this had patched the
1484 generated configure script, which of course I cannot do.)
1486 LATER: After change 22 below (new libtool files), the configure script
1487 seems to know about icc (and also ecc). Therefore, I have commented out
1488 this hack in configure.in.
1490 7. Added support for pkg-config (2 patches were sent in).
1492 8. Negated POSIX character classes that used a combination of internal tables
1493 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1494 [[:^ascii]]. Typically, they would match almost any characters. The other
1495 POSIX classes were not broken in this way.
1497 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1498 to find the match, as PCRE was deluded into thinking that the match had to
1499 start at the start point or following a newline. The same bug applied to
1500 patterns with negative forward assertions or any backward assertions
1501 preceding ".*" at the start, unless the pattern required a fixed first
1502 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1504 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1505 starting at the last subject character, bytes beyond the end of the subject
1506 string were read.
1508 11. Renamed the variable "class" as "classbits" to make life easier for C++
1509 users. (Previously there was a macro definition, but it apparently wasn't
1510 enough.)
1512 12. Added the new field "tables" to the extra data so that tables can be passed
1513 in at exec time, or the internal tables can be re-selected. This allows
1514 a compiled regex to be saved and re-used at a later time by a different
1515 program that might have everything at different addresses.
1517 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1518 -R library as well as a -L library.
1520 14. The debugging options of pcretest (-d on the command line or D on a
1521 pattern) showed incorrect output for anything following an extended class
1522 that contained multibyte characters and which was followed by a quantifier.
1524 15. Added optional support for general category Unicode character properties
1525 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1526 support. It adds about 90K to the size of the library. The meanings of the
1527 inbuilt class escapes such as \d and \s have NOT been changed.
1529 16. Updated pcredemo.c to include calls to free() to release the memory for the
1530 compiled pattern.
1532 17. The generated file chartables.c was being created in the source directory
1533 instead of in the building directory. This caused the build to fail if the
1534 source directory was different from the building directory, and was
1535 read-only.
1537 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1538 file. No doubt somebody will tell me if they don't make sense... Also added
1539 Dan Mooney's comments about building on OpenVMS.
1541 19. Added support for partial matching via the PCRE_PARTIAL option for
1542 pcre_exec() and the \P data escape in pcretest.
1544 20. Extended pcretest with 3 new pattern features:
1546 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1547 write the compiled pattern to the file whose name is "rest-of-line".
1548 This is a straight binary dump of the data, with the saved pointer to
1549 the character tables forced to be NULL. The study data, if any, is
1550 written too. After writing, pcretest reads a new pattern.
1552 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1553 compiled pattern from the given file. There must not be any
1554 occurrences of "<" in the file name (pretty unlikely); if there are,
1555 pcretest will instead treat the initial "<" as a pattern delimiter.
1556 After reading in the pattern, pcretest goes on to read data lines as
1557 usual.
1559 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1560 and 16-bit fields in a compiled pattern, to simulate a pattern that
1561 was compiled on a host of opposite endianness.
1563 21. The pcre-exec() function can now cope with patterns that were compiled on
1564 hosts of opposite endianness, with this restriction:
1566 As for any compiled expression that is saved and used later, the tables
1567 pointer field cannot be preserved; the extra_data field in the arguments
1568 to pcre_exec() should be used to pass in a tables address if a value
1569 other than the default internal tables were used at compile time.
1571 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1572 now diagnosed as an error. Previously, most of the time, a negative number
1573 would have been treated as zero, but if in addition "ovector" was passed as
1574 NULL, a crash could occur.
1576 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1577 new versions from the libtool 1.5 distribution (the last one is a copy of
1578 a file called libtool.m4). This seems to have fixed the need to patch
1579 "configure" to support Darwin 1.3 (which I used to do). However, I still
1580 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1581 workstation).
1583 24. Changed the PCRE licence to be the more standard "BSD" licence.
1586 Version 4.5 01-Dec-03
1587 ---------------------
1589 1. There has been some re-arrangement of the code for the match() function so
1590 that it can be compiled in a version that does not call itself recursively.
1591 Instead, it keeps those local variables that need separate instances for
1592 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1593 needs to "recurse". Keeping track of where control must go is done by means
1594 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1595 hide most of the details from the main code, and operates only if
1596 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1597 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1598 operating.
1600 To make it easier for callers to provide specially tailored get/free
1601 functions for this usage, two new functions, pcre_stack_malloc, and
1602 pcre_stack_free, are used. They are always called in strict stacking order,
1603 and the size of block requested is always the same.
1605 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1606 PCRE has been compiled to use the stack or the heap for recursion. The
1607 -C option of pcretest uses this to show which version is compiled.
1609 A new data escape \S, is added to pcretest; it causes the amounts of store
1610 obtained and freed by both kinds of malloc/free at match time to be added
1611 to the output.
1613 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1614 what's available on my current Linux desktop machine.
1616 3. When matching a UTF-8 string, the test for a valid string at the start has
1617 been extended. If start_offset is not zero, PCRE now checks that it points
1618 to a byte that is the start of a UTF-8 character. If not, it returns
1619 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1620 this is necessary because there may be backward assertions in the pattern.
1621 When matching the same subject several times, it may save resources to use
1622 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1624 4. The code for checking the validity of UTF-8 strings has been tightened so
1625 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1626 containing "overlong sequences".
1628 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1629 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1630 should have been "&", but it just so happened that all the cases this let
1631 through by mistake were picked up later in the function.
1633 6. I had used a variable called "isblank" - this is a C99 function, causing
1634 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1636 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1637 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1638 so the version is shown; (c) stop "make test" echoing "./RunTest".
1640 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1642 9. The prototype for memmove() for systems that don't have it was using
1643 size_t, but the inclusion of the header that defines size_t was later. I've
1644 moved the #includes for the C headers earlier to avoid this.
1646 10. Added some adjustments to the code to make it easier to compiler on certain
1647 special systems:
1649 (a) Some "const" qualifiers were missing.
1650 (b) Added the macro EXPORT before all exported functions; by default this
1651 is defined to be empty.
1652 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1653 that it reads its output file name as an argument instead of writing
1654 to the standard output and assuming this can be redirected.
1656 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1657 class containing characters with values greater than 255, PCRE compilation
1658 went into a loop.
1660 12. A recursive reference to a subpattern that was within another subpattern
1661 that had a minimum quantifier of zero caused PCRE to crash. For example,
1662 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1663 recursion. If the recursively-called subpattern itself had a zero repeat,
1664 that was OK.
1666 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1667 buffer into which it was copied (for escape processing) was still set at
1668 1024, so long lines caused crashes.
1670 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1671 "internal error: code overflow...". This applied to any character class
1672 that was followed by a possessive quantifier.
1674 15. Modified the Makefile to add libpcre.la as a prerequisite for
1675 libpcreposix.la because I was told this is needed for a parallel build to
1676 work.
1678 16. If a pattern that contained .* following optional items at the start was
1679 studied, the wrong optimizing data was generated, leading to matching
1680 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1681 matching string must start with a or b or c. The correct conclusion for
1682 this pattern is that a match can start with any character.
1685 Version 4.4 13-Aug-03
1686 ---------------------
1688 1. In UTF-8 mode, a character class containing characters with values between
1689 127 and 255 was not handled correctly if the compiled pattern was studied.
1690 In fixing this, I have also improved the studying algorithm for such
1691 classes (slightly).
1693 2. Three internal functions had redundant arguments passed to them. Removal
1694 might give a very teeny performance improvement.
1696 3. Documentation bug: the value of the capture_top field in a callout is *one
1697 more than* the number of the hightest numbered captured substring.
1699 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1700 in incorrectly linking with a previously installed version. They now link
1701 explicitly with libpcre.la.
1703 5. configure.in no longer needs to recognize Cygwin specially.
1705 6. A problem in pcre.in for Windows platforms is fixed.
1707 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1708 pcretest, it used to include the size of the study block as part of its
1709 output. Unfortunately, the structure contains a field that has a different
1710 size on different hardware architectures. This meant that the tests that
1711 showed this size failed. As the block is currently always of a fixed size,
1712 this information isn't actually particularly useful in pcretest output, so
1713 I have just removed it.
1715 8. Three pre-processor statements accidentally did not start in column 1.
1716 Sadly, there are *still* compilers around that complain, even though
1717 standard C has not required this for well over a decade. Sigh.
1719 9. In pcretest, the code for checking callouts passed small integers in the
1720 callout_data field, which is a void * field. However, some picky compilers
1721 complained about the casts involved for this on 64-bit systems. Now
1722 pcretest passes the address of the small integer instead, which should get
1723 rid of the warnings.
1725 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1726 both compile and run time, and gives an error if an invalid UTF-8 sequence
1727 is found. There is a option for disabling this check in cases where the
1728 string is known to be correct and/or the maximum performance is wanted.
1730 11. In response to a bug report, I changed one line in Makefile.in from
1732 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1733 to
1734 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1736 to look similar to other lines, but I have no way of telling whether this
1737 is the right thing to do, as I do not use Windows. No doubt I'll get told
1738 if it's wrong...
1741 Version 4.3 21-May-03
1742 ---------------------
1744 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1745 Makefile.
1747 2. Some refactoring to improve the quality of the code:
1749 (i) The utf8_table... variables are now declared "const".
1751 (ii) The code for \cx, which used the "case flipping" table to upper case
1752 lower case letters, now just substracts 32. This is ASCII-specific,
1753 but the whole concept of \cx is ASCII-specific, so it seems
1754 reasonable.
1756 (iii) PCRE was using its character types table to recognize decimal and
1757 hexadecimal digits in the pattern. This is silly, because it handles
1758 only 0-9, a-f, and A-F, but the character types table is locale-
1759 specific, which means strange things might happen. A private
1760 table is now used for this - though it costs 256 bytes, a table is
1761 much faster than multiple explicit tests. Of course, the standard
1762 character types table is still used for matching digits in subject
1763 strings against \d.
1765 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1766 ending in _t are). So I've renamed it as ESC_tee.
1768 3. The first argument for regexec() in the POSIX wrapper should have been
1769 defined as "const".
1771 4. Changed pcretest to use malloc() for its buffers so that they can be
1772 Electric Fenced for debugging.
1774 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1775 to read one or more bytes before the start of the subject string. Often this
1776 had no effect on PCRE's behaviour, but in some circumstances it could
1777 provoke a segmentation fault.
1779 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1780 to try to read one or more bytes before the start of the subject string.
1782 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1783 UTF-8 support could misbehave in various ways if the subject string
1784 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1785 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1786 back over UTF-8 characters.)
1789 Version 4.2 14-Apr-03
1790 ---------------------
1792 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1794 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1795 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1796 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1797 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1798 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1800 Note: automatic setting of the BUILD variables is not yet working
1801 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1802 compile-time but not at link-time
1803 [LINK]: use for linking executables only
1804 make different versions for Windows and non-Windows
1805 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1806 libraries
1807 [LINK_FOR_BUILD]: new variable
1808 [OBJEXT]: use throughout
1809 [EXEEXT]: use throughout
1810 <winshared>: new target
1811 <wininstall>: new target
1812 <dftables.o>: use native compiler
1813 <dftables>: use native linker
1814 <install>: handle Windows platform correctly
1815 <clean>: ditto
1816 <check>: ditto
1817 copy DLL to top builddir before testing
1819 As part of these changes, -no-undefined was removed again. This was reported
1820 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1821 in any case.
1823 3. Some tidies to get rid of compiler warnings:
1825 . In the match_data structure, match_limit was an unsigned long int, whereas
1826 match_call_count was an int. I've made them both unsigned long ints.
1828 . In pcretest the fact that a const uschar * doesn't automatically cast to
1829 a void * provoked a warning.
1831 . Turning on some more compiler warnings threw up some "shadow" variables
1832 and a few more missing casts.
1834 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1835 option, a class that contained a single character with a value between 128
1836 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1838 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1839 option, a class that contained several characters, but with at least one
1840 whose value was between 128 and 255 caused PCRE to crash.
1843 Version 4.1 12-Mar-03
1844 ---------------------
1846 1. Compiling with gcc -pedantic found a couple of places where casts were
1847 needed, and a string in dftables.c that was longer than standard compilers are
1848 required to support.
1850 2. Compiling with Sun's compiler found a few more places where the code could
1851 be tidied up in order to avoid warnings.
1853 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1854 first of these names is deprecated in the latest Autoconf in favour of the name
1855 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1856 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1857 analogy I have changed it to CFLAGS_FOR_BUILD.
1859 4. Added -no-undefined to the linking command in the Makefile, because this is
1860 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1861 linking step for the pcreposix library.
1863 5. PCRE was failing to diagnose the case of two named groups with the same
1864 name.
1866 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1867 literal character that is needed in the subject for a match, and scans along to
1868 ensure that it is present before embarking on the full matching process. This
1869 saves time in cases of nested unlimited repeats that are never going to match.
1870 Problem: the scan can take a lot of time if the subject is very long (e.g.
1871 megabytes), thus penalizing straightforward matches. It is now done only if the
1872 amount of subject to be scanned is less than 1000 bytes.
1874 7. A lesser problem with the same optimization is that it was recording the
1875 first character of an anchored pattern as "needed", thus provoking a search
1876 right along the subject, even when the first match of the pattern was going to
1877 fail. The "needed" character is now not set for anchored patterns, unless it
1878 follows something in the pattern that is of non-fixed length. Thus, it still
1879 fulfils its original purpose of finding quick non-matches in cases of nested
1880 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1883 Version 4.0 17-Feb-03
1884 ---------------------
1886 1. If a comment in an extended regex that started immediately after a meta-item
1887 extended to the end of string, PCRE compiled incorrect data. This could lead to
1888 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1890 2. Moved to autoconf 2.53 and libtool 1.4.2.
1892 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1893 the special perltest8 script is no longer needed - all the tests can be run
1894 from a single perltest script.
1896 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1897 by \s. It has now been removed in PCRE. This means it isn't recognized as
1898 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1899 class [:space:] *does* include VT, thereby creating a mess.
1901 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1902 space and tab.
1904 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1905 its new features into the main test script, reducing the number of scripts.
1907 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1908 were backward compatible, and made the (?i) apply to the whole pattern, as if
1909 /i were given. Now it behaves more logically, and applies the option setting
1910 only to what follows. PCRE has been changed to follow suit. However, if it
1911 finds options settings right at the start of the pattern, it extracts them into
1912 the global options, as before. Thus, they show up in the info data.
1914 8. Added support for the \Q...\E escape sequence. Characters in between are
1915 treated as literals. This is slightly different from Perl in that $ and @ are
1916 also handled as literals inside the quotes. In Perl, they will cause variable
1917 interpolation. Note the following examples:
1919 Pattern PCRE matches Perl matches
1921 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1922 \Qabc\$xyz\E abc\$xyz abc\$xyz
1923 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1925 For compatibility with Perl, \Q...\E sequences are recognized inside character
1926 classes as well as outside them.
1928 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1929 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1930 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1931 signed/unsigned warnings.
1933 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1934 option for pcretest, so I've replaced it by a simple function that does just
1935 that job.
1937 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1938 "pcregrep -".
1940 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1941 Java package. This provides some syntactic sugar for simple cases of what my
1942 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1943 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1944 item, you can use this simplified notation. Note that only makes sense with
1945 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1946 greediness, whatever the setting of the PCRE_UNGREEDY option.
1948 13. A change of greediness default within a pattern was not taking effect at
1949 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1950 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1951 was abstracted outside.
1953 14. PCRE now supports the \G assertion. It is true when the current matching
1954 position is at the start point of the match. This differs from \A when the
1955 starting offset is non-zero. Used with the /g option of pcretest (or similar
1956 code), it works in the same way as it does for Perl's /g option. If all
1957 alternatives of a regex begin with \G, the expression is anchored to the start
1958 match position, and the "anchored" flag is set in the compiled expression.
1960 15. Some bugs concerning the handling of certain option changes within patterns
1961 have been fixed. These applied to options other than (?ims). For example,
1962 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1963 been the other way round. Some of this was related to change 7 above.
1965 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1966 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1967 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1968 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1970 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1971 mode. Unlike ".", it always matches newline, whatever the setting of
1972 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1973 assertions. Perl allows it, but it doesn't (in general) work because it can't
1974 calculate the length of the lookbehind. At least, that's the case for Perl
1975 5.8.0 - I've been told they are going to document that it doesn't work in
1976 future.
1978 18. Added an error diagnosis for escapes that PCRE does not support: these are
1979 \L, \l, \N, \P, \p, \U, \u, and \X.
1981 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1982 reading past the end of the pattern in cases such as /[abcd/.
1984 20. PCRE was getting more memory than necessary for patterns with classes that
1985 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1987 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1988 compiling PCRE for use with Virtual Pascal.
1990 22. Small fix to the Makefile to make it work properly if the build is done
1991 outside the source tree.
1993 23. Added a new extension: a condition to go with recursion. If a conditional
1994 subpattern starts with (?(R) the "true" branch is used if recursion has
1995 happened, whereas the "false" branch is used only at the top level.
1997 24. When there was a very long string of literal characters (over 255 bytes
1998 without UTF support, over 250 bytes with UTF support), the computation of how
1999 much memory was required could be incorrect, leading to segfaults or other
2000 strange effects.
2002 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2003 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2004 there was a subsequent back reference to those brackets. This meant that, for
2005 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2006 possible to check for precisely this case. All we can do is abandon the
2007 optimization if .* occurs inside capturing brackets when there are any back
2008 references whatsoever. (See below for a better fix that came later.)
2010 26. The handling of the optimization for finding the first character of a
2011 non-anchored pattern, and for finding a character that is required later in the
2012 match were failing in some cases. This didn't break the matching; it just
2013 failed to optimize when it could. The way this is done has been re-implemented.
2015 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2017 28. Added a new feature that provides some of the functionality that Perl
2018 provides with (?{...}). The facility is termed a "callout". The way it is done
2019 in PCRE is for the caller to provide an optional function, by setting
2020 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2021 global variable. By default it is unset, which disables all calling out. To get
2022 the function called, the regex must include (?C) at appropriate points. This
2023 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2024 This provides a means of identifying different callout points. When PCRE
2025 reaches such a point in the regex, if pcre_callout has been set, the external
2026 function is called. It is provided with data in a structure called
2027 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2028 matching continues; if it returns a non-zero value, the match at the current
2029 point fails. However, backtracking will occur if possible. [This was changed
2030 later and other features added - see item 49 below.]
2032 29. pcretest is upgraded to test the callout functionality. It provides a
2033 callout function that displays information. By default, it shows the start of
2034 the match and the current position in the text. There are some new data escapes
2035 to vary what happens:
2037 \C+ in addition, show current contents of captured substrings
2038 \C- do not supply a callout function
2039 \C!n return 1 when callout number n is reached
2040 \C!n!m return 1 when callout number n is reached for the mth time
2042 30. If pcregrep was called with the -l option and just a single file name, it
2043 output "<stdin>" if a match was found, instead of the file name.
2045 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2046 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2047 pcre_exec(). This saves a malloc/free per call. The default value of
2048 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2049 when configuring.
2051 32. The default maximum size of a compiled pattern is 64K. There have been a
2052 few cases of people hitting this limit. The code now uses macros to handle the
2053 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2054 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2055 configuring. Tests 2 and 5 work only with 2-byte links because they output
2056 debugging information about compiled patterns.
2058 33. Internal code re-arrangements:
2060 (a) Moved the debugging function for printing out a compiled regex into
2061 its own source file (printint.c) and used #include to pull it into
2062 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2063 separate copies.
2065 (b) Defined the list of op-code names for debugging as a macro in
2066 internal.h so that it is next to the definition of the opcodes.
2068 (c) Defined a table of op-code lengths for simpler skipping along compiled
2069 code. This is again a macro in internal.h so that it is next to the
2070 definition of the opcodes.
2072 34. Added support for recursive calls to individual subpatterns, along the
2073 lines of Robin Houston's patch (but implemented somewhat differently).
2075 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2076 allow it to read and process whole directories in Win32. This code was
2077 contributed by Lionel Fourquaux; it has not been tested by me.
2079 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2080 used to name a group. Names consist of alphanumerics and underscores, and must
2081 be unique. Back references use the syntax (?P=name) and recursive calls use
2082 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2083 numbers. The function pcre_fullinfo() can be used after compilation to extract
2084 a name/number map. There are three relevant calls:
2086 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2087 PCRE_INFO_NAMECOUNT yields the number of entries
2088 PCRE_INFO_NAMETABLE yields a pointer to the map.
2090 The map is a vector of fixed-size entries. The size of each entry depends on
2091 the length of the longest name used. The first two bytes of each entry are the
2092 group number, most significant byte first. There follows the corresponding
2093 name, zero terminated. The names are in alphabetical order.
2095 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2096 case instead of 255. Making it the same both with and without UTF-8 support
2097 means that the same test output works with both.
2099 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2100 calling malloc() with a zero argument.
2102 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2103 optimization. I've improved things by keeping a bitmap of backreferences with
2104 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2105 fact referenced, the optimization can be applied. It is unlikely that a
2106 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2107 the match to follow \n) will appear inside brackets with a number greater than
2108 31, but if it does, any back reference > 31 suppresses the optimization.
2110 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2111 of disabling numbered capturing parentheses. Any opening parenthesis that is
2112 not followed by ? behaves as if it were followed by ?: but named parentheses
2113 can still be used for capturing (and they will acquire numbers in the usual
2114 way).
2116 41. Redesigned the return codes from the match() function into yes/no/error so
2117 that errors can be passed back from deep inside the nested calls. A malloc
2118 failure while inside a recursive subpattern call now causes the
2119 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2121 42. It is now possible to set a limit on the number of times the match()
2122 function is called in a call to pcre_exec(). This facility makes it possible to
2123 limit the amount of recursion and backtracking, though not in a directly
2124 obvious way, because the match() function is used in a number of different
2125 circumstances. The count starts from zero for each position in the subject
2126 string (for non-anchored patterns). The default limit is, for compatibility, a
2127 large number, namely 10 000 000. You can change this in two ways:
2129 (a) When configuring PCRE before making, you can use --with-match-limit=n
2130 to set a default value for the compiled library.
2132 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2133 a different value is set. See 45 below.
2135 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2137 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2138 of things that can be changed at compile time. The first argument specifies
2139 what is wanted and the second points to where the information is to be placed.
2140 The current list of available information is:
2144 The output is an integer that is set to one if UTF-8 support is available;
2145 otherwise it is set to zero.
2149 The output is an integer that it set to the value of the code that is used for
2150 newline. It is either LF (10) or CR (13).
2154 The output is an integer that contains the number of bytes used for internal
2155 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2159 The output is an integer that contains the threshold above which the POSIX
2160 interface uses malloc() for output vectors. See item 31 above.
2164 The output is an unsigned integer that contains the default limit of the number
2165 of match() calls in a pcre_exec() execution. See 42 above.
2167 44. pcretest has been upgraded by the addition of the -C option. This causes it
2168 to extract all the available output from the new pcre_config() function, and to
2169 output it. The program then exits immediately.
2171 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2172 order to support additional features. One way would have been to define
2173 pcre_exec2() (for example) with extra arguments, but this would not have been
2174 extensible, and would also have required all calls to the original function to
2175 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2176 is used for passing in "extra" data from pcre_study().
2178 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2179 contains the following fields:
2181 flags a bitmap indicating which of the following fields are set
2182 study_data opaque data from pcre_study()
2183 match_limit a way of specifying a limit on match() calls for a specific
2184 call to pcre_exec()
2185 callout_data data for callouts (see 49 below)
2187 The flag bits are also defined in pcre.h, and are
2193 The pcre_study() function now returns one of these new pcre_extra blocks, with
2194 the actual study data pointed to by the study_data field, and the
2195 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2196 before. That is, this change is entirely upwards-compatible and requires no
2197 change to existing code.
2199 If you want to pass in additional data to pcre_exec(), you can either place it
2200 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2201 block.
2203 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2204 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2205 times with different match limits, until it finds the minimum value needed for
2206 pcre_exec() to complete. The value is then output. This can be instructive; for
2207 most simple matches the number is quite small, but for pathological cases it
2208 gets very large very quickly.
2210 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2211 returns the size of the data block pointed to by the study_data field in a
2212 pcre_extra block, that is, the value that was passed as the argument to
2213 pcre_malloc() when PCRE was getting memory in which to place the information
2214 created by pcre_study(). The fourth argument should point to a size_t variable.
2215 pcretest has been extended so that this information is shown after a successful
2216 pcre_study() call when information about the compiled regex is being displayed.
2218 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2219 because what follows is always an absolute path. (Later: it turns out that this
2220 is more than cosmetic for MinGW, because it doesn't like empty path
2221 components.)
2223 49. Some changes have been made to the callout feature (see 28 above):
2225 (i) A callout function now has three choices for what it returns:
2227 0 => success, carry on matching
2228 > 0 => failure at this point, but backtrack if possible
2229 < 0 => serious error, return this value from pcre_exec()
2231 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2232 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2233 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2234 use by callout functions. It will never be used by PCRE itself.
2236 (ii) The pcre_extra structure (see 45 above) has a void * field called
2237 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2238 pcre_callout_block structure has a field of the same name. The contents of
2239 the field passed in the pcre_extra structure are passed to the callout
2240 function in the corresponding field in the callout block. This makes it
2241 easier to use the same callout-containing regex from multiple threads. For
2242 testing, the pcretest program has a new data escape
2244 \C*n pass the number n (may be negative) as callout_data
2246 If the callout function in pcretest receives a non-zero value as
2247 callout_data, it returns that value.
2249 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2250 there were some redundant $(CFLAGS) in commands that are now specified as
2251 $(LINK), which already includes $(CFLAGS).
2253 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2254 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2255 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2256 one-byte characters throughout. Note that case-insensitive matching applies
2257 only to characters whose values are less than 256. PCRE doesn't support the
2258 notion of cases for higher-valued characters.
2260 (i) A character class whose characters are all within 0-255 is handled as
2261 a bit map, and the map is inverted for negative classes. Previously, a
2262 character > 255 always failed to match such a class; however it should
2263 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2265 (ii) A negated character class with a single character < 255 is coded as
2266 "not this character" (OP_NOT). This wasn't working properly when the test
2267 character was multibyte, either singly or repeated.
2269 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2270 mode, for example: \x{100}{2,3}.
2272 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2273 singly or repeated) now correctly test multibyte characters. However,
2274 PCRE doesn't recognize any characters with values greater than 255 as
2275 digits, spaces, or word characters. Such characters always match \D, \S,
2276 and \W, and never match \d, \s, or \w.
2278 (v) Classes may now contain characters and character ranges with values
2279 greater than 255. For example: [ab\x{100}-\x{400}].
2281 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2282 PCRE in UTF-8 mode.
2284 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2285 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2286 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2287 value.)
2289 53. The single man page has become too large. I have therefore split it up into
2290 a number of separate man pages. These also give rise to individual HTML pages;
2291 these are now put in a separate directory, and there is an index.html page that
2292 lists them all. Some hyperlinking between the pages has been installed.
2294 54. Added convenience functions for handling named capturing parentheses.
2296 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2297 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2298 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2299 are faulted.
2301 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2302 calling configure. These values are used when compiling the dftables.c program
2303 which is run to generate the source of the default character tables. They
2304 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2305 you will need to set these values.
2307 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2310 Version 3.9 02-Jan-02
2311 ---------------------
2313 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2315 2. If --disable-static was given, the building process failed when trying to
2316 build pcretest and pcregrep. (For some reason it was using libtool to compile
2317 them, which is not right, as they aren't part of the library.)
2320 Version 3.8 18-Dec-01
2321 ---------------------
2323 1. The experimental UTF-8 code was completely screwed up. It was packing the
2324 bytes in the wrong order. How dumb can you get?
2327 Version 3.7 29-Oct-01
2328 ---------------------
2330 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2331 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2332 this didn't happen under Solaris 8, where I normally test things.
2334 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2335 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2336 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2337 this fix an hour or so after the initial 3.7 release.)
2340 Version 3.6 23-Oct-01
2341 ---------------------
2343 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2344 offsets passed as NULL with zero offset count.
2346 2. The config.guess and config.sub files had not been updated when I moved to
2347 the latest autoconf.
2350 Version 3.5 15-Aug-01
2351 ---------------------
2353 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2354 had been forgotten.
2356 2. By using declared but undefined structures, we can avoid using "void"
2357 definitions in pcre.h while keeping the internal definitions of the structures
2358 private.
2360 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2361 user point of view, this means that both static and shared libraries are built
2362 by default, but this can be individually controlled. More of the work of
2363 handling this static/shared cases is now inside libtool instead of PCRE's make
2364 file.
2366 4. The pcretest utility is now installed along with pcregrep because it is
2367 useful for users (to test regexs) and by doing this, it automatically gets
2368 relinked by libtool. The documentation has been turned into a man page, so
2369 there are now .1, .txt, and .html versions in /doc.
2371 5. Upgrades to pcregrep:
2372 (i) Added long-form option names like gnu grep.
2373 (ii) Added --help to list all options with an explanatory phrase.
2374 (iii) Added -r, --recursive to recurse into sub-directories.
2375 (iv) Added -f, --file to read patterns from a file.
2377 6. pcre_exec() was referring to its "code" argument before testing that
2378 argument for NULL (and giving an error if it was NULL).
2380 7. Upgraded Makefile.in to allow for compiling in a different directory from
2381 the source directory.
2383 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2384 options bits, the pointer it was passed was to an int instead of to an unsigned
2385 long int. This mattered only on 64-bit systems.
2387 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2388 generated) instead of pcre.in, which it its source. Also made the same change
2389 in several of the .c files.
2391 10. A new release of gcc defines printf() as a macro, which broke pcretest
2392 because it had an ifdef in the middle of a string argument for printf(). Fixed
2393 by using separate calls to printf().
2395 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2396 script, to force use of CR or LF instead of \n in the source. On non-Unix
2397 systems, the value can be set in config.h.
2399 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2400 absolute limit. Changed the text of the error message to make this clear, and
2401 likewise updated the man page.
2403 13. The limit of 99 on the number of capturing subpatterns has been removed.
2404 The new limit is 65535, which I hope will not be a "real" limit.
2407 Version 3.4 22-Aug-00
2408 ---------------------
2410 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2412 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2415 Version 3.3 01-Aug-00
2416 ---------------------
2418 1. If an octal character was given, but the value was greater than \377, it
2419 was not getting masked to the least significant bits, as documented. This could
2420 lead to crashes in some systems.
2422 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2423 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2425 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2426 These just pass their arguments on to (pcre_free)(), but they are provided
2427 because some uses of PCRE bind it to non-C systems that can call its functions,
2428 but cannot call free() or pcre_free() directly.
2430 4. Add "make test" as a synonym for "make check". Corrected some comments in
2431 the Makefile.
2433 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2434 Makefile.
2436 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2437 command called pgrep for grepping around the active processes.
2439 7. Added the beginnings of support for UTF-8 character strings.
2441 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2442 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2443 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2444 out for the ar command.)
2447 Version 3.2 12-May-00
2448 ---------------------
2450 This is purely a bug fixing release.
2452 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2453 of ZA. This was just one example of several cases that could provoke this bug,
2454 which was introduced by change 9 of version 2.00. The code for breaking
2455 infinite loops after an iteration that matches an empty string was't working
2456 correctly.
2458 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2459 when matched against abbab (for example). After matching an empty string, it
2460 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2461 caused it to match further down the string than it should.
2463 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2464 was there because it doesn't seem to be needed, and it causes trouble on some
2465 systems, as it is not a Standard C header. It has been removed.
2467 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2468 were reported on the Macintosh. The changes were from
2470 while ((c = *(++ptr)) != 0 && c != '\n');
2471 to
2472 while ((c = *(++ptr)) != 0 && c != '\n') ;
2474 Totally extraordinary, but if that's what it takes...
2476 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2477 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2478 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2479 assumes the way PCRE uses memmove() (always moving upwards).
2481 6. PCRE is being used in one environment where strchr() is not available. There
2482 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2483 faster code anyway.
2486 Version 3.1 09-Feb-00
2487 ---------------------
2489 The only change in this release is the fixing of some bugs in Makefile.in for
2490 the "install" target:
2492 (1) It was failing to install pcreposix.h.
2494 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2497 Version 3.0 01-Feb-00
2498 ---------------------
2500 1. Add support for the /+ modifier to perltest (to output $` like it does in
2501 pcretest).
2503 2. Add support for the /g modifier to perltest.
2505 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2506 matches null strings.
2508 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2509 pattern. Perl treats empty patterns specially - it reuses the most recent
2510 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2511 effect.
2513 5. The POSIX interface was broken in that it was just handing over the POSIX
2514 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2515 required a bigger vector, with some working space on the end. This means that
2516 the POSIX wrapper now has to get and free some memory, and copy the results.
2518 6. Added some simple autoconf support, placing the test data and the
2519 documentation in separate directories, re-organizing some of the
2520 information files, and making it build pcre-config (a GNU standard). Also added
2521 libtool support for building PCRE as a shared library, which is now the
2522 default.
2524 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2525 09 are not valid octal constants. Single digits will be used for minor values
2526 less than 10.
2528 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2529 existing programs that set these in the POSIX interface can use PCRE without
2530 modification.
2532 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2533 return all that pcre_info() returns, plus additional data. The pcre_info()
2534 function is retained for compatibility, but is considered to be obsolete.
2536 10. Added experimental recursion feature (?R) to handle one common case that
2537 Perl 5.6 will be able to do with (?p{...}).
2539 11. Added support for POSIX character classes like [:alpha:], which Perl is
2540 adopting.
2543 Version 2.08 31-Aug-99
2544 ----------------------
2546 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2547 trying to match at the startoffset position, but instead was moving forward to
2548 the next newline as if a previous match had failed.
2550 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2551 and could get into a loop if a null string was matched other than at the start
2552 of the subject.
2554 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2555 be distinguished at compile time, and for completeness also added PCRE_DATE.
2557 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2558 in GnuWin32 environments.
2561 Version 2.07 29-Jul-99
2562 ----------------------
2564 1. The documentation is now supplied in plain text form and HTML as well as in
2565 the form of man page sources.
2567 2. C++ compilers don't like assigning (void *) values to other pointer types.
2568 In particular this affects malloc(). Although there is no problem in Standard
2569 C, I've put in casts to keep C++ compilers happy.
2571 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2572 should be (const char *).
2574 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2575 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2576 However, I haven't made this a standard facility. The documentation doesn't
2577 mention it, and the Makefile doesn't support it.
2579 5. The Makefile now contains an "install" target, with editable destinations at
2580 the top of the file. The pcretest program is not installed.
2582 6. pgrep -V now gives the PCRE version number and date.
2584 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2585 causing the entire string to be ignored, instead of just the last character.
2587 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2588 non-matching string, it can take a very, very long time, even for strings of
2589 quite modest length, because of the nested recursion. PCRE now does better in
2590 some of these cases. It does this by remembering the last required literal
2591 character in the pattern, and pre-searching the subject to ensure it is present
2592 before running the real match. In other words, it applies a heuristic to detect
2593 some types of certain failure quickly, and in the above example, if presented
2594 with a string that has no trailing " it gives "no match" very quickly.
2596 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2597 other alternatives are tried instead.
2600 Version 2.06 09-Jun-99
2601 ----------------------
2603 1. Change pcretest's output for amount of store used to show just the code
2604 space, because the remainder (the data block) varies in size between 32-bit and
2605 64-bit systems.
2607 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2608 start matching at. This allows lookbehinds to work when searching for multiple
2609 occurrences in a string.
2611 3. Added additional options to pcretest for testing multiple occurrences:
2613 /+ outputs the rest of the string that follows a match
2614 /g loops for multiple occurrences, using the new startoffset argument
2615 /G loops for multiple occurrences by passing an incremented pointer
2617 4. PCRE wasn't doing the "first character" optimization for patterns starting
2618 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2619 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2620 the letter 'x'. On long subject strings, this gives a significant speed-up.
2623 Version 2.05 21-Apr-99
2624 ----------------------
2626 1. Changed the type of magic_number from int to long int so that it works
2627 properly on 16-bit systems.
2629 2. Fixed a bug which caused patterns starting with .* not to work correctly
2630 when the subject string contained newline characters. PCRE was assuming
2631 anchoring for such patterns in all cases, which is not correct because .* will
2632 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2633 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2634 must be retried after every newline in the subject.
2637 Version 2.04 18-Feb-99
2638 ----------------------
2640 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2641 computation of the store needed to hold the pattern was incorrect (too large).
2642 If such patterns were nested a few deep, this could multiply and become a real
2643 problem.
2645 2. Added /M option to pcretest to show the memory requirement of a specific
2646 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2648 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2649 compiled in such a way that the backtracking after subsequent failure was
2650 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2651 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2654 Version 2.03 02-Feb-99
2655 ----------------------
2657 1. Fixed typo and small mistake in man page.
2659 2. Added 4th condition (GPL supersedes if conflict) and created separate
2660 LICENCE file containing the conditions.
2662 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2663 Perl, that is the internal \ allows the delimiter to be included in the
2664 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2665 the final delimiter, add \ to the end of the pattern (to test the error).
2667 4. Added the convenience functions for extracting substrings after a successful
2668 match. Updated pcretest to make it able to test these functions.
2671 Version 2.02 14-Jan-99
2672 ----------------------
2674 1. Initialized the working variables associated with each extraction so that
2675 their saving and restoring doesn't refer to uninitialized store.
2677 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2678 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2679 fix the problem.
2681 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2682 calls, and wasn't printing the correct value for compiling calls. Increased the
2683 default value of LOOPREPEAT, and the number of significant figures in the
2684 times.
2686 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2688 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2689 a building problem on Windows NT with a FAT file system.
2692 Version 2.01 21-Oct-98
2693 ----------------------
2695 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2696 to character tables built by pcre_maketables() in the current locale. If NULL
2697 is passed, the default tables are used.
2700 Version 2.00 24-Sep-98
2701 ----------------------
2703 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2704 it any more.
2706 2. Allow quantification of (?>) groups, and make it work correctly.
2708 3. The first character computation wasn't working for (?>) groups.
2710 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2711 end of the subject) and add 5.005's \z, which really does match only at the
2712 very end of the subject.
2714 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2716 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2717 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2718 localized options. All options to pcre_study() were also removed.
2720 7. Add other new features from 5.005:
2722 $(?<= positive lookbehind
2723 $(?<! negative lookbehind
2724 (?imsx-imsx) added the unsetting capability
2725 such a setting is global if at outer level; local otherwise
2726 (?imsx-imsx:) non-capturing groups with option setting
2727 (?(cond)re|re) conditional pattern matching
2729 A backreference to itself in a repeated group matches the previous
2730 captured string.
2732 8. General tidying up of studying (both automatic and via "study")
2733 consequential on the addition of new assertions.
2735 9. As in 5.005, unlimited repeated groups that could match an empty substring
2736 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2737 runtime if any iteration does actually match an empty substring.
2739 10. Include the RunTest script in the distribution.
2741 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2742 discrepancies, some of which were old and were also with respect to 5.004. They
2743 have now been fixed.
2746 Version 1.09 28-Apr-98
2747 ----------------------
2749 1. A negated single character class followed by a quantifier with a minimum
2750 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2751 program crashes, or just wrong answers. This did not apply to negated classes
2752 containing more than one character, or to minima other than one.
2755 Version 1.08 27-Mar-98
2756 ----------------------
2758 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2760 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2761 latter must appear before anything that relies on it in the pattern.
2764 Version 1.07 16-Feb-98
2765 ----------------------
2767 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2768 repeat of a potentially empty string).
2771 Version 1.06 23-Jan-98
2772 ----------------------
2774 1. Added Markus Oberhumer's little patches for C++.
2776 2. Literal strings longer than 255 characters were broken.
2779 Version 1.05 23-Dec-97
2780 ----------------------
2782 1. Negated character classes containing more than one character were failing if
2783 PCRE_CASELESS was set at run time.
2786 Version 1.04 19-Dec-97
2787 ----------------------
2789 1. Corrected the man page, where some "const" qualifiers had been omitted.
2791 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2792 input syntax.
2794 3. Fixed memory leak which occurred when a regex with back references was
2795 matched with an offsets vector that wasn't big enough. The temporary memory
2796 that is used in this case wasn't being freed if the match failed.
2798 4. Tidied pcretest to ensure it frees memory that it gets.
2800 5. Temporary memory was being obtained in the case where the passed offsets
2801 vector was exactly big enough.
2803 6. Corrected definition of offsetof() from change 5 below.
2805 7. I had screwed up change 6 below and broken the rules for the use of
2806 setjmp(). Now fixed.
2809 Version 1.03 18-Dec-97
2810 ----------------------
2812 1. A erroneous regex with a missing opening parenthesis was correctly
2813 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2814 on some systems.
2816 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2817 it was reported that one broken compiler failed on the former because "code" is
2818 also an independent variable.
2820 3. The erroneous regex a[]b caused an array overrun reference.
2822 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2823 fail on data ending with that character. (It was going on too far, and checking
2824 the next character, typically a binary zero.) This was specific to the
2825 optimized code for single-character negative classes.
2827 5. Added a contributed patch from the TIN world which does the following:
2829 + Add an undef for memmove, in case the the system defines a macro for it.
2831 + Add a definition of offsetof(), in case there isn't one. (I don't know
2832 the reason behind this - offsetof() is part of the ANSI standard - but
2833 it does no harm).
2835 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2836 most of the places where whitespace preceded '#'. I have given up and
2837 allowed the remaining 2 cases to be at the margin.
2839 + Rename some variables in pcre to eliminate shadowing. This seems very
2840 pedantic, but does no harm, of course.
2842 6. Moved the call to setjmp() into its own function, to get rid of warnings
2843 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2845 7. Constructs such as \d{8,} were compiling into the equivalent of
2846 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2847 outcome, but in this particular case used more store than had been allocated,
2848 which caused the bug to be discovered because it threw up an internal error.
2850 8. The debugging code in both pcre and pcretest for outputting the compiled
2851 form of a regex was going wrong in the case of back references followed by
2852 curly-bracketed repeats.
2855 Version 1.02 12-Dec-97
2856 ----------------------
2858 1. Typos in pcre.3 and comments in the source fixed.
2860 2. Applied a contributed patch to get rid of places where it used to remove
2861 'const' from variables, and fixed some signed/unsigned and uninitialized
2862 variable warnings.
2864 3. Added the "runtest" target to Makefile.
2866 4. Set default compiler flag to -O2 rather than just -O.
2869 Version 1.01 19-Nov-97
2870 ----------------------
2872 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2873 like /([ab]*)*/, that is, for classes with more than one character in them.
2875 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2876 as /((?>a*))*/ (a PCRE_EXTRA facility).
2879 Version 1.00 18-Nov-97
2880 ----------------------
2882 1. Added compile-time macros to support systems such as SunOS4 which don't have
2883 memmove() or strerror() but have other things that can be used instead.
2885 2. Arranged that "make clean" removes the executables.
2888 Version 0.99 27-Oct-97
2889 ----------------------
2891 1. Fixed bug in code for optimizing classes with only one character. It was
2892 initializing a 32-byte map regardless, which could cause it to run off the end
2893 of the memory it had got.
2895 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2898 Version 0.98 22-Oct-97
2899 ----------------------
2901 1. Fixed bug in code for handling temporary memory usage when there are more
2902 back references than supplied space in the ovector. This could cause segfaults.
2905 Version 0.97 21-Oct-97
2906 ----------------------
2908 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2910 2. Optimized negated single characters not to use a bit map.
2912 3. Brought error texts together as macro definitions; clarified some of them;
2913 fixed one that was wrong - it said "range out of order" when it meant "invalid
2914 escape sequence".
2916 4. Changed some char * arguments to const char *.
2918 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2920 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2921 pcretest.
2924 Version 0.96 16-Oct-97
2925 ----------------------
2927 1. Added a simple "pgrep" utility to the distribution.
2929 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2930 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2931 where "ddd" means "one or more decimal digits".
2933 3. Fixed serious bug. If a pattern had a back reference, but the call to
2934 pcre_exec() didn't supply a large enough ovector to record the related
2935 identifying subpattern, the match always failed. PCRE now remembers the number
2936 of the largest back reference, and gets some temporary memory in which to save
2937 the offsets during matching if necessary, in order to ensure that
2938 backreferences always work.
2940 4. Increased the compatibility with Perl in a number of ways:
2942 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2943 to request this handling. The option can be set at compile or exec time.
2945 (b) $ matches before a terminating newline by default; an option
2946 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2947 mode). The option can be set at compile or exec time.
2949 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2950 the same as Perl's. If the decimal number it represents is less than 10
2951 or there aren't that many previous left capturing parentheses, an octal
2952 escape is read. Inside a character class, it's always an octal escape,
2953 even if it is a single digit.
2955 (d) An escaped but undefined alphabetic character is taken as a literal,
2956 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2957 escapes.
2959 (e) {0} is now permitted. (The previous item is removed from the compiled
2960 pattern).
2962 5. Changed all the names of code files so that the basic parts are no longer
2963 than 10 characters, and abolished the teeny "globals.c" file.
2965 6. Changed the handling of character classes; they are now done with a 32-byte
2966 bit map always.
2968 7. Added the -d and /D options to pcretest to make it possible to look at the
2969 internals of compilation without having to recompile pcre.
2972 Version 0.95 23-Sep-97
2973 ----------------------
2975 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2976 \x20 at the start of a run of normal characters. These were being treated as
2977 real characters, instead of the source characters being re-checked.
2980 Version 0.94 18-Sep-97
2981 ----------------------
2983 1. The functions are now thread-safe, with the caveat that the global variables
2984 containing pointers to malloc() and free() or alternative functions are the
2985 same for all threads.
2987 2. Get pcre_study() to generate a bitmap of initial characters for non-
2988 anchored patterns when this is possible, and use it if passed to pcre_exec().
2991 Version 0.93 15-Sep-97
2992 ----------------------
2994 1. /(b)|(:+)/ was computing an incorrect first character.
2996 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2997 but not actually doing anything yet.
2999 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3000 as Perl does (e.g. [-az] or [az-]).
3002 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3003 all possible positions.
3005 5. Split up into different modules to avoid including unneeded functions in a
3006 compiled binary. However, compile and exec are still in one module. The "study"
3007 function is split off.
3009 6. The character tables are now in a separate module whose source is generated
3010 by an auxiliary program - but can then be edited by hand if required. There are
3011 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3012 toupper() in the code.
3014 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3015 make them global. Abolish the function for setting them, as the caller can now
3016 set them directly.
3019 Version 0.92 11-Sep-97
3020 ----------------------
3022 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3023 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3025 2. Caseless matching was not working in character classes if the characters in
3026 the pattern were in upper case.
3028 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3030 4. Make PCRE_ANCHORED public and accept as a compile option.
3032 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3033 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3034 pass them.
3036 6. Give an error if bad option bits passed at compile or run time.
3038 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3039 pcretest to cause it to pass that flag.
3041 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3042 options, and the first character, if set.
3044 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3047 Version 0.91 10-Sep-97
3048 ----------------------
3050 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3051 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3053 2. PCRE was looping on encountering an indefinitely repeated back reference to
3054 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3055 Perl does - treats the match as successful.
3057 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5